Nancy Bartlett v. Tony Evers ( 2020 )


Menu:
  •                                                              
    2020 WI 68
    SUPREME COURT          OF   WISCONSIN
    CASE NO.:              2019AP1376-OA
    COMPLETE TITLE:        Nancy Bartlett, Richard Bowers, Jr. and Ted
    Keneklis,
    Petitioners,
    v.
    Tony Evers, in his official capacity as Governor
    of the State of Wisconsin, Joel Brennan, in his
    official capacity as Secretary of the Wisconsin
    Department of Administration, Wisconsin
    Department of Administration, Craig Thompson, in
    his official capacity as Secretary of
    the Wisconsin Department of Transportation,
    Wisconsin Department of Transportation, Peter
    Barca, in his official capacity as Secretary of
    the Wisconsin Department of Revenue, and
    Wisconsin Department of Revenue,
    Respondents.
    ORIGINAL ACTION
    OPINION FILED:         July 10, 2020
    SUBMITTED ON BRIEFS:
    ORAL ARGUMENT:         April 20, 2020
    SOURCE OF APPEAL:
    COURT:
    COUNTY:
    JUDGE:
    JUSTICES:
    ROGGENSACK, C.J., filed an opinion concurring in part and
    dissenting in part.    ANN WALSH BRADLEY, J., filed an opinion
    concurring in part and dissenting in part, in which DALLET, J.,
    joined.   KELLY, J., filed an opinion concurring in part and
    dissenting in part, in which REBECCA GRASSL BRADLEY, J. joined.
    HAGEGDORN, J., filed a concurring opinion, in which ZIEGLER, J.,
    joined.
    NOT PARTICIPATING:
    ATTORNEYS:
    For the petitioners, there were briefs filed by Richard M.
    Esenberg, Anthony LoCoco, Lucas T. Vebber, Luke N. Berg, and
    Wisconsin Institute for Law & Liberty, Milwaukee. There was an
    oral argument by Richard M. Esenberg.
    For the respondents, there were briefs filed by Colin T.
    Roth and Maura FJ Whelan, assistant attorneys general; with whom
    on the brief was Joshua L. Kaul, attorney general. There was an
    oral argument by Colin T. Roth.
    An   amicus    curiae    brief     was   filed    on   behalf    of   The
    Legislature   by   Misha   Tseytlin,     Kevin   M.   LeRoy,   and   Troutman
    Sanders LLP, Chicago, Illinois. There was an oral argument by
    Misha Tseytlin.
    2
    
    2020 WI 68
    NOTICE
    This opinion is subject to further
    editing and modification.   The final
    version will appear in the bound
    volume of the official reports.
    No.    2019AP1376-OA
    STATE OF WISCONSIN                       :            IN SUPREME COURT
    Nancy Bartlett, Richard Bowers, Jr. and
    Ted Keneklis,
    Petitioners,
    v.
    Tony Evers, in his official capacity as
    Governor of the State of Wisconsin,                             FILED
    Joel Brennan, in his official capacity as
    Secretary of the Wisconsin Department of                   JUL 10, 2020
    Administration, Wisconsin Department of
    Administration, Craig Thompson, in his official               Sheila T. Reiff
    capacity as Secretary of the Wisconsin                     Clerk of Supreme Court
    Department of Transportation, Wisconsin
    Department of Transportation, Peter Barca, in
    his official capacity as Secretary of the
    Wisconsin Department of Revenue, and Wisconsin
    Department of Revenue,
    Respondents.
    ¶1   PER   CURIAM.   We   review   the     petitioners'         original
    action requesting a declaration that Governor Evers exceeded his
    constitutional authority to partially veto appropriation bills.
    The petitioners assert that four series of partial vetoes in
    2019 Wis. Act 9——the state's 2019-21 biennial budget bill——are
    unconstitutional.
    No.     2019AP1376-OA
    ¶2     The parties refer to the provisions based on their
    content      before     the     vetoes:      (1) the        school    bus     modernization
    fund;       (2) the     local    roads       improvement        fund;       (3) the        vapor
    products tax and (4) the vehicle fee schedule.
    ¶3     The petitioners contend that the four series of vetoes
    are    unconstitutional.            Article          V,    Section     10(1)(b)       of    the
    Wisconsin Constitution provides that the governor may approve
    appropriation bills "in whole or in part."
    ¶4     No rationale has the support of a majority.                          However,
    a     majority    has     reached       a    conclusion        with     respect       to    the
    constitutionality         of     each       series    of     vetoes.        Five   justices
    conclude that the vetoes to the school bus modernization fund
    are unconstitutional.              The same five also                 conclude that the
    vetoes to the local roads improvement fund are unconstitutional.1
    Four justices conclude that the vetoes to the vapor products tax
    are unconstitutional.2            Five justices conclude that the vetoes to
    the vehicle fee schedule are constitutional.3
    ¶5     Chief Justice Roggensack concludes that the vetoes to
    the     school    bus      modernization             fund     and     the     local        roads
    Chief Justice Roggensack and Justices Ziegler, Rebecca
    1
    Grassl Bradley, Kelly and Hagedorn conclude that these series of
    vetoes are unconstitutional.
    Justices Ziegler, Rebecca Grassl Bradley, Kelly and
    2
    Hagedorn conclude that the vetoes to the vapor products tax are
    unconstitutional.
    Chief Justice Roggensack and Justices Ann Walsh Bradley,
    3
    Ziegler, Dallet and Hagedorn conclude that the vetoes to the
    vehicle fee schedule are constitutional.
    2
    No.    2019AP1376-OA
    improvement fund are unconstitutional because they "resulted in
    topics and subject matters that were not found in the enrolled
    bill."      Chief    Justice       Roggensack's      concurrence/dissent,          ¶99.
    She also concludes that the vetoes to the vapor products tax and
    vehicle fee schedule are constitutional because they did not
    alter "the topic or subject matter of the part approved."                          Id.,
    ¶106.
    ¶6    Justice Ann Walsh Bradley and Justice Dallet conclude
    that the four series of vetoes are constitutional because they
    "result[ed] in objectively complete, entire, and workable laws."
    Justice     Ann      Walsh       Bradley's       concurrence/dissent,             ¶170.
    Consequently, they would not grant relief.
    ¶7    Justice      Kelly      and    Justice      Rebecca    Grassl       Bradley
    conclude that the four series of vetoes are unconstitutional.
    Justice Kelly's concurrence/dissent, ¶230.                    They conclude that
    the   vetoes   violate       the    Wisconsin     Constitution's         origination
    clause,     amendment     clause          and   legislative       passage       clause.
    Id., ¶¶223, 225-26, 228.
    ¶8    Justice Hagedorn and Justice Ziegler conclude that the
    vetoes to the school bus modernization fund, the local roads
    improvement       fund       and      the       vapor      products       tax       are
    unconstitutional.         Justice         Hagedorn's     concurrence,       ¶¶269–75.
    They also conclude that the vetoes to the vehicle fee schedule
    are constitutional because they merely negated a policy proposal
    advanced by the legislature.              Id., ¶268.
    ¶9    Accordingly, rights are declared such that the vetoes
    to    the   school     bus     modernization         fund,    the       local    roads
    3
    No.    2019AP1376-OA
    improvement fund and the vapor products tax are unconstitutional
    and invalid.      Relief is granted such that the portions of the
    enrolled bills that were vetoed are in full force and effect as
    drafted   by   the   legislature.       See   State   ex   rel.    Sundby   v.
    Adamany, 
    71 Wis. 2d 118
    , 125, 
    237 N.W.2d 910
     (1976).              The vetoes
    to the vehicle fee schedule are constitutional, and no relief is
    granted with respect to these vetoes.
    By the Court.-Rights declared; relief granted in part and
    denied in part.
    4
    No.   2019AP1376-OA.pdr
    ¶10        PATIENCE DRAKE ROGGENSACK, C.J.                  (concurring in part,
    dissenting in part).                This is an original action brought by
    three    taxpayers, Nancy           Bartlett,    Richard        Bowers,    Jr.     and    Ted
    Keneklis       (Taxpayers)      against      Governor      Tony     Evers      and      other
    government       officials      and    agencies.         Taxpayers        challenge       the
    validity of several vetoes Governor Evers made to the 2019–21
    biennial       budget.1        Specifically,      they    challenge        a   series     of
    vetoes that changed a school bus modernization fund into an
    alternative fuel fund.              They also challenge another series that
    removed        conditions       from    a    local       road     improvement           fund,
    effectively changing it into a fund for "local grants" or "local
    supplements."        Third, they challenge a series of vetoes that
    altered a vehicle fee schedule by changing the amount truck
    owners    must     pay    to    register     their    vehicles.           Lastly,        they
    challenge one veto that altered a section that imposed a tax on
    "vapor products" by expanding the definition of vapor product to
    include liquid heated by a vaping device.                   Taxpayers assert that
    these    vetoes    went     beyond     the     governor's       partial     veto     power,
    which     is    provided       in    Article     V,   Section      10(1)(b)        of     the
    Wisconsin Constitution:             "Appropriation bills may be approved in
    whole or in part by the governor, and the part approved shall
    become law."
    ¶11        I conclude that the part approved by the governor,
    i.e., the consequences of the partial veto, must not alter the
    1  "The Wisconsin budget process covers two fiscal years at a
    time——a biennium."    Benjamin W. Proctor, Comment, Wisconsin's
    Chief Legislator: The Governor's Partial Veto Authority and the
    New Tipping Point, 
    90 Marq. L. Rev. 739
    , 739 n.3 (2007).
    1
    No.    2019AP1376-OA.pdr
    topic or subject matter of the "whole" bill before the veto.2
    Stated      otherwise,      such   a     veto    does    not      alter     the   stated
    legislative idea that initiated the enrolled bill.                           Therefore,
    Governor Evers could not use his partial veto power to change
    the school bus modernization fund into an alternative fuel fund.
    Nor could he use his partial veto to change the local road
    improvement         fund   into    a    fund     for    local     grants     or   local
    supplements, devoid of any requirements that it be used for
    local roads.          I partially concur with the per curiam opinion
    that these two series of vetoes are invalid and have no effect
    on the law enacted by the legislature.                         I further partially
    concur that he lawfully used his partial veto power to alter the
    amount      truck     owners   must      pay     to    register     their     vehicles.
    However, I partially dissent from the per curiam opinion because
    he also lawfully used his partial veto to alter the definition
    of vapor product.          This veto should stand.
    I.    BACKGROUND
    ¶12    On June 25 and 26, 2019, the Wisconsin State Assembly
    and   Senate,       respectively,       passed    the   2019–21       biennial    budget
    bill.      The enrolled bill was presented to Governor Evers, who
    signed it with several vetoes on July 3, 2019.3                           On July 31,
    2019, Taxpayers filed an original action, which was amended on
    2"Once identical versions of a bill pass both the state
    assembly and the state senate, the bill is referred to as an
    'enrolled bill' and is ready for the governor's consideration."
    
    Id.
     at 741 n.19.
    3   2019 Wis. Act 9.
    2
    No.   2019AP1376-OA.pdr
    August 19, 2019.         We took jurisdiction.          The legislature filed
    an amicus brief, generally supporting Taxpayers.
    A.    The School Bus Modernization Fund
    ¶13    The     first   series   of       vetoes   changed    a    school   bus
    modernization fund into an alternative fuel fund.                     For context,
    the State of Wisconsin is a beneficiary of a trust created by a
    consent    decree      following   litigation     against    Volkswagen.        The
    terms of the trust establish various permissible uses:
    [T]he state could utilize funding from the trust to
    scrap, and then repower or replace certain eligible
    vehicles and equipment, including: (a) Class 8 local
    freight trucks and port drayage trucks; (b) Class 4
    through 8 school buses, shuttle buses, or transit
    buses; (c) freight switchers; (d) ferries and tugs;
    (e) ocean going vessels shore power; (f) Class 4
    through 7 local freight trucks; (g) airport ground
    support   equipment;   (h) forklifts and   port  cargo
    handling equipment; and (i) light duty zero emission
    vehicle supply equipment (electric or hydrogen vehicle
    charging stations).[4]
    During the 2017–19 biennium, Wisconsin used the settlement funds
    "for replacing eligible state vehicles and for awarding grants
    to transit systems to replace eligible public transit vehicles."5
    4 Executive Session Record for Paper #505 from the Record of
    Committee Proceedings on 2019 Assembly Bill 56 (Paper #505) at 3
    (June                          6,                          2019),
    https://docs.legis.wisconsin.gov/misc/lfb/budget/2019_21_biennia
    l_budget/102_budget_papers/505_volkswagen_settlement_volkswagen_
    settlement.pdf.
    5 Joint Committee on Finance Motion #129 (Motion #129) (June
    6,                                                         2019),
    https://docs.legis.wisconsin.gov/misc/lfb/jfcmotions/2019/2019_0
    6_06/008_volkswagen_settlement/002_motion_129_volkswagen_settlem
    ent.pdf.
    3
    No.   2019AP1376-OA.pdr
    ¶14    For   2019–21,   Governor    Evers     proposed   a    budget   that
    would have expanded uses of the settlement funds to include "the
    installation of charging stations for vehicles with an electric
    motor."6    The Legislature's Joint Committee on Finance rejected
    Governor Evers' proposal, instead opting to create a school bus
    modernization fund to aid school boards in purchasing "energy
    efficient" school buses.7
    ¶15    Governor   Evers   utilized      his   partial    veto    power     to
    attempt    to   accomplish   his   initial    proposal.       To     do   so,   he
    partially vetoed § 55c and vetoed the entirety of § 9101(2i).
    ¶16    The markup of § 55c reads:
    16.047(4s) of the statutes is created to read:
    16.047 (4s) SCHOOL BUS REPLACEMENT GRANTS.      (a) In
    this subsection: 1.    "School board" has the meaning
    given in s. 115.001(7).2.       "School bus" has the
    meaning given in s. 121.51(4).(b)   The department [of
    administration] shall establish a program to award
    grants of settlement funds from the appropriation
    under s. 20.855(4)(h) to school boards for the
    replacement of school buses owned and operated by the
    school boards with school buses that are energy
    efficient, including school buses that use alternative
    fuels.   Any school board may apply for a grant under
    the program. (c) As a condition of receiving a grant
    under this subsection, the school board shall provide
    matching funds equal to the amount of the grant award.
    (d) A school board may use settlement funds awarded
    under this subsection only for the payment of costs
    6 2019 Assembly Bill 56, §§ 52, 53 & 54; see also Paper
    #505, at 2 (explaining the governor wanted to "[e]xpand DOA's
    authority to use settlement monies to award grants for the
    replacement of public transit vehicles to also include awarding
    grants for the installation of charging stations for electric
    vehicles").
    7 Joint Stipulation of Facts and Joint Statement that There
    Are No Material Disputed Facts (Joint Statement), ¶¶21–22.
    4
    No.    2019AP1376-OA.pdr
    incurred by the school board to replace school buses
    in accordance with the settlement guidelines.
    As partially vetoed, the section states: "The department shall
    establish a program to award grants of settlement funds from the
    appropriation under s. 20.855(4)(h) for alternative fuels."
    ¶17    Governor Evers vetoed the entirety of § 9101(2i):
    (2i) VOLKSWAGEN SETTLEMENT FUNDS. Of the settlement
    funds in s. 20.855(4)(h), during the 2019–21 fiscal
    biennium, the department of administration shall
    allocate $3,000,000 for grants under s. 16.047 (4s)
    for the replacement of school buses.
    B.    The Local Road Improvement Fund
    ¶18    The second series of vetoes removed conditions from a
    local road improvement fund, effectively changing it into a fund
    for "local grants" or "local supplements," which did not require
    expenditures for local roads.            For context: "[the Department of
    Transportation]      DOT    administers        the    Local     Roads    Improvement
    Program    (LRIP)    to    assist   political        subdivisions       in   improving
    seriously    deteriorating       local   roads       by   reimbursing        political
    subdivisions    for       certain     improvements.           LRIP      includes   an
    entitlement component and a discretionary component."8
    ¶19    Governor Evers partially vetoed §§ 126 and 184s and
    vetoed the entirety of § 1095m.               Section 126, schedule item 
    Wis. Stat. § 20.395
    (2)(fc),       of     the     enrolled      bill      appropriated
    $90,000,000    for       local   road    improvement       as    a    discretionary
    supplement.9     The markup reads:            "(fc) Local roads improvement
    8 Legislative Reference Bureau Analysis of 2019 Assembly
    Bill    56   (Analysis   of   Bill    56),   at    90,   https://
    docs.legis.wisconsin.gov/2019/related/proposals/ab56.pdf.
    9   Joint Statement, ¶24.
    5
    No.   2019AP1376-OA.pdr
    discretionary     supplement . . . 90,000,000      [and       Governor     Evers
    wrote in 75,000,000]."        As partially vetoed, the scheduled item
    states:    "Local supplement . . . 75,000,000."
    ¶20    Governor       Evers     also    partially     vetoed      § 184s:
    "20.395(2)(fc) of the statutes is created to read:                  20.395(2)
    (fc) Local roads improvement discretionary supplement.               From the
    general fund, as a continuing appropriation, the amounts in the
    schedule    for     the     local   roads    improvement       discretionary
    supplemental grant program under s. 86.31 (3s)."                As partially
    vetoed,    the    section   states:    "Local   supplement.         From     the
    general fund, as a continuing appropriation, the amounts in the
    schedule for local grant [sic]."
    ¶21    Governor Evers vetoed the entirety of § 1095m:
    86.31 (3s) of the statutes is created to read:
    86.31    (3s)   DISCRETIONARY     SUPPLEMENTAL    GRANTS.
    (a) Funds provided under s. 20.395 (2) (fc) shall be
    distributed under this subsection as discretionary
    grants    to  reimburse    political   subdivisions   for
    improvements.     The department shall solicit and
    provide discretionary grants under this subsection
    until all funds appropriated under s. 20.395 (2) (fc)
    have been expended.     (b) 1.    From the appropriation
    under s. 20.395 (2) (fc), the department shall
    allocate $32,003,200 in fiscal year 2019–20, to fund
    county trunk highway improvements.        2.    From the
    appropriation under s. 20.395(2) (fc), the department
    shall allocate $35,149,400 in fiscal year 2019–20, to
    fund    town   road   improvements.        3. From    the
    appropriation under s. 20.395 (2) (fc), the department
    shall allocate $22,847,000 in fiscal year 2019–20, to
    fund municipal street improvement projects.           (c)
    Notwithstanding sub. (4), a political subdivision may
    apply to the department under this subsection for
    reimbursement of not more than 90 percent of eligible
    costs of an improvement.
    6
    No.   2019AP1376-OA.pdr
    C.    The Vehicle Fee Schedule
    ¶22    A    third   series      of    vetoes     altered   the    amount      truck
    owners must pay to register their vehicles.                        Registration fees
    had    varied      depending      on    the    weight     class    of    the   vehicle.
    Section 1988b of the enrolled bill would have made the fee for
    four    weight     classes      the    same.       In   so   doing,     it   would    have
    increased the fee for two weight classes and decreased the fee
    for two others.         Governor Evers used his partial veto powers to
    retain the legislature's proposed fee increases and void its
    proposed decreases.             In the marked-up language, italicized words
    represent         deletions      by    the     legislature,       underlined         words
    represent insertions by the legislature and crossed-out words
    represent partial vetoes by Governor Evers:
    341.25(2)(a) to (cm) of the statutes are amended to
    read: 341.25 (2)(a) Not more than 4,500 $ 75.00 100.00
    (b) Not   more   than   6,000 . . . . . . . . . . 84.00
    100.00 (c) Not more than 8,000 . . . . . . . . . .
    106.00 100.00 (cm) Not more than 10,000 . . . . . . .
    . . . 155.00 100.00
    ¶23    The parties stipulated to a table that summarizes the
    changes:10
    10   Id., ¶33.
    7
    No.   2019AP1376-OA.pdr
    Maximum              Pre-Act 9     Annual Fee            Annual Fee
    Gross                Annual        Approved by           Chosen by
    Weight in            Fee           the                   Governor
    Pounds                             Legislature           Evers
    Not more             $75.00        $100.00               $100.00
    than 4,500
    Not more             $84.00        $100.00               $100.00
    than 6,000
    Not more             $106.00       $100.00               $106.00
    than 8,000
    Not more             $155.00       $100.00               $155.00
    than 10,000
    D.    The Vapor Products Tax
    ¶24    The      last    challenged          veto        altered     a     section    that
    imposed a tax on "vapor products" by expanding the definition of
    vapor product to include liquid heated by a vaping device.                                For
    context, sometimes vaping fluid is sold separately from vaping
    devices.     An analogy is pipe tobacco, which is sold separately
    from pipes.      Section 1754 of the enrolled bill defined vapor
    products to include the hardware that produces vapor from the
    application    of    a     heating     element         to    liquid.          However,    the
    definition    did     not       encompass       the     liquid.             Governor    Evers
    partially vetoed a clause in the definition, which expanded it
    to include the liquid:
    139.75 (14) of the statutes is created to read: 139.75
    (14) "Vapor product" means a noncombustible product
    that produces vapor or aerosol for inhalation from the
    application of a heating element to a liquid or other
    substance that is depleted as the product is used,
    regardless of whether the liquid or other substance
    contains nicotine.
    8
    No.   2019AP1376-OA.pdr
    II.    DISCUSSION
    A.   Standard of Review
    ¶25     As this is an original action, we have no lower court
    opinion to review.11                We are required to interpret Article V,
    Section       10(1)(b)        to     decide      the   pending       controversy,      which
    presents a question of law.                    Koschkee v. Taylor, 
    2019 WI 76
    , ¶9,
    
    387 Wis. 2d 552
    , 
    929 N.W.2d 600
    .
    ¶26     Taxpayers ask us to overturn our precedent in part.
    They bear the burden of persuading us to do so.                                     State v.
    Breitzman, 
    2017 WI 100
    , ¶5 n.4, 
    378 Wis. 2d 431
    , 
    904 N.W.2d 93
    .
    B.    Overview of the Partial Veto Power
    ¶27     Taxpayers       argue          that   some    of     our    decisions    have
    deviated       from     the        original      meaning     of     Article   V,     Section
    10(1)(b) and that we should return to                             the original meaning.
    They     assert,      "[a]s         originally       enacted,      Article    V,     Section
    10(1)(b) of the Wisconsin Constitution authorized the governor
    to     approve     or    disapprove            legislative        proposals   capable     of
    separate       enactment       but      appearing      in   a     single    bill,    nothing
    more."        Therefore, I thoroughly analyze the constitutional text
    and     our    precedent.               In    addition,     I     consider    failed     and
    successful amendments to the governor's partial veto power to
    demonstrate that the people of Wisconsin have actively responded
    to our decisions when they have deemed it proper to do so.
    1.    Amendment of Article V, Section 10
    Original jurisdiction is proper under Wis. Const. art.
    11
    VII, § 3(2).    We have invoked our original jurisdiction to
    interpret the scope of the governor's partial veto powers on
    eight prior occasions, which are discussed below.
    9
    No.    2019AP1376-OA.pdr
    ¶28    The Wisconsin Constitution, as originally adopted in
    1848, did not allow the governor to veto less than an entire
    bill.     At that time, no state constitution authorized the veto
    of less than an entire bill.                 Such authority first appeared in
    the   constitution      of     the    Confederate       States       in     1861    and   was
    limited to appropriations bills.                Henry Campbell Black, Relation
    of the Executive Power to Legislation 103 (1919).                                   By 1919,
    thirty-seven states allowed their governor to veto less than an
    entire      appropriations       bill.         Id.          Notably,        these     states
    generally     adopted       "item"    vetoes.        For    example,         the    Illinois
    Constitution authorized the governor to disapprove "any one or
    more items or sections" of an appropriations bill.                                  State ex
    rel. Wis. Tel. Co. v. Henry, 
    218 Wis. 302
    , 311, 
    260 N.W. 486
    (1935)      (quoting    Ill.    Const.       art.      V,    §     16     (1935)).        One
    contemporary source defined an "item" as "any part of a bill
    [making appropriations] which is sufficiently distinct that it
    may be separated without serious damage to the essential force
    of the residue."         John Mabry Mathews, American State Government
    223 (1926).
    ¶29    In     1911,    Wisconsinites          began        debating     whether     to
    authorize         the   governor        to    veto      less        than      an      entire
    appropriations bill because the legislature started "packaging
    multiple     appropriation       measures       into    larger,           omnibus    bills."
    Richard      A.    Champagne,        Staci   Duros      &    Madeline        Kasper,      The
    Wisconsin     Governor's       Partial       Veto,   Reading        the     Constitution,
    June 2019, at 1, 3-4.          This became known as "logrolling":
    [T]he practice of jumbling together in one act
    inconsistent subjects in order to force a passage by
    10
    No.   2019AP1376-OA.pdr
    uniting minorities with different interests when the
    particular provisions could not pass on their separate
    merits, with riders of objectionable legislation
    attached to general appropriation bills in order to
    force the governor to veto the entire bill and thus
    stop the wheels of government or approve the obnoxious
    act.
    State ex rel. Martin v. Zimmerman, 
    233 Wis. 442
    , 447-48, 
    289 N.W. 662
     (1940).
    ¶30    Before 1911, the legislative practice was to pass on
    each appropriation in a separate bill.                   Champagne et al., The
    Wisconsin Governor's Partial Veto, at 3.                      By 1913, Governor
    Francis     E.    McGovern    began    to    publically     complain      about   the
    changes to the appropriations process.               Id.; State ex rel. Wis.
    Senate v. Thompson, 
    144 Wis. 2d 429
    , 438, 
    424 N.W.2d 385
     (1988).
    He argued that the legislature was passing "omnibus bills" with
    "fifty to one hundred items."                Champagne et al., The Wisconsin
    Governor's        Partial    Veto,     at    3   (quoting     Associated     Press,
    McGovern Criticizes State Legislature, Janesville Daily Gazette,
    Sept. 18, 1913, at 1).            Furthermore, the legislature would wait
    until the current budget was close to expiring.                       Champagne et
    al., The Wisconsin Governor's Partial Veto, at 3.                     He said this
    practice "tied the hands of the executive, and he practically
    had no alternative except to approve the appropriations as a
    whole."     
    Id.
     (quoting McGovern Criticizes State Legislature, at
    1).       Ultimately,       Governor    McGovern     lost     his    campaign     for
    increased        veto   powers.        Champagne    et      al.,    The   Wisconsin
    Governor's Partial Veto, at 4.
    ¶31    The next substantial push for increased gubernatorial
    power came in 1925.           That year, two proposals were considered.
    11
    No.    2019AP1376-OA.pdr
    The first never made it out of committee.                  
    Id.
     at 5 & n.32.          The
    second proposal failed by a vote of 14 to 9 in the Senate.                           Id.
    at 6.      It read, in part:         "The governor may disapprove or reduce
    items or parts of items in any bill appropriating money.                              So
    much of such bill as he approves shall upon his signing become
    law."      1925 Senate Joint Resolution 23.
    ¶32    In 1927, Senator William Titus introduced a similar
    resolution:     "Appropriation bills may be approved in whole or in
    part by the governor, and the part approved shall become law,
    and the part objected to shall be returned in the same manner as
    provided for other bills."                 Champagne et al.,           The Wisconsin
    Governor's      Partial     Veto,     at     6    (quoting     1927    Senate      Joint
    Resolution      35;    1927    Enrolled          Joint   Resolution         37).     The
    resolution passed both houses.                   One newspaper explained, "This
    would allow that executive to return unfavored appropriations to
    the legislators, at the same time passing others in the same
    bill thus speeding the legislative work."                    Champagne et al., The
    Wisconsin Governor's Partial Veto, at 7 n.38 (quoting Beats Plan
    for Repeal of Car Tax, Capital Times, March 15, 1927).                               The
    resolution again passed both houses in 1929, and it was ratified
    by   the    people     in   November       1930.12       Champagne     et     al.,   The
    Wisconsin Governor's Partial Veto, at 7.
    ¶33    Both     the   failed    1925       resolution   and     the    successful
    1930 amendment are believed to have been drafted by Edwin Witte,
    the Chief of the Legislative Reference Library (the predecessor
    The Wisconsin Constitution provides that a proposed
    12
    amendment must be approved by two consecutive legislatures and
    then ratified by the people. Wis. Const. art. XII, § 1.
    12
    No.       2019AP1376-OA.pdr
    to     the    Legislative       Reference        Bureau),        and     drafting         files
    describe an item veto.              See Frederick B. Wade, The Origin &
    Evolution of the Partial Veto Power, Wis. Lawyer, Mar. 2008, at
    12, 14; Mary E. Burke, Comment,                       The Wisconsin Partial Veto:
    Past, Present and Future, 
    1989 Wis. L. Rev. 1395
    , 1402 n.44.
    The drafting file for the 1927 resolution indicates that Senator
    Titus requested the Legislative Reference Library to draft a
    resolution "to allow the Governor to veto items in appropriation
    bills."       A cover sheet in the drafting file reads, "res. to
    permit Gov. to veto items in app. bills."                        The library wrote to
    Senator Titus, "Enclosed herewith is a revised draft of the
    Joint Resolution you asked us to prepare, to allow the Governor
    to veto items in appropriation bills."                       See John S. Weitzer,
    Comment, The Wisconsin Partial Veto:                     Where Are We and How Did
    We     Get    Here?    The     Definition        of     "Part"    and        the       Test   of
    Severability, 
    76 Marq. L. Rev. 625
    , 631 n.35 (1993) (summarizing
    the    drafting       file).     The   1929      drafting        file    has       a    similar
    reference to "allow[ing] the governor to veto items."                              Wade, The
    Origin & Evolution of Partial Veto Power, at 14.
    ¶34    The drafting files do not indicate why, if the drafter
    intended an item veto, he used the word, part.                                 Champagne et
    al.,    The    Wisconsin       Governor's     Partial      Veto,        at    6.        Notably
    though, some contemporary sources used the term "partial" veto
    to describe an item veto.               Black, Relation of the Executive
    Power to Legislation, at 101 (chapter titled "The Selective or
    Partial Veto" describing an item veto used in many states).
    13
    No.   2019AP1376-OA.pdr
    ¶35    The campaign for ratification of the 1930 amendment
    also described an item veto.               Champagne et al., The Wisconsin
    Governor's Partial Veto, at 5.              For example, Witte——the believed
    drafter——wrote a brief supporting its ratification.                                 Edwin E.
    Witte,   Brief     in    Support     of    the       Proposed         Amendment       to       the
    Constitution       to    Allow      the     Governor             to       Veto     Items        in
    Appropriation      Bills    (1930).        Its       first       sentence        reads:    "The
    governor's veto of items in appropriation bills is an essential
    part of an executive budget system."                         
    Id.
              As one article,
    published    in    the   Wisconsin     Lawyer         in    2008,      summarizes:         "The
    brief    uses      the     words    item        and        items      a     total     of        19
    times. . . .       Under these circumstances, it appears that Witte
    viewed the terms part and item as interchangeable synonyms for
    expressing the item veto concept."                   Wade, The Origin & Evolution
    of the Partial Veto Power, at 14.
    ¶36    Several     newspaper        articles          at     the      time     of        the
    constitutional amendment described an item veto.                                 For example,
    The League of Women Voters' "explanation of the proposal" said
    it   would   "enable     the     governor       to    veto       single      items        in   an
    appropriations      bill     without      vetoing          the     entire         bill."         A
    Proposed Amendment, Wausau Daily Record-Herald, Oct. 28, 1930,
    at 8.    A Capital Times article quoted Senator Thomas Duncan, who
    introduced the 1929 resolution, as saying, "[t]he item veto is
    absolutely indispensable."            It would "merely giv[e] back to the
    governor     the   power"      he   had    when       "most        appropriations          were
    divided into separate bills."               Duncan Tells Need for New Vote
    Powers, Capital Times, Oct. 14, 1930, at 7.                                 Similarly, the
    14
    No.   2019AP1376-OA.pdr
    Wisconsin State Journal reported him saying the new veto power
    was "not revolutionary, but on the contrary [was] in successful
    operation in 37 states."                     Veto Rule Better Law Step, Claim,
    Wis. St. J., Oct. 13, 1930, at 7.
    ¶37    Following the amendment's ratification, sources also
    described        it    as    an     item     veto.           For    example,         the    1931–32
    Wisconsin        Blue       Book     explained         the    amendment         permitted          the
    governor "to veto single items in appropriation bills."                                            The
    Wisconsin Blue Book 583 n.1 (1931).
    2.    Our Precedent
    ¶38    We first interpreted the governor's partial veto power
    in   Henry.13          Since       then,   we    have    interpreted            the   governor's
    partial     veto        powers       seven      more    times.            As    our        decisions
    demonstrate, governors have become more creative and aggressive
    with their partial vetoes.                   Yet, our decisions explain only two
    relevant limits:14             (1) the part approved must be a complete,
    entire     and    workable          law;   and    (2) the          part   approved          must    be
    germane to the topic or subject matter of the enrolled bill
    before      the        veto.          Constitutional               amendments         also      have
    added:     "In        approving       an     appropriation           bill       in    part,        the
    At passage, the provisions now in Article V, Section
    13
    10(1)(b) were not in a subsection, but were italicized and
    inserted into Section 10.     The wording was a bit different:
    "Appropriation bills may be approved in whole or in part by the
    governor, and the part approved shall become law, and the part
    objected to shall be returned in the same manner as provided for
    other bills." Wis. Const. art. V, § 10 (1930).
    For write-in vetoes, where a governor crosses out a
    14
    number and writes in a lesser number, we have articulated
    additional restrictions, which are described below.
    15
    No.   2019AP1376-OA.pdr
    governor     may   not   create       a   new    word    by   rejecting      individual
    letters in the words of the enrolled bill, and may not create a
    new sentence by combining parts of 2 or more sentences of the
    enrolled bill."      Wis. Const. art. V, § 10(1)(c).
    a.       Early Cases
    ¶39   In the midst of the Great Depression, Wisconsinites
    were   suffering.        The   legislature         passed     an    emergency      relief
    package.     Henry, 218 Wis. at 307–08.                 As one comment summarizes,
    "To raise revenue for the relief efforts, the nine-section bill
    included six sections providing authority to impose emergency
    income taxes.       Another section of the bill appropriated funds
    for    relief   efforts    and    specified        how     the     funds    were   to   be
    distributed.       Two    other   sections        stated      legislative       intent."
    Burke, The Wisconsin Partial Veto, at 1401.                        The governor "when
    presented with the bill, vetoed the legislative intent sections
    and the distribution subsections of the appropriation section."
    Id.    The assembly did not override his vetoes.
    ¶40   The Wisconsin Telephone Co., a taxpayer, commenced an
    original action, arguing:
    [T]he governor's disapproval of parts of the bill, as
    originally passed, by the legislature, and his
    approval   of   the  remaining   parts   thereof,  was
    unauthorized under [Wis. Const. art. V, § 10] because
    the constitutional grant of power to the governor by
    that section to approve parts of an appropriation bill
    and to disapprove parts thereof does not grant power
    to him to approve the appropriation, and disapprove a
    proviso or condition inseparably connected to the
    appropriation,   nor   to  disapprove   parts   of  an
    appropriation bill that are not an appropriation.
    Henry, 218 Wis. at 309.
    16
    No.    2019AP1376-OA.pdr
    ¶41    We did not decide whether the governor had the power
    to reject provisos or conditions that are inseparably connected.
    Id.       Instead,        we      concluded        that     "the      parts        which      were
    disapproved        by    the     governor    were     not      provisos      or     conditions
    which were inseparably connected to the appropriation."                                         Id.
    But we acknowledged that there was a plausible argument that the
    governor could not veto inseparable provisos or conditions.                                     Id.
    at 309–10 (citing State ex rel. Teachers & Officers v. Holder,
    
    23 So. 643
     (Miss. 1898)).
    ¶42    We    also       concluded      that       the    governor       could          "pass
    independently        on    every      separable      piece      of    legislation          in   an
    appropriation           bill."        Henry,       218     Wis.      at     315.         In     our
    interpretation of the term, "part," which was employed in the
    amendment of Article V, Section 10, we reasoned that the partial
    veto power must be broader than an item veto.                               Id. at 310–14.
    We    also   concluded         that    "part"      should      be    given     its       "usual,
    customary,         and    accepted        meaning         [as] . . . '[o]ne              of     the
    portions, equal or unequal, into which anything is divided, or
    regarded as divided; something less than a whole.'"                                 Id. at 313
    (quoting Part, Webster's New Int'l Dictionary 1781 (2d ed.)).
    We    observed     that     the    part     approved       constituted        a     "complete,
    entire,      and    workable       law,     for    the     appropriation           for     relief
    purposes, of the money to be raised, as tax revenues thereunder,
    and for the allotment and use of that appropriation."                                      Henry,
    218 Wis. at 314.          The vetoes were upheld.
    ¶43    In State ex rel. Finnegan v. Dammann, 
    220 Wis. 143
    ,
    
    264 N.W. 622
     (1936), we were asked to decide whether a bill on
    17
    No.    2019AP1376-OA.pdr
    which the governor asserted a partial veto was an appropriation
    bill.         We     concluded         that    the    enrolled           bill     was    not       an
    appropriations bill.               
    Id. at 148-49
    .           Therefore, the governor's
    attempted veto was "ineffective because the subject matter of
    the    bill        did   not    fall     within      the    constitutional              provision
    authorizing         a    partial      veto."        
    Id. at 149
    .          "Finnegan      added
    nothing to Henry's analysis of the definition of 'part' and the
    test of severability."                 Weitzer, The Wisconsin Partial Veto, at
    637.
    ¶44     Four years after Finnegan, we decided Martin.                                 As one
    comment summarizes, "the legislature enacted a bill changing the
    amount       of     state      funds     appropriated           as     aid      for     dependent
    children."          Burke, The Wisconsin Partial Veto, at 1405.                          As with
    Henry,       the    governor       vetoed      sections         and    subsections           of   the
    enrolled bill.             The Secretary of State refused to publish the
    act     on    procedural        grounds,       which       are        unimportant       for       our
    purposes, as well as substantive grounds.                                 As a substantive
    matter,       he    argued     that     the    partial          vetoes    "so     changed         the
    legislative          program     or     policy . . . as           to     render        the    parts
    approved . . . invalid."                Martin, 233 Wis. at 450.
    ¶45     We began by construing Article V, Section 10.                                       We
    concluded that the partial veto amendment was not ambiguous, and
    as "amended in 1930 it must be construed as a whole."                                        Id. at
    447.     We explained that the amendment's "purpose was to prevent,
    if     possible,         the   adoption        of    omnibus          appropriation          bills,
    logrolling,          the    practice      of     jumbling         together        in    one       act
    inconsistent subjects in order to force a passage by uniting
    18
    No.    2019AP1376-OA.pdr
    minorities      with       different          interests      when         the     particular
    provisions could not pass on their separate merits."                                  Id. at
    447-48.      We then rejected the Secretary of State's argument,
    relying on Henry:
    It must be conceded that the governor's partial
    disapproval did effectuate a change in policy; so did
    the partial veto of the bill involved in the case of
    [Henry], supra, which this court held to be valid.
    The question here is whether the approved parts, taken
    as a whole, provide a complete workable law. We have
    concluded that they do, and we must give them effect
    as such.
    Id. at 450.
    ¶46    For    the     next    four      decades,      "the    partial        veto   was
    rarely    used."          Champagne      et    al.,   The       Wisconsin         Governor's
    Partial Veto, at 1.                "Aside from the 1931 and 1933 biennial
    budget bills, in which there were 12 partial vetoes, subsequent
    governors     either       did     not   partially       veto      any     provisions      or
    partially vetoed only one or two provisions in budget bills
    until the 1969 legislative session."                  Id.
    b.   Later Cases
    ¶47    We next addressed the partial veto in State ex rel.
    Sundby v. Adamany, 
    71 Wis. 2d 118
    , 
    237 N.W.2d 910
     (1976).                                  As
    one   comment      explains,       "[i]n   Sundby,       Governor         [Patrick]    Lucey
    vetoed      clauses    of     sentences.           Previously,            partial     vetoes
    involved only sections and subsections of appropriation bills."
    Weitzer, The Wisconsin Partial Veto, at 639 n.89.                               "The subject
    matter of the portion of the appropriations bill to which these
    partial vetoes appl[ied] involved tax levy limits imposed on
    19
    No.   2019AP1376-OA.pdr
    towns, villages, cities and counties."               Sundby, 
    71 Wis. 2d at 121
    .     The markup read:
    If the [governing body of the political subdivision]
    desires to increase its tax levy above the limitations
    specified in this section, it shall publish such
    intent in a class I notice under ch. 985 in the
    official town newspaper.   The notice shall include a
    statement of the purpose and the amount of the
    proposed levy and the amount by which it wishes to
    exceed the limits imposed by this section. If, within
    20 days after publication of the notice, a petition is
    filed with the town clerk signed by a number of
    electors equal to, or in excess of, 5% of the number
    of electors casting ballots in the town in the last
    gubernatorial election, the question of the proposed
    amount of increase in levy above the limitations
    specified in this section shall be submitted to a
    referendum at a spring election, general election or
    special election.
    
    Id.
        at    122–23.       "In   substance,   the    governor's   veto    made
    mandatory the local referendums which the bill, as passed by the
    legislature, made optional."           
    Id. at 124
    .
    ¶48    We explained that the constitutions of other states
    focused on item vetoes and limited partial vetoes to "item or
    items."      
    Id. at 128
    .    However:
    The Wisconsin Constitution, by way of contrast,
    confers upon its chief executive the power to object
    to "part" of the bill and, in construing this power,
    this court has indicated that the chief executive has
    a   greater  range   of  options  pursuant   to  such
    terminology as to the manner in which he may exercise
    the partial veto than he might have if the power were
    limited to "items."
    
    Id.
    ¶49    We addressed two arguments not thoroughly analyzed in
    our prior decisions for curtailing the governor's partial veto
    power.       First, we considered how separation of powers analysis
    20
    No.   2019AP1376-OA.pdr
    should impact our understanding of the amendment of Article V,
    Section      10.          In   particular,       we      considered     that    "[t]he
    legislative power is vested by the Wisconsin Constitution in the
    senate and the assembly."                 
    Id. at 131
    .       But we then explained
    that the governor plays a role in the legislative process.                           
    Id.
    at 131–34.         Second, we addressed whether the partial veto power
    could "bring about an affirmative change in the result intended
    by the legislature" or merely "negative what the legislature has
    done."       
    Id. at 134
    .      We    rejected      the   distinction     between
    affirmative and negative changes.                
    Id.
         We stated:
    Every veto has both a negative and affirmative ring
    about it.     There is always a change of policy
    involved.   We think the constitutional requisites of
    [Wis. Const.] art. V, [§] 10, fully anticipate that
    the governor's action may alter the policy as written
    in the bill sent to the governor by the legislature.
    Id.    We upheld the vetoes, noting as we had in Henry that the
    provisions were "separable."               Id. at 135.
    ¶50   Two     years     after      Sundby,   we    decided     State    ex   rel.
    Kleczka v. Conta, 
    82 Wis. 2d 679
    , 
    264 N.W.2d 539
     (1978).                             The
    governor's markup read:              "(1) Every individual filing an income
    tax statement may designate that their income tax liability be
    increased by $1 for deposit into the Wisconsin Election Campaign
    Fund for the use of eligible candidates under s. 11.50."                        
    Id. at 685
    .      The consequence of this veto was that taxpayers could
    choose to provide $1 to the campaign fund without increasing
    their tax liability.           
    Id.
    ¶51   The petitioners and the legislature's amicus made two
    arguments.     First, the petitioners argued that the partial veto
    21
    No.   2019AP1376-OA.pdr
    "created an appropriation where none existed before."                                 
    Id. at 704
    .      Second,       the    petitioners          and     the    amicus    argued         that
    "voluntary contributions were a proviso or condition upon which
    the appropriation depended and that such proviso or condition
    were ipso facto inseverable from the appropriation itself."                                 
    Id.
    ¶52   We    rejected       the     first           argument     because       it      was
    "incorrect, under the facts, for the petitioners to assert that
    the bill as altered by the Governor created an appropriation
    where none existed before. . . .                   Rather, it affected the source
    from which the appropriated funds were to be derived."                                
    Id.
     at
    704–05.
    ¶53   Next, we acknowledged that "[s]everability is indeed
    the test of the Governor's constitutional authority to partially
    veto a bill."           
    Id. at 705
    .            We explained that the test for
    severability       is     whether       the        part     approved       constitutes        a
    complete, entire and workable law.                   
    Id.
     at 705–06.          In Henry, we
    had    suggested        that   some      provisos          or     conditions        might    be
    inseparable; in Kleczka, we said that discussion in Henry was
    simply dicta.       Kleczka, 
    82 Wis. 2d at
    712–14.                          Henry did not
    need    to   speculate         about     the       constitutionality           of     vetoing
    provisos     or    conditions          because       the        relevant    sections        and
    subsections were not provisos or conditions.                           In Kleczka, we,
    therefore, upheld the veto.
    ¶54   Justice Hansen authored the first separate writing in
    this line of cases, concurring in part and dissenting in part.
    He noted that "[i]n recent years, partial vetoes have not only
    increased greatly in number; they have been applied to ever
    22
    No.   2019AP1376-OA.pdr
    smaller portions of bills."               
    Id. at 719
     (Hansen, J., concurring
    in part and dissenting in part).                     This concerned him, and he
    stated, "the standard adopted by the court poses no discernible
    obstacle to the use of deletions to produce a complete, entire
    and workable bill concerning a subject utterly unrelated to that
    of the bill as passed by the legislature."                       
    Id. at 723
    .         His
    separation of powers analysis came to the opposite conclusion of
    the majority:         "At some point this creative negative constitutes
    the enacting of legislation by one person, and at precisely that
    point    the     governor        invades       the    exclusive        power   of    the
    legislature to make laws."              
    Id. at 720
    .
    ¶55     He    stated:        "the     partial       veto     power     should      be
    exercised      only    as   to    the     individual     components,       capable    of
    separate    enactment,        which     have     been   joined    together     by    the
    legislature in an appropriation bill.                        That is, the portions
    stricken must be able to stand as a complete and workable bill."
    
    Id. at 726
    .      Stated otherwise, Justice Hansen would have applied
    the complete, entire and workable law requirement to both the
    part approved and the part rejected.
    ¶56     A few years after Kleczka, in 1983, Governor Anthony
    Earl was the first to veto individual letters, which has since
    become   known    as    the      "Vanna    White"15     or   "pick-a-letter"        veto.
    Benjamin W. Proctor, Comment, Wisconsin's Chief Legislator:                           The
    Governor's Partial Veto Authority and the New Tipping Point, 90
    15 Vanna White is a television personality on Wheel of
    Fortune.   James K. Conant, Wisconsin Politics and Government:
    America's Laboratory of Democracy 46 (2006).
    23
    No.   2019AP1376-OA.pdr
    Marq. L. Rev. 739, 750 (2007).             In a law review article he later
    authored, he stated:
    In the 1983–85 budget bill, I vetoed letters and
    digits to reduce a paragraph of five sentences into a
    one-sentence paragraph of twenty-two words.        This
    time, the legislature was not interested in the
    political result; it looked only at the philosophical
    question   of  the   balance   of   power  between  the
    legislative and executive branches.       It determined
    decisively that as a representative of the executive
    branch, I had gone too far.     The veto was overridden
    unanimously by the state assembly and with only one
    dissenting vote in the senate.
    Anthony S. Earl, Personal Reflections on the Partial Veto, 
    77 Marq. L. Rev. 437
    , 440 (1994).
    ¶57    Just    a    few       years   later,   Governor     Tommy    Thompson
    utilized the Vanna White veto.                 He struck "phrases, digits,
    letters, and word fragments in an executive budget bill, so as
    to create new words, sentences, and dollar amounts."                    Champagne
    et al., The Wisconsin Governor's Partial Veto, at 12.
    ¶58    Governor Thompson's vetoes were not overridden, and
    the constitutionality of some of them came before us in Wis.
    Senate, 
    144 Wis. 2d 429
    .              In total, thirty-seven vetoes were
    challenged.     To give one example:
    [O]ne section of the budget bill would have created a
    statutory provision allowing courts to detain for "not
    more   than  48   hours"  any   juvenile  violating  a
    delinquency proceeding court order. Governor Thompson
    vetoed the term "48 hours" and creatively substituted
    "ten days" by vetoing individual letters and words
    from another sentence in that section.
    Burke, The Wisconsin Partial Veto, at 1396.                    To give another
    example,   he   reduced       a    $750,000    appropriation    to   $75,000   by
    vetoing a "0."    
    Id.
    24
    No.   2019AP1376-OA.pdr
    ¶59    Reiterating our analysis from Kleczka——that the part
    approved must be a complete, entire and workable law——we upheld
    the partial vetoes.          Wis. Senate, 
    144 Wis. 2d at
    449–50.
    ¶60    We also explained that the consequences of any partial
    veto must be a law that remains consistent with the topic or
    subject      matter     of   the     "whole"        bill.          
    Id. at 437
    .          "This
    limit[ed] the ability of a governor to strike just any word in a
    sentence."      Champagne et al., The Wisconsin Governor's Partial
    Veto, at 19; see also Gordon B. Baldwin, The Partial Veto Power
    Threatens      Democracy:       A    Rebuttal,            5    Graven     Images       267,       268
    (2002).
    ¶61    There have been two cases regarding the partial veto
    power since Wis. Senate: Citizens Utility Bd. v. Klauser, 
    194 Wis. 2d 484
    , 
    534 N.W.2d 608
     (1995) and Risser v. Klauser, 
    207 Wis. 2d 176
    , 
    558 N.W.2d 108
     (1997).                           In Citizens Utility Board,
    we    concluded       that   the     governor        was       permitted       "to     strike      a
    numerical      sum     appropriated           in    the        bill     and     to     insert      a
    different, smaller number as the appropriated sum."                                     Citizens
    Utility Bd., 
    194 Wis. 2d at 488
    .                     In Risser, we concluded that
    the    governor's       "write-in        veto       may       be   exercised         only    on    a
    monetary     figure     which       is   an    appropriation             amount."           Risser,
    Wis. 2d at 181.
    ¶62    Notably, in both of these opinions, we reiterated the
    limitation we had described as a "germaneness" limitation.                                        Id.
    at 183; Citizens Utility Bd., 
    194 Wis. 2d at 506
    .                                    In Citizens
    Utility      Board,    we    explained        the    write-in         veto     "survives          the
    'topicality'      or    'germaneness'           requirement           set      forth    in     Wis.
    25
    No.    2019AP1376-OA.pdr
    Senate.       The new provision approved by the governor——'$250,000'—
    —relates to the same subject matter as the original legislative
    enactment,          viz.,   a   money      appropriation           to     be        utilized       by
    [Citizens Utility Board] as a public interest advocacy entity."
    Citizens Utility Bd., 
    194 Wis. 2d at 505
    .                          In Risser, while we
    mentioned       a     germaneness       limitation,          we    did        not    apply     it.
    However,       we     did   state    that        "a    governor's         power        to    craft
    legislation          necessarily      must        have       constitutional               limits."
    Risser, 
    207 Wis. 2d at 197
    .
    3.    Failed and Successful Amendments
    ¶63     The    executive      and    legislative           branches          are     acutely
    aware of our decisions in this area.                         There have been numerous
    proposals to amend the partial veto power.                              Champagne et al.,
    The    Wisconsin       Governor's        Partial        Veto,     at     Appendix           Tbl.    3
    (listing proposals from 1935 to 2013).                       Indeed, the same year as
    Henry,       "state     legislators        proposed          limiting         the     governor's
    partial veto authority to 'appropriation items.'                                The proposal,
    however, failed to pass either the Assembly or the Senate."
    Burke, This Wisconsin Partial Veto, at 1403.                              A similar series
    of    events    followed        Martin     and    Kleczka.          
    Id.
           at     1406     n.77;
    Proctor, Wisconsin's Chief Legislator, at 763 n.156.
    ¶64     Twice, the partial veto power has been successfully
    curtailed       by     amendment,        once     in     1990     and         once    in     2008.
    Together, these amendments are set out at Article V, Section
    10(1)(c):       "In    approving      an    appropriation           bill       in     part,    the
    governor      may     not   create    a    new        word   by   rejecting           individual
    letters in the words of the enrolled bill, and may not create a
    26
    No.    2019AP1376-OA.pdr
    new sentence by combining parts of 2 or more sentences of the
    enrolled     bill."       The   1990   amendment,        which       prohibited       the
    creation of words by deleting letters, was a response to Wis.
    Senate.      Burke, The Wisconsin Partial Veto, at 1426.                        The 2008
    amendment was a response to partial vetoes made by Governor Jim
    Doyle and prohibited the creation of new sentences by combining
    parts of two or more sentences of the enrolled bill.                            Proctor,
    Wisconsin's Chief Legislator, at 752–54.
    C.   Stare Decisis
    ¶65   In   the    case-at-hand,       Taxpayers    ask        us    to   overturn
    Henry    because    it    adopted,     in    their     view,    an        overly    broad
    definition of "part."           Alternatively, they ask us to overrule
    Kleczka and "hold that the governor may not exercise the partial
    veto in a way that transforms the meaning and purpose of a law
    into something entirely new."               In particular, they ask us to
    reconsider Kleczka's rejection of the suggestion in Henry that
    the    governor    cannot   veto     "provisos    or    conditions          which    were
    inseparably connected to the appropriation."                    The Legislature's
    amicus asks us to adopt the test proposed by Justice Hansen's
    separate writing in Kleczka:            that both the part approved and
    the part rejected must be complete, entire and workable laws.
    ¶66   "We are respectful of the doctrine of stare decisis."
    State v. Roberson, 
    2019 WI 102
    , ¶49, 
    389 Wis. 2d 190
    , 
    935 N.W.2d 813
    .    As we have explained:
    [Adhering to precedent] ensures that existing law will
    not be abandoned lightly.   When existing law is open
    to revision in every case, deciding cases becomes a
    mere exercise of judicial will, with arbitrary and
    unpredictable results.   Consequently, this court has
    27
    No.   2019AP1376-OA.pdr
    held that any departure from the doctrine of stare
    decisis demands special justification.
    Id., ¶49 (quoting Schultz v. Natwick, 
    2002 WI 125
    , ¶37, 
    257 Wis. 2d 19
    , 
    653 N.W.2d 266
    ) (modifications in the original).
    ¶67    Because        Taxpayers'        argument      is    grounded      in
    originalism,   I   note    that    even     prominent   originalists    respect
    stare decisis.       As Justice Scalia once stated:              "You have to
    make stare decisis an exception to any philosophy of judicial
    interpretation."      Law and Justice Scalia, Hoover Institution at
    23:30–38                   (Mar.                    16,                  2009),
    https://www.youtube.com/watch?v=zE9biZT_z1k&t=1435s                        (last
    visited June 27, 2020); see also Antonin Scalia, A Matter of
    Interpretation:      Federal Courts and the Law 140 (1997) ("[S]tare
    decisis is not a part of my originalist philosophy; it is a
    pragmatic exception to it.").             In one of Justice Scalia's best
    known writings, he explained:
    In [originalism's] undiluted form, at least, it is
    medicine that seems too strong to swallow.       Thus,
    almost every originalist would adulterate it with the
    doctrine of stare decisis——so that Marbury v. Madison
    would stand even if [a prominent legal scholar] should
    demonstrate unassailably that it got the meaning of
    the Constitution wrong.
    Antonin Scalia, Originalism: The Lesser Evil, 57 U. Cinn. L.
    Rev. 849, 861 (1989).
    ¶68    When we are asked to overturn precedent, we consider
    whether:
    (1) Changes or developments in the law have undermined
    the rationale behind a decision; (2) there is a need
    to make a decision correspond to newly ascertained
    facts; (3) there is a showing that the precedent has
    become detrimental to coherence and consistency in the
    28
    No.   2019AP1376-OA.pdr
    law; (4) the prior decision is "unsound in principle;"
    or (5) the prior decision is "unworkable in practice."
    Roberson,    
    389 Wis. 2d 190
    ,   ¶50     (citing         Bartholomew    v.    Wis.
    Patients Comp. Fund & Compcare Health Servs. Ins. Corp., 
    2006 WI 91
    , ¶33, 
    293 Wis. 2d 38
    , 
    717 N.W.2d 216
    ).                   "We also may consider
    'whether [our past decision] has produced a settled body of
    law.'"       Roberson,   
    389 Wis. 2d 190
    ,     ¶50    (quoting     Johnson
    Controls, Inc. v. Employers Ins. of Wausau, 
    2003 WI 108
    , ¶99,
    
    264 Wis. 2d 60
    , 
    665 N.W.2d 257
    ) (modifications in original).
    ¶69     To begin with the request to overturn Henry, Taxpayers
    argue:
    [S]tare decisis is "at its weakest when [this Court]
    interpret[s]     the    Constitution       because    [its]
    interpretations[s]     can     be     altered    only    by
    constitutional amendment."         Franchise Tax Bd. of
    California v. Hyatt, 
    139 S. Ct. 1485
    , 1499 (2019)
    (quoting Agostini v. Felton, 
    521 U.S. 203
    , 235
    (1997)).      Appropriately,    then,    this   Court   has
    recognized    it   need    not    "retain    constitutional
    interpretations that were objectively wrong when
    made."   Koschkee, 
    387 Wis. 2d 552
    , ¶8 n.5.         And, as
    already    explained,    Henry's      interpretation    was
    "objectively wrong when made."
    By "objectively wrong," Taxpayers mean that                     Henry   is not in
    accord   with   the   original    meaning      of     the    1930   constitutional
    amendment.      In addition to the history of Article V, Section
    10(1)(b),    Taxpayers    refer     us    to    other        provisions     of    the
    Wisconsin Constitution that they assert support their argument.
    For example, they cite Article VIII, Section 8, which provides:
    On the passage in either house of the legislature of
    any law which . . . makes, continues or renews an
    appropriation of public or trust money . . . three-
    fifths of all the members elected to such house shall
    in all such cases be required to constitute a quorum
    therein.
    29
    No.    2019AP1376-OA.pdr
    Taxpayers further argue that Henry has not created a "reliance
    interest."       They also contend that Henry has proven "unworkable
    in practice" because it has led to, arguably, abusive practices
    by increasingly creative governors.
    ¶70     Governor Evers responds with several points.                        First,
    he argues that Taxpayers' reliance on Koschkee is misplaced.
    Koschkee "overruled a single decision from three years earlier
    that    had    'no    common      legal    rationale'    for    its     mandate."     He
    argues that Henry is different because of "[t]he near century of
    consistent       partial-veto        decisions"       stemming        from   it.      In
    essence, Governor Evers argues that Henry has produced a settled
    body    of     law,    and   he    claims    a     reliance     interest.       Second,
    Governor Evers cites the constitutional amendments in 1990 and
    2008.        They are, according to him, a "part of the corpus of
    settled law that must be uprooted if [Taxpayers] win" because
    "both     amendments         presuppose       that     Article     V,     § 10(1)(b)[]
    empowers the Governor to veto any 'part' of an appropriation
    bill, no matter how small."                Third, Governor Evers contends that
    partial       veto    decisions     have    been     workable    in    practice.      He
    claims we have had "no problem drawing a line between valid and
    invalid vetoes."
    ¶71     I reject Taxpayers' request to overturn Henry.                      First,
    I cannot say that Henry was objectively wrong.                          An objectively
    wrong opinion is not merely an opinion that was "mistaken."                          Cf.
    State v. Friedlander, 
    2019 WI 22
    , ¶18, 
    385 Wis. 2d 633
    , 
    923 N.W.2d 849
     (explaining the difference between an opinion that is
    objectively wrong and an opinion that is mistaken in the context
    30
    No.   2019AP1376-OA.pdr
    of statutory interpretation); State v. Fuerte, 
    2017 WI 104
    , ¶61,
    
    378 Wis. 2d 504
    ,    
    904 N.W.2d 773
          (same)      (Abrahamson,           J.,
    dissenting).             An    objectively           wrong      opinion        is    one    whose
    interpretation of the law is not plausible.                            State v. Lagundoye,
    
    2004 WI 4
    , ¶¶72–75, 
    268 Wis. 2d 77
    , 
    674 N.W.2d 526
     (Abrahamson,
    C.J.,     dissenting).               That      cannot         be     said      about       Henry's
    interpretation of Article V, Section 10(1)(b).
    ¶72    As we explained in Henry, the words "item" and "part"
    are not synonyms.              Henry, 218 Wis. at 310–11.                           Furthermore,
    nearly every state constitution that authorized an "item veto"
    at the time of the 1930 amendment used the word "item."                                    Id. at
    310–12.      Indeed, the failed 1925 resolution, likely drafted by
    the same person that drafted the 1930 amendment, used the word
    "item."      Therefore, if the intent of the 1930 amendment was to
    create an "item veto," it easily could have been done.                                           In
    addition, Henry was decided in 1935——a mere five years after the
    amendment.          Therefore, as the earliest case interpreting the
    amendment,        to    some   extent,         Henry    is      itself      evidence       of    the
    original meaning of the 1930 amendment.
    ¶73    More       fundamentally,              the        successful,          subsequent
    amendments        to    Article      V,   set    out       in   § 10(1)(c),          prohibit      a
    governor from "creat[ing] a new word by rejecting letters in the
    words of the enrolled bill" and from "creat[ing] a new sentence
    by combining parts of 2 or more sentences of the enrolled bill."
    Article      V,    § 10(1)(c).            If    we     were     to    read     § 10(1)(b)         as
    permitting the veto of only an item, then there would have been
    no    need    for      § 10(1)(c),        which       prohibits        the     governor         from
    31
    No.    2019AP1376-OA.pdr
    removing letters to create a new word or creating new sentences
    with    words       from   two   or    more      sentences.        Stated        otherwise,
    § 10(1)(c) would have no effect after an "item" is vetoed, as
    nothing of the "item" would have been left.                     However, § 10(1)(c)
    has effect because by vetoing "part," smaller portions of an
    enrolled bill can be altered, as shown by Wis. Senate, on which
    § 10(1)(c) placed limits.
    ¶74     Taxpayers'        references           to    other         constitutional
    provisions are not persuasive; indeed, the references highlight
    why Taxpayers have not established that Henry is objectively
    wrong.        Taxpayers      ask      us   to    minimize    the     role       Article    V,
    § 10(1)(c) plays in our interpretation, even though it sets out
    successful       amendments,       which        clearly     relate       to     Article     V,
    § 10(1)(b).         Yet, they ask us to consider other provisions that
    are    not    clearly      related.         For      example,   Taxpayers         have    not
    explained how Article VIII, Section 8 supports their argument.
    It provides quorum requirements for votes on fiscal bills.                                What
    that has to do with the partial veto power, which takes place
    after such a vote, is unclear.
    ¶75     Second,     our     decisions,         consisting     of        eight   cases
    dating back eighty-five years, have produced a "settled body of
    law" despite naysayers' attempts to unsettle it.                              Roberson, 
    389 Wis. 2d 190
    ,    ¶50   (quoting        Johnson     Controls,     
    264 Wis. 2d 60
    ,
    ¶99).        Indeed, we have previously rejected a similar argument
    about original meaning.               Wis. Senate, 
    144 Wis. 2d at
    461 n.18.
    We    cannot    rehash     original        meaning——and      its     interaction          with
    stare decisis——every time a partial veto comes before us.                              There
    32
    No.    2019AP1376-OA.pdr
    is good reason that prominent originalists have recognized stare
    decisis as an exception to their judicial philosophy.                    Scalia,
    Originalism:    The    Lesser     Evil,    at    861.         Furthermore,     the
    political branches, as well as the media and legal scholars, are
    aware of our interpretations of Article V, Section 10(1)(b), and
    Wisconsinites   actively    have    debated      the    proper   scope   of    the
    governor's partial veto power.           As already explained, there have
    been numerous attempts to amend the partial veto power, two of
    which were successful.
    ¶76   At bottom, item veto advocates, despite substantial
    effort, have not been able to convince their fellow citizens to
    adopt an item veto.     At this point, as we said in Henry:
    If the legislature and people wish the governor to
    have only the power to veto items in an appropriation
    bill, a constitutional amendment may be desirable. It
    should, however, be understood that this court has no
    power to toy with the constitutional grant of a
    partial veto to the governor and to replace it with a
    veto power that may be more sensible and palatable.
    Any claimed excesses on the part of the governor in
    the exercise of this broad partial veto authority are
    correctable not by this court, but by the people,
    either at the ballot box or by constitutional
    amendment.
    Wis. Senate, 
    144 Wis. 2d at 465
    .
    ¶77   Third,      Taxpayers    are    incorrect      in    suggesting     that
    Henry has proved unworkable in practice because governors have
    exercised creative partial vetoes which we have evaluated.                      An
    opinion   may   be     unworkable    in     practice      when     courts     have
    difficulty applying it.         See State v. Harris, 
    2010 WI 79
    , ¶43,
    
    326 Wis. 2d 685
    , 
    786 N.W.2d 409
    .                We have not had difficulty
    33
    No.   2019AP1376-OA.pdr
    interpreting challenged vetoes in light of our past decisions;
    therefore, Henry has not proved unworkable in practice.
    ¶78     Taxpayers    alternatively       argue   we     should     overturn
    Kleczka because it is "detrimental to coherence and consistency
    in the law."     They assert it is inconsistent with our decisions
    interpreting separation of powers.            They cite League of Women
    Voters of Wis. v. Evers, 
    2019 WI 75
    , 
    387 Wis. 2d 511
    , 929,
    N.W.2d    209,   Tetra   Tech   EC,   Inc.   v.   DOR,    
    2018 WI 75
    ,    
    382 Wis. 2d 496
    , 
    914 N.W.2d 21
     (lead) and Gabler v. Crime Victims
    Rights Bd., 
    2017 WI 67
    , 
    376 Wis. 2d 147
    , 
    897 N.W.2d 384
    .                     They
    also cite Federalist No. 58 as authority for a particular model
    of separation of powers.
    ¶79     The Legislature in its amicus brief adds:
    When precedent does "not even discuss" a critical
    aspect of the relevant text, stare decisis does not
    require the [c]ourt to persist in a prior, deficient
    interpretation.   State v. Denny, 
    2017 WI 17
    , ¶¶67–70,
    
    373 Wis. 2d 390
    , 
    891 N.W.2d 144
    .     In the context of
    Article V, Section 10, this [c]ourt has correctly
    interpreted one portion of the text, reading "part
    approved becomes law" to mean "a complete, entire, and
    workable law."    Wis[.] Senate, 
    144 Wis. 2d at 437
    .
    Yet, this [c]ourt has not given attention to another
    portion of the text, which explains when the "rejected
    part" "become[s] law."     This has created a serious
    separation-of-powers problem, wherein the Governor can
    effectively enact law by vetoing sentence fragments.
    Legislature Amicus Br. at 3.          To explain, the Legislature makes
    a temporal argument about when the part rejected becomes law.
    The part approved becomes law when it is signed by the governor;
    the part rejected does not.           The part rejected is returned to
    the legislature and becomes law if and only if it is "approved
    by two-thirds of the members present."               Wis. Const. art. V,
    34
    No.    2019AP1376-OA.pdr
    § 10(2)(b).          Therefore,         the    part       rejected,       according    to   the
    Legislature's amicus, must be capable of separate enactment at a
    later date, independent of the part approved.
    ¶80    Governor Evers responds that no inconsistency has been
    created.      The cases cited by Taxpayers dealt with issues bearing
    no resemblance to the governor's partial veto power.                                In League
    of Women Voters, we concluded that "[h]ow the Legislature meets,
    when   it     meets,       and    what    descriptive            titles    the     Legislature
    assigns       to    those        meetings       or       their     operating       procedures
    constitute         parts    of    the    legislative         process        with    which   the
    judicial branch 'has no jurisdiction or right' to interfere."
    League       of    Women     Voters,          
    387 Wis. 2d 511
    ,     ¶37     (internal
    quotations removed).             In Tetra Tech, we concluded that we do not
    give     great       weight       deference          to     administrative           agencies'
    conclusions of law.               Tetra Tech, 
    382 Wis. 2d 496
    , ¶108.                         In
    Gabler, we concluded that an executive agency could not, "acting
    pursuant to authority delegated by the legislature, review a
    Wisconsin          court's       exercise           of     discretion,        declare       its
    application of the law to be in error, and then sanction the
    judge for making a decision the agency disfavors[.]"                                   Gabler,
    
    376 Wis. 2d 147
    , ¶36.
    ¶81    Governor Evers responds that the Legislature's amicus
    is inconsistent with historical practice:
    The phrase "shall become law" simply describes
    the transformation that occurs when a bill is
    presented to the Governor for his approval.
    . . . .
    35
    No.   2019AP1376-OA.pdr
    The Legislature would instead read "shall become
    law" as imposing a complete-and-workable-law test
    wherever the phrase appears. But that makes no sense
    applied to the rejected part of an appropriation bill.
    Unlike the part approved——which immediately becomes
    law under Article V, § 10(1)(b)——the rejected part
    never needs to function as a stand-alone law. Either
    it remains rejected and never becomes law, or, upon a
    successful legislative override, it rejoins the part
    approved and "the bill as originally passed by the
    legislature becomes law."     Richard A. Champagne &
    Madeline Kasper, Wis. Legis. Reference Bureau, The
    Veto Override Process in Wisconsin 1 (2019).
    ¶82   I    reject      Taxpayers'    request      to    overturn     Kleczka.
    Their argument presumes that states are obligated to follow a
    particular    model   of   separation        of   powers   that   delegates    the
    "power of the purse" totally to the legislature.
    ¶83   However, our jurisprudence consistently describes the
    governor's role in the budgeting process as "quasi-legislative."
    Wis. Senate, 
    144 Wis. 2d at 454
     (quoting Henry, 218 Wis. 2d at
    314).   In Wis. Senate, we stated:
    [The   1930    amendment]   gave   the    governor   a
    constitutionally recognized role in the legislative
    budgetary function.      The legislature itself has
    recognized the governor's legislative role in the
    budget area by ceding to the governor the initial
    responsibility for preparing the biennial budget
    report and requiring him to submit his executive
    budget bill together with suggestions for the best
    methods for raising the needed revenues.     It was no
    coincidence that the same 1929 legislature which
    passed [ch. 97, Laws of 1929], adopting the executive
    budget system for this state, thereby creating a
    statutory role for the governor in the budgetary
    process, also passed——for the requisite second time——
    the [] joint resolution proposing the constitutional
    amendment to [Wis. Const.] art. V, [§] 10 to provide
    for the governor's partial veto authority. These acts
    were all part of the complete overhaul of the budget
    system in this state that took place at that time.
    The partial veto power the governor may exercise over
    36
    No.    2019AP1376-OA.pdr
    appropriation bills is simply one tool he                   has    for
    controlling his own executive budget bill.
    Wis. Senate, 
    144 Wis. 2d at
    454–55 (internal citations omitted).
    Taxpayers simply ignore these statements because they do not fit
    their understanding of separation of powers.
    ¶84    Furthermore,       our   jurisprudence     is      not   unique    in
    describing a quasi-legislative role for the governor.                   A veto
    power, regardless of its contours, is inherently legislative.
    The United States Supreme Court has said so in a number of
    cases.    For example, it has explained:
    It is said that the approval by the President of a
    bill passed by Congress is not strictly an executive
    function, but is legislative in its nature; in this
    view, it is argued, conclusively shows that his
    approval can legally occur only on a day when both
    Houses are actually sitting in the performance of
    legislative functions. Undoubtedly the President when
    approving bills passed by Congress may be said to
    participate in the enactment of laws which the
    Constitution requires him to execute.       But that
    consideration does not determine the question before
    us.    As the Constitution, while authorizing the
    President to perform certain functions of a limited
    number that are legislative in their general nature,
    does not restrict the exercise of those functions to
    the particular days on which the two Houses of
    Congress are actually sitting in the transaction of
    public business, the court cannot impose such a
    restriction upon the Executive.
    La Abra Silver Min. Co. v. United States, 
    175 U.S. 423
    , 453
    (1899); see also Edwards v. United States, 
    286 U.S. 482
    , 490
    (1932)     ("The     President      acts       legislatively        under    the
    Constitution,      but   he   is    not    a   constituent     part    of    the
    Congress."); cf. Rateree v. Rockett, 
    852 F.2d 946
    , 951 (7th Cir.
    1988) ("[W]hen the Vice President of the United States votes in
    the Senate to break a tie, U.S. Const. art. I § III cl. 4, he
    37
    No.   2019AP1376-OA.pdr
    acts legislatively, not executively.                  Similarly, the President
    acts legislatively when he approves or vetoes bills passed by
    Congress.").
    ¶85   Taxpayers seem to assume that the governor cannot have
    a quasi-legislative role because creating law is a core power of
    the   legislature.          Under   this      theory,     the      power    to     create
    legislation cannot be shared.            At least two problems exist with
    this assumption.
    ¶86   First, as demonstrated by rulemaking, and as we have
    long concluded, the legislature may delegate its power to make
    law to the executive.         Martinez v. DILHR, 
    165 Wis. 2d 687
    , 697,
    
    478 N.W.2d 582
     (1992) (citing Schmidt v. Local Affairs & Dev.
    Dep't, 
    39 Wis. 2d 46
    , 56, 
    158 N.W.2d 306
     (1968)) ("Legislative
    power may be delegated to an administrative agency as long as
    adequate standards for conducting the allocated power are in
    place.").      Such    a     delegation       would     be    impossible         if    the
    executive    were     not    permitted        to   have      at     least    a     quasi-
    legislative role in our constitutional structure.
    ¶87   Second, this theory does not account for the text of
    the Wisconsin Constitution.            As Kelczka said, we must look first
    to the text of the Wisconsin Constitution, not references to
    philosophical works, such as Montesquieu's The Spirit of Law.
    Kleczka, 
    82 Wis. 2d at
    710 n.3 (explaining how Montesquieu and
    the Federalist Papers should impact our understanding of the
    partial veto power).          Taxpayers would have us reverse this by
    first   considering     philosophical         works     and     then     consider      the
    constitutional      text.       Such     an    analysis       would       ignore      that
    38
    No.   2019AP1376-OA.pdr
    Wisconsinites         are   free     to    assign       powers    traditional         to   one
    branch      of     government       to     another       branch       by    constitutional
    amendment.16         I also cannot ignore how the constitutional text
    has been understood for nearly a century.
    ¶88     In     addition,      whether      the     Federalist         Papers    support
    Taxpayers' position is unclear.                  As we explained in Kleczka, the
    Federalist Papers can be read to support an expansive reading of
    the        partial      veto        power.               "The     authors            of    The
    Federalist . . . repeatedly                 alluded        to     the       tendency,       in
    republican forms of government, to the aggrandizement of the
    legislative        branch    at     the     expense      of     the    other     branches."
    
    Id.
     (citing        Federalist       No.    73    (Hamilton);          No.   49   (Madison);
    No. 48      (Madison)).            Indeed,       the    legislature's         practice      of
    logrolling spawned the need for Article V, Section 10(1)(b).
    ¶89     Moreover,        I      cannot           accept     the        position       of
    Legislature's amicus that we should apply the complete, entire
    and   workable       law    test     to    the    part     rejected.           The    textual
    analysis provided by Governor Evers fits historical practice:
    the phrase "shall become law" describes the transformation that
    occurs when proposed legislation takes on legally binding force.
    It does not indicate that the part rejected must be a complete,
    entire and workable law.                  Governors and legislatures have long
    As Judge Posner explained when the partial veto power was
    16
    challenged in federal court: "That it is unusual, even quirky,
    does not make it unconstitutional.      It violates no federal
    constitutional   provision    because   the    [United   States]
    Constitution does not fix the balance of power between branches
    of state government."   Risser v. Thompson, 
    930 F.2d 549
    , 554
    (7th Cir. 1991).
    39
    No.   2019AP1376-OA.pdr
    understood that the part rejected rejoins the part approved if
    the legislature overrides the governor's veto.                          Governor Evers
    cites a document by the Legislative Reference Bureau that says
    as   much.     Champagne       &    Kasper,      The    Veto    Override     Process    in
    Wisconsin.       Also,        our    decisions         are    consistent     with     this
    understanding.      Citizens Utility Bd., 
    194 Wis. 2d at 488
    .                           If
    the governor were to veto "$100,000" and write in "$90,000," all
    would understand that a legislative override of the veto would
    mean   that    $10,000    is        added   to    the        $90,000    to   return    the
    appropriation to its original number.                   See 
    id.
    D.   Application
    ¶90    Having broken no new ground, I employ our decisions
    and continue the constitutional analysis of "part" in the four
    vetoes that were challenged.                Taxpayers do not dispute that the
    "part approved" constitutes a complete, entire and workable law.
    Rather, the dispute before us is whether Governor Evers' partial
    vetoes went too far by altering the topic or subject matter of
    the enrolled bills.            Stated otherwise, we have a dispute over
    whether the parts approved alter the stated legislative idea for
    which the enrolled bill was passed.
    1.    Topic or Subject Matter
    ¶91    The legislature controls whether an idea will result
    in an enrolled bill that will be presented to the governor for
    signature.      A veto that does not alter legislative control of
    the topic or subject matter of enrolled bills has been referred
    to as "germane."          Wis. Senate, 
    144 Wis. 2d at 437
    .                          Stated
    otherwise, such a veto does not alter the stated legislative
    40
    No.    2019AP1376-OA.pdr
    idea that initiated the enrolled bill.                     The text of Article V,
    § 10(1)(b), which employs the term, "part," twice in the same
    sentence and connects "part" to the "whole" bill states:                          "bills
    may be approved in whole or in part by the governor, and the
    part   approved      shall       become    law."      A    plain      reading    of    the
    constitutional text connects the "part" approved by the governor
    to the "whole" bill because it is only a "part" of that "whole"
    bill that is vetoed.               When the part approved by the governor
    does not alter the topic or subject matter of the whole bill
    presented to him for signature, the part approved maintains the
    legislature's choice of topic or subject matter that underlies
    the "whole" bill.           Stated otherwise, when legislative topic or
    subject matter is maintained, the "part" approved and the "part"
    that was not approved remain portions of the same "whole" bill,
    consistent with the constitutional text of § 10(1)(b).                          Clearly,
    the evaluation of "part" and "whole" in § 10(1)(b) depends on
    how broadly the topic or subject matter is defined.
    ¶92   For     example,       we     have     previously        concluded        that
    $250,000 is a "part" of $350,000, and, therefore, the governor
    may veto $350,000 and write in $250,000.                     Citizens Utility Bd.,
    
    194 Wis. 2d at 505-06
    .             We explained that "$250,000 is 'part' of
    $350,000[] because $250,000 is 'something less than' $350,000,
    and    $250,000      goes    'to    make    up,     with     others . . . a       larger
    number,'     i.e.,    $350,000."           
    Id.
        (quoting    Part,     Webster's       New
    Int'l Dictionary 1781 (2d ed.)).
    ¶93   We also rejected an argument that "part" means only
    "physical     part[s]       of   the     bill."     Citizens       Utility      Bd.,    194
    41
    No.    2019AP1376-OA.pdr
    Wis. 2d at 503–04.                 To explain, "[i]f the governor strikes a
    $100 appropriation and writes in $80, the amount the governor
    attempts         to   veto    is    $20.         However,          '$20'     does      not    appear
    anywhere in the bill.               '$20' is not physically part of the bill.
    It    is    part      of    the    bill       only        conceptually."            
    Id. at 503
    .
    Nevertheless,          we     have           permitted       write-in        vetoes         because,
    conceptually, the amount remaining after the veto is a part of
    the bill.         
    Id. at 510
    .            Stated otherwise, the idea contemplated
    by the legislature in funding an identified entity or described
    project remains after the veto.                            If the entity or project is
    funded      to    a   lesser       degree       because       of    a   write-in        veto,       the
    legislative idea that initiated the enrolled bill remains after
    the veto nevertheless.                   Similarly, an enrolled bill's topic or
    subject matter is part of its makeup.
    ¶94       When the topic or subject matter of a bill is altered
    through veto from that of the whole bill that was presented for
    the governor's signature to a topic or subject matter conceived
    by    the       governor,         the    veto        is    outside      of       the   governor's
    constitutional authority.                    When the veto is used in that manner,
    the "part approved" cannot be defined as a "part" of the "whole"
    bill passed by the legislature because it is inconsistent with
    the   constitutional              meaning       of     "part"      in   Article        V,    Section
    10(1)(b).
    ¶95       Secondary sources have discussed the topic and subject
    matter limitation on vetoes.                         I note that their understanding,
    which      is    referred     to        as    germaneness,         is   consistent           with   my
    analysis in this case.                  In particular, Jack Stark in discussing
    42
    No.   2019AP1376-OA.pdr
    vetoes       made   by    Governor    Doyle   that     triggered     the   2008
    constitutional amendment, stated:
    The case law has recently produced a significant
    restriction, holding that the material left after a
    veto must be germane to (have the same subject matter
    as) the material from which it was fashioned. If the
    vetoes of the most recent budget bill that got the
    most attention had been challenged, they would most
    likely have been reviewed in light of that principle.
    With two related vetoes the Governor effected a
    transfer of several hundred million dollars from the
    transportation fund to the general fund.     The money
    transferred would ultimately increase school aid.   In
    both of those vetoes, the germaneness requirement
    appears to have been violated.    Most of the material
    that was vetoed was about particular transportation
    projects, and some of it was about the unfunded
    liability of the state's retirement system.
    Jack Stark, Symposium, Is the Wisconsin Constitution Obsolete? A
    Conference on the Wisconsin Constitution, 
    90 Marq. L. Rev. 411
    ,
    417-18       (2007);     see   also   Champagne   et   al.,    The   Wisconsin
    Governor's Partial Veto, at 18–19.
    2.     Application of Topic or Subject Matter Limitation
    a.   School Bus Modernization and Local Road Improvement Funds
    ¶96    Taxpayers argue:
    Sections 55c and 9101(2i) of Act 9 allocated $3
    million of certain settlement funds for modernizing
    school buses, with specific conditions as to how that
    program should operate.    Governor Evers transformed
    this into an open-ended grant "for alternative fuels"
    with no conditions, and then directed by fiat that the
    agency in charge spend up to $10 million "for electric
    vehicle charging stations."   This is so far removed
    from what the Legislature intended to create that
    there is no question that the portions Evers' vetoed
    were non-severable.
    ¶97    As for the local road improvement fund, they state:
    43
    No.   2019AP1376-OA.pdr
    Sections 126, 184s, and 1095m of Act 9 allocated
    $90 million for the improvement of local roads, along
    with   specific   sub-allocations  for   county  trunk
    highways, town roads, and municipal streets. Governor
    Evers used the partial veto to transform this into a
    $75 million allocation "for local grant [sic]." This
    veto entirely eliminated the core purpose of the award
    (local road improvements), instead creating a generic
    slush fund with no meaningful constraints.
    ¶98   Governor Evers has made no response to these points.
    Quoting from the dissent in Wis. Senate, he seems to acknowledge
    in   a    footnote        of   his    brief    that    "what    remains       [must]   be
    germane."          Wis.    Senate,     
    144 Wis. 2d at 474
        (Bablitch,      J.,
    dissenting).         But he does not explain how what he labels as
    "parts that remain" are in accord with their originating actions
    of the enrolled bill.             Instead, he argues that he can veto "any
    part,     no   matter      how    small"     unless    prohibited      by    Article    V,
    Section 10(1)(c).
    ¶99   I   agree       with   Taxpayers;      these     vetoes      resulted   in
    topics and subject matters that were not found in the enrolled
    bill, i.e., they were not a "part" of the enrolled bill.                          Stated
    otherwise, the enrolled bill says nothing about an "alternative
    fuel fund."         The parts of the enrolled bill that remain after
    this veto have nothing to do with school buses; indeed, the
    remaining part has nothing to do with schools or even education.
    Governor Evers has publically stated he wants to use the fund
    for electric charging stations, a use not contemplated by any
    part of the enrolled bill and one specifically rejected by the
    legislature.
    ¶100 Notably,         Governor       Evers   vetoed     the        entirety   of
    § 9101(2i),        which       "allocate[d]        $3,000,000    for     grants    under
    44
    No.   2019AP1376-OA.pdr
    s. 16.047     (4s)     for   the       payment          of   school       buses."       (Emphasis
    added.)        Section       9101(2i)             further     demonstrates              that   the
    legislative idea of § 16.047(4s) was to replace school buses.
    The   legislative      idea       of    § 16.047(4s)           was    not,        for    example,
    limiting carbon emissions.
    ¶101 Legislative history confirms that the legislative idea
    was to replace school buses.                      Settlement funds in the previous
    biennium     were    used    to    replace             "eligible     state      vehicles"      and
    "public transit vehicles."17                 Governor Evers sought to "[e]xpand
    DOA's authority to use settlement monies to award grants for
    replacement of public transit vehicles to also include grants
    for   the     installation             of     charging        stations          for      electric
    vehicles."18     Governor Evers' proposed expansion was rejected in
    favor of one more analogous to previous uses of the settlement
    funds.
    ¶102 Similarly,         the       partial          vetoes      of     the    local       road
    improvement fund, which created a generic fund, are precisely
    the kinds of vetoes commentators have assumed would violate the
    topic or subject matter limitation.                           Stark, Is the Wisconsin
    Constitution        Obsolete,     at        418    ("With     two     related       vetoes     the
    Governor effected a transfer of several hundred million dollars
    from the transportation fund to the general fund.                                       The money
    transferred would ultimately increase school aid.                                  In both of
    those vetoes, the germaneness requirement appears to have been
    violated.").
    17   Motion #129.
    18   Paper #505.
    45
    No.    2019AP1376-OA.pdr
    ¶103 The       legislative     idea       was    to    fund    an     ongoing     road
    improvement program.19        Section 1095m, vetoed in its entirety by
    Governor Evers, made this clear.                  It allocated specific amounts
    to    fund     "county     truck     highway          improvements,"          "town     road
    improvements" and "municipal street improvements."                                 A general
    undirected fund was not part of a fund created to improve local
    roads      because   a   general   fund     can       be    spent    on    virtually       any
    subject, i.e., topics and subject matters never considered by
    the   legislature.         Indeed,    a     general         fund    could     be    used    to
    accomplish goals explicitly rejected by the legislature during
    its deliberative process.
    ¶104 I      cannot    uphold        these       vetoes.             Accordingly,       I
    partially concur with the per curiam opinion that these vetoes
    have no effect on the provisions in the enrolled bills that the
    legislature enacted.
    b.   Vehicle Fee Schedule and Vapor Products Tax
    ¶105 Taxpayers have not carried their burden with respect
    to    the     remaining     vetoes.          With          regard    to      the     vehicle
    registration fees, Taxpayers argue:
    Governor   Evers   accepted the   increases  and
    rejected the decreases, creating a new fee schedule
    that is neither graduated nor equalized.          The
    question, under traditional severability analysis, is
    whether the Legislature would have intended the fee
    increases on lighter trucks without the corresponding
    decreases for heavier trucks. Given that the obvious
    purpose of the statutory change was to equalize the
    fee schedule, the answer is no.
    19   Analysis of Bill 56, at 90.
    46
    No.       2019AP1376-OA.pdr
    This is an inherently different argument than what Taxpayers
    raised in regard to the school bus modernization fund and the
    local    road   improvement        fund.        The       part    approved        is    clearly
    related    to   the    subject     matter       of    vehicle       registration          fees.
    These vetoes are consistent with those that we approved in Wis.
    Senate and that long have been considered within the governor's
    partial veto power.          Burke, The Wisconsin Partial Veto, at 1396.
    ¶106 A similar analysis applies to the veto that altered
    the     definition     of    vapor    product.              The    veto        expanded     the
    definition of vapor product, thereby expanding what could be
    taxed.     But it did not alter the topic or subject matter of the
    part approved.         Rather, it would seem all products that would
    have been taxed under the enrolled bill will continue to be
    taxed.     Furthermore, the liquid used in vaping devices is within
    the   scope     of    the    phrase    vapor         product       as    used      in    common
    parlance.       Had the legislature left vapor product undefined,
    reasonable people may have assumed it encompassed liquid sold
    separately.
    E.   Remedy
    ¶107 The Legislature's amicus has asked us to consider a
    remedy that is purely prospective.                        As it explains, while our
    decisions       "'[n]ormally'          apply          'retrospectively,'                purely
    prospective     application——which          does       not       apply    a    new     decision
    even to the case at hand——is appropriate where retrospective
    application     of    a     'new   principle         of    law'     would       'unsettl[e]'
    reliance interests."           See State v. Beaver Dam Area Dev. Corp.,
    
    2008 WI 90
    , ¶¶95–101, 
    312 Wis. 2d 84
    , 
    752 N.W.2d 295
    .                                    Had I
    47
    No.   2019AP1376-OA.pdr
    accepted the Legislature's argument and concluded that the part
    rejected by the governor should be a complete, workable law, I
    might view its request differently.                     However, I reject this
    request    because   I     break      no   new    ground    with    this   decision.
    Indeed,    the    topic    and     subject       matter    limitation,     sometimes
    referred to as germaneness, has been discussed in three prior
    cases.    It is not a new principle of law.                  Risser, 
    207 Wis. 2d at 183
    ; Citizens Utility Bd., 
    194 Wis. 2d at 506
    ; Wis. Senate,
    
    144 Wis. 2d at
    451–52.
    III.      CONCLUSION
    ¶108 I conclude that the part approved by the governor,
    i.e., the consequences of the partial veto, must not alter the
    topic or subject matter of the "whole" bill before the veto.
    Stated otherwise, such a veto does not alter the legislative
    idea    that   initiated    the       enrolled     bill.     Therefore,     Governor
    Evers could not use his partial veto power to alter the school
    bus modernization fund into an alternative fuel fund.                      Nor could
    he use it to alter the local road improvement fund into a fund
    for local grants or local supplements, devoid of any requirement
    that it be used for local roads.                 These two series of vetoes are
    invalid and have no effect on those laws.                      However, Governor
    Evers lawfully used his partial veto power to alter the amount
    truck    owners   must    pay    to    register     their   vehicles.       He   also
    lawfully exercised his partial veto power in regard to vaping
    products.      These vetoes stand.
    48
    No.     2019AP1376-OA.awb
    ¶109 ANN        WALSH      BRADLEY,             J.        (concurring        in        part,
    dissenting in part).            In an important case like this, where the
    people   of     Wisconsin      need     clarity,            we    instead     sow    confusion.
    Evidence of the lack of clarity is highlighted by the very fact
    that this case has generated four separate writings with various
    rationales.       And not one of them has garnered a majority vote of
    this    court.      Thus,       we    are    left          with    no    clear      controlling
    rationale or test for the future.
    ¶110 I agree with that part of the per curiam opinion that
    upholds the vehicle fee schedule veto.                             The Governor lawfully
    used his partial veto power when he altered the amount truck
    owners must pay to register their vehicles.
    ¶111 Employing different rationales or tests, the majority
    of justices err, however, by determining that the other three
    vetoes at issue are unconstitutional and must be struck down on
    the basis of arguments neither argued nor briefed by any party.
    In   doing    so,      Chief    Justice          Roggensack's           concurrence/dissent
    seeks to create a subjective test that unnecessarily inserts the
    court    into     policy       disputes      between             the    other     branches      of
    government,      and    is     likely       to       lead    to    more     uncertainty        and
    litigation      over    partial       vetoes          by    future      governors        of   this
    state.
    ¶112 Not only does Chief Justice Roggensack's opinion base
    this decision on a theory that no party has advanced, but it is
    also based on a theory that has never been actually applied.
    The opinion's proffered "topic or subject matter" test morphs
    into an alternative test as the analysis unfolds.                                    That test
    1
    No.    2019AP1376-OA.awb
    eschews    the    "topic     or      subject        matter"     language      and    instead
    focuses on an amorphous concept of what was the "legislative
    idea"    behind   the    bill.         Both        iterations    of    the    test      invite
    manipulation      and    inject        subjectivity           into    what    was    once    a
    clearly    objective       test.         Such        subjective       and     manipulative
    determinations        have      no     place       in    addressing         the   important
    question of the constitutionality of the use of the governor's
    partial veto power.
    ¶113 Justice Kelly's concurrence/dissent suffers from the
    same infirmity as does Chief Justice Roggensack's:                                  it, too,
    embraces a test neither advanced by any party nor ever applied
    in any case.          Advocating for invalidating all four vetoes at
    issue,    Justice     Kelly's         writing       would     overrule       or   modify     a
    multitude of cases, spanning 85 years of precedent, and would
    render two constitutional amendments superfluous.
    ¶114 Likewise, Justice Hagedorn's concurrence relies on a
    theory not argued by the parties.                       The opinion would "revisit"
    and overrule a number of precedential cases.                               It also injects
    subjectivity into the determination of the constitutionality of
    an exercise of the partial veto power, ultimately determining
    that three of the four vetoes are unconstitutional.
    ¶115 Rather than embrace tests neither previously argued
    nor   applied,    I     would     instead      turn      to    and    uphold      our   well-
    established precedent.               It recognizes, time and again, that the
    Wisconsin governor's veto power is incredibly broad.                                Contrary
    to the determinations based on untested theories set forth in
    the various separate writings, I conclude that our precedent
    2
    No.    2019AP1376-OA.awb
    inexorably leads to the determination that all four vetoes at
    issue, including the Governor's vetoes related to the school bus
    modernization      fund,   local   road    improvement       fund,    and     vapor
    products tax are constitutionally permissible exercises of the
    partial veto power.
    ¶116 Accordingly, I concur in part and dissent in part to
    the per curiam opinion.
    I
    ¶117 Petitioners brought this case as an original action
    against Governor Tony Evers, seeking to invalidate four partial
    vetoes the Governor made to the 2019-21 biennial budget bill.
    They asked this court to determine whether under the partial
    veto power as granted by the Wisconsin Constitution1 the governor
    may permissibly strike portions of a law that are "essential,
    integral,    and     interdependent       parts   of    those        which    were
    approved."      Additionally, they ask us to address whether the
    governor may strike words so as to transform the meaning and
    purpose of a law, essentially turning it into a different law.
    ¶118 The argument petitioners made rested on the assertion
    that this court should overrule a laundry list of longstanding
    precedents    regarding      the    governor's     partial        veto       power.
    However,    they   focus   their   discussion     on   two     specific      cases,
    State ex rel. Wisconsin Telephone Co. v. Henry, 
    218 Wis. 302
    ,
    
    260 N.W. 486
     (1935), and State ex rel. Kleczka v. Conta, 82
    1  Pursuant to Article V, § 10(1)(b) of the Wisconsin
    Constitution, "Appropriation bills may be approved in whole or
    in part by the governor, and the part approved shall become
    law."
    3
    No.    2019AP1376-OA.awb
    Wis. 2d 679,         
    264 N.W.2d 539
             (1978).            Petitioners       readily
    acknowledge that without overruling our long-term precedents,
    their argument cannot stand.2
    ¶119 Chief          Justice       Roggensack's            opinion      declines      to
    overrule      any   of     our    precedents         and   upholds     only    two   of   the
    vetoes         at        issue.                 Chief        Justice          Roggensack's
    concurrence/dissent,             ¶¶71,    82.        In    contrast,       Justice   Kelly's
    concurrence/dissent           would       affect      a    sea    change      in   the    law,
    overruling or modifying multiple cases and upholding none of the
    four       vetoes   at   issue.          Justice     Hagedorn's      concurrence         would
    "revisit" some of our prior cases (although it does not say
    which ones), and would strike down three of the vetoes at issue
    while upholding one.             I address each opinion in turn.
    II
    ¶120 Chief          Justice       Roggensack's            opinion      grounds      its
    analysis with a citation to State ex rel. Wisconsin Senate v.
    Thompson, 
    144 Wis. 2d 429
    , 437, 
    424 N.W.2d 385
     (1988), asserting
    that Wisconsin Senate "explained that the consequences of any
    partial veto must be a law that remains consistent with the
    topic or subject matter of the 'whole' bill."                                Chief Justice
    Roggensack's concurrence/dissent, ¶60.                        The actual language of
    At oral argument, counsel for Petitioners acknowledged
    2
    that accepting Petitioners' position would require the court to
    overrule several cases, which include: State ex rel. Wis. Tel.
    Co. v. Henry, 
    218 Wis. 302
    , 
    260 N.W. 486
     (1935); State ex rel.
    Martin v. Zimmerman, 
    233 Wis. 442
    , 
    289 N.W. 662
     (1940); State ex
    rel. Sundby v. Adamany, 
    71 Wis. 2d 118
    , 
    237 N.W.2d 910
     (1976);
    State ex rel. Kleczka v. Conta, 
    82 Wis. 2d 679
    , 
    264 N.W.2d 539
    (1978); State ex rel. Wis. Senate v. Thompson, 
    144 Wis. 2d 429
    ,
    
    424 N.W.2d 385
     (1988); and Citizens Util. Bd. v. Klauser, 
    194 Wis. 2d 484
    , 
    534 N.W.2d 608
     (1995).
    4
    No.    2019AP1376-OA.awb
    Wisconsin   Senate        sets    forth   that      "the   consequences      of    any
    partial veto must be a law that is germane to the topic or
    subject matter of the vetoed provisions."                     Wis.     Senate, 
    144 Wis. 2d at 437
    .
    ¶121 In the opinion's view, the vetoes that "change the
    school bus modernization fund into an alternative fuel fund" and
    "change the local road improvement fund into a fund for local
    grants or local supplements" fail this inquiry.                      Chief Justice
    Roggensack's       concurrence/dissent,          ¶11.         The     school       bus
    modernization      fund    veto    altered    the    original      law's   topic    or
    subject matter because, as the opinion posits, "the enrolled
    bill says nothing about an 'alternative fuel fund.'                        The parts
    of the enrolled bill that remain after this veto have nothing to
    do with school buses; indeed, the remaining part has nothing to
    do with schools or even education."              Id., ¶99.
    ¶122 Similarly,         Chief    Justice     Roggensack's       opinion    views
    the local road improvement fund veto as altering the topic or
    subject matter of the original law as passed by the legislature.
    It contends that "[a] general undirected fund was not part of a
    fund created to improve local roads because a general fund can
    be spent on virtually any subject, i.e., topics and subject
    matters    never    considered      by    the    legislature."         Id.,       ¶103.
    Consequently, the opinion concludes that these two vetoes are an
    unconstitutional use of the governor's partial veto and are thus
    invalid.
    ¶123 The first problem with this approach is that no party
    advocated for it.         Thus, it has not been tested by the rigors of
    5
    No.    2019AP1376-OA.awb
    appellate advocacy, i.e., briefing and oral argument.                             Deciding
    a case based on a theory not argued by any party not only
    blindsides the parties and sidesteps their input, but it also
    too often results in an inadequate guidance and ill-conceived
    legal analysis.         Yet Chief Justice Roggensack's opinion, on its
    own, selects language from the Wisconsin Senate opinion that
    never before has been the basis of a partial veto decision,
    rewords it, and runs with it.
    ¶124 In Wisconsin Senate, the court wrote:
    We also accept, and for the first time in this case
    give explicit judicial recognition to, the long-
    standing practical and administrative interpretation
    or modus vivendi between governors and legislatures,
    that the consequences of any partial veto must be a
    law that is germane to the topic or subject matter of
    the vetoed provisions.
    Wis. Senate, 
    144 Wis. 2d at 437
    .
    ¶125 The       opinion    claims        that    it   breaks      no   new     ground.
    Chief      Justice    Roggensack's           concurrence/dissent,           ¶107.        But
    neither Wisconsin Senate nor any other case has been explicitly
    decided based on the "topic or subject matter" limitation the
    Wisconsin Senate court referred to as "germaneness."                              In other
    words, no veto has ever been struck down because the resulting
    law   is    not    related     to    the     topic    or   subject      matter      of   the
    original law.
    ¶126 To      support     its     rationale,        the   opinion      cites     three
    instances      where     the        topic     or     subject    matter       limitation,
    "sometimes referred to as germaneness," has been "discussed" in
    prior cases:         Wisconsin Senate, 
    144 Wis. 2d at 451-52
    , Risser v.
    Klauser,     
    207 Wis. 2d 176
    ,           183,   
    558 N.W.2d 108
           (1997),      and
    6
    No.    2019AP1376-OA.awb
    Citizens Utility Board v. Klauser, 
    194 Wis. 2d 484
    , 506, 
    534 N.W.2d 608
           (1995).                 Chief      Justice           Roggensack's
    concurrence/dissent, ¶107.          However, none of these cases used
    topic or subject matter as a reason for striking down a partial
    veto.
    ¶127 Although        the    Wisconsin        Senate    court       stated        the
    limitation that "the consequences of any partial veto must be a
    law that is germane to the topic or subject matter of the vetoed
    provisions[,]"      it      ultimately       held     that         "the      governor
    may . . . veto individual words, letters and digits, and also
    may reduce appropriations by striking digits, as long as what
    remains after veto is a complete, entire, and workable law."
    Wis. Senate, 
    144 Wis. 2d at 437
    .            Wisconsin Senate does not give
    any examples of what is "germane to the topic or subject matter
    of the vetoed provisions" and what is not.                 See 
    id.
            It further
    does not provide any guidance in making such a determination.
    ¶128 In Risser, the court merely cites in passing that "the
    disapproval of part of an appropriation bill may not result in a
    provision which is 'totally new, unrelated or non-germane' to
    the original bill."         Risser, 
    207 Wis. 2d at
    183 (citing Wis.
    Senate, 
    144 Wis. 2d at 451-53
    ).            However, it did not include any
    analysis of germaneness.
    ¶129 In      Citizens      Utility     Board,    the     court        provided    a
    cursory analysis of topic or subject matter ("germaneness"), but
    it was limited to the following:
    There also can be no dispute that sec. 15 of 1993
    Senate Bill 44, as partially vetoed by the governor,
    survives the "topicality" or "germaneness" requirement
    as set forth in Wisconsin Senate.    The new provision
    7
    No.    2019AP1376-OA.awb
    approved by the governor——"$250,000"——relates to the
    same subject matter as the original legislative
    enactment, viz., a money appropriation to be utilized
    by CUB as a public interest advocacy entity.
    Citizens Util. Bd., 
    194 Wis. 2d at 505
    .                        Again, little can be
    gleaned from this regarding the meaning of Wisconsin Senate's
    germaneness        limitation,      which          the    opinion     terms    "topic      or
    subject matter."
    ¶130 Far     from    supporting         the       argument   presented        in   the
    opinion, these cases serve to demonstrate the novelty of its
    theory.      Although Wisconsin Senate articulated the "germaneness"
    limitation, no case has rejected a gubernatorial partial veto
    for defying it or even truly defined what it means.                           Rather than
    "breaking     no    new    ground,"      Chief       Justice    Roggensack's         opinion
    thus bases its decision on a scantily referenced limitation,
    rewords it, and attempts to transform it into the dispositive
    test for a partial veto analysis.                        If this court is to address
    the meaning of the "germaneness" language in Wisconsin Senate,
    we should wait for a case where the parties present the issue
    rather than raise it of our own accord without the benefit of
    advocacy.
    ¶131 The second problem with the approach advanced in the
    opinion is that it provides no clarity where clarity is sorely
    needed.      The proffered "topic or subject matter" test morphs
    into    an    alternative         test    as       the     analysis       unfolds.        The
    alternative test eschews the "topic or subject matter" language
    and instead focuses on an amorphous concept of what was the
    "legislative       idea    that    initiated         the    enrolled      bill."      Chief
    Justice Roggensack's concurrence/dissent, ¶11.                         But neither test
    8
    No.   2019AP1376-OA.awb
    provides any guidance at all.                  Further, such alternatives will
    surely breed more litigation regarding what test to apply and
    the   meaning    of   such    terms       as       "topic,"    "subject       matter,"   or
    "legislative idea" behind an enrolled bill.3
    ¶132 Previous      cases       are       clear      that   in     evaluating       the
    constitutionality of a governor's exercise of the partial veto,
    we apply an objective test.                 Premised on the language of our
    state      constitution,      this         "objective          test     permit[s]        the
    affirmative use of the partial veto power as long as the parts
    remaining after the veto are a complete and workable law."                            Wis.
    Senate, 
    144 Wis. 2d at 453
    .
    ¶133 Chief      Justice    Roggensack's            opinion's      approach     moves
    away from an objective analysis, and exposes determinations on
    the   constitutionality         of    a    partial        veto   to     the    subjective
    preferences of judges.               To explain, the "topic" or "subject
    matter" of an enrolled bill is subject to manipulation.                           It is a
    function of the lens through which the bill is viewed.                             As the
    opinion acknowledges, "topic" or "subject matter" can be broadly
    or        narrowly     viewed.                 Chief          Justice         Roggensack's
    concurrence/dissent, ¶91 ("Clearly, the evaluation of 'part' and
    'whole'     in   § 10(1)(b)     depends            on   how   broadly    the     topic   or
    subject matter is defined.").
    3To further illustrate the amorphous concept of "the
    legislative idea that initiated the enrolled bill," an image
    comes to mind:   two legislators, after hours, are sitting at a
    local pub across the street from the state capitol.      As one
    drinks a beer, he looks at his fellow legislator, announcing,
    "Hey, I have an idea."    Who knows whose idea and what kind of
    idea will meet this amorphous "legislative idea" test, and the
    opinion fails to explain.
    9
    No.   2019AP1376-OA.awb
    ¶134 Favoring         a     narrow       interpretation          of    "topic,"         the
    opinion states with respect to the school bus modernization fund
    veto:     "The parts of the enrolled bill that remain after this
    veto have nothing to do with school buses; indeed, the remaining
    part has nothing to do with schools or even education."                                     Id.,
    ¶99.
    ¶135 But       through      a    broad    lens,     the    "topic"         or    "subject
    matter" of the bill changes.                 The "topic" or "subject matter" of
    the    legislation         could    easily      be   viewed      as    not     school      buses
    specifically, but vehicle efficiency generally.                                Through this
    lens,     the    remaining          alternative       fuel       provision         is    surely
    "germane" to the "topic" or "subject matter" of the legislation.
    ¶136 Similarly,            the   local    road    improvement          fund      veto    is
    characterized         by    the    opinion      as   the    creation         of   a     "general
    undirected fund" that "was not part of a fund created to improve
    local roads because a general fund can be spent on virtually any
    subject, i.e., topics and subject matters never considered by
    the legislature."            Id., ¶103.          But is the "topic" or "subject
    matter"    of    the       original       legislation       local       road      improvement
    specifically      or        the     appropriation          of    money       to    localities
    generally?       Both are reasonable readings, and deciding between
    the two requires a subjective determination.
    ¶137 The approach of Chief Justice Roggensack's opinion has
    taken an area of Wisconsin law that has been quite clear and
    based on an objective test, and injected it with subjectivity.
    Our     case    law    clearly          indicates     that       the     governor        has    a
    constitutional partial veto power that is broad, in fact much
    10
    No.    2019AP1376-OA.awb
    broader than that provided by other states.                     Wis. Senate, 
    144 Wis. 2d at
    439-40 (citing Henry, 218 Wis. at 313); see also John
    S. Wietzer, The Wisconsin Partial Veto:                  Where Are We and How
    Did We Get Here?           The Definition of "Part" and the Test of
    Severability, 
    76 Marq. L. Rev. 625
    , 645-46 (1993) (referring to
    Wisconsin's partial veto power as "uniquely broad"); Anthony S.
    Earl, Personal Reflections on the Partial Veto, 
    77 Marq. L. Rev. 437
    , 438 (1994) (discussing the governor's "broad power to veto
    parts of appropriation bills").
    ¶138 Yet the opinion exposes that broad veto power to the
    serendipity of what lens the judge subjectively chooses.                           This
    would have the effect of inevitably inserting the court into
    policy    disputes   between     the    other     branches      of   government,      a
    result this court has previously considered undesirable.                            See
    State    ex   rel.   Sundby     v.    Adamany,    
    71 Wis. 2d 118
    ,        134,    
    237 N.W.2d 910
     (1976) (rejecting the argument that an affirmative
    policy change constitutes an unconstitutional use of the partial
    veto power).
    ¶139 A commentator has correctly observed three reasons for
    steering clear of subjective considerations in the evaluation of
    the constitutionality of partial vetoes.                See Wietzer, supra, at
    648.     First, "a subjective test for partial veto validity would
    foster    uncertainty      in   the    legislative      process . . . ."            Id.
    Second,    "subjective     tests      would   place     the   court    between      the
    executive     and    the      legislature,       with     the       court   assuming
    legislative powers . . . ."             Id.      Finally, "a subjective test
    11
    No.   2019AP1376-OA.awb
    would    involve       the     courts      every       time     a    partial         veto   dispute
    arose."    Id.
    ¶140 These       concerns          ring     true.            Indeed,     the     budgeting
    process of this state benefits from certainty.                                  However, Chief
    Justice Roggensack's opinion takes us farther from that goal.
    It leaves every partial veto subject to challenge by litigation,
    where pursuant to the opinion's approach, judges can manipulate
    the result by injecting their subjective policy preferences into
    the analysis of the constitutionality of a partial veto.
    III
    ¶141 I turn next to address Justice Kelly's opinion, which
    invalidates all four vetoes.                      It proposes "that we respect the
    constitution's structural limitations on what it means for a
    bill      to     be         approved        'in        part.'"             Justice          Kelly's
    concurrence/dissent, ¶217.                  Accordingly, Justice Kelly's writing
    suggests       that    we    add     to    the     current      "complete,            entire,   and
    workable law" test:                 "After exercising the partial veto, the
    remaining part of the bill must not only be a 'complete, entire,
    and     workable       law,'    it        must    also     be       a    law    on     which    the
    legislature       actually          voted;       and    the     part      of    the     bill    not
    approved       must    be     one    of     the        proposed         laws    in    the    bill's
    collection."          Id.
    ¶142 This        approach          suffers        from        several         infirmities.
    First, it embraces a test not argued or briefed by either party.
    Thus, it has not had the benefit of being tested by the fires of
    advocacy.
    12
    No.    2019AP1376-OA.awb
    ¶143 Second,      it    cavalierly       discards       and        overrules       or
    modifies multiple cases constituting 85 years of precedent in
    derogation of the doctrine of stare decisis.4                    It would abandon
    our   partial    veto    precedent      because      Justice     Kelly       deems    our
    precedent, in his view, "wrongly decided."                    Id., ¶206.          I would
    take a more modest approach.
    ¶144 Stare decisis, the principle that requires courts to
    "stand by things decided," is fundamental to the rule of law.
    Johnson Controls, Inc. v. Emp'rs Ins. of Wausau, 
    2003 WI 108
    ,
    ¶94, 
    264 Wis. 2d 60
    , 
    665 N.W.2d 257
    .                   "This court follows the
    doctrine of stare decisis scrupulously because of our abiding
    respect for the rule of law."            
    Id.
    ¶145 "Fidelity to precedent ensures that existing law will
    not be abandoned lightly.          When existing law is open to revision
    in    every   case,     deciding      cases    becomes    a    mere        exercise       of
    judicial      will,   with     arbitrary       and     unpredictable             results."
    Hinrichs v. DOW Chem. Co., 
    2020 WI 2
    , ¶67, 
    389 Wis. 2d 669
    , 
    937 N.W.2d 37
     (quoting Schultz v. Natwick, 
    2002 WI 125
    , ¶37, 
    257 Wis. 2d 19
    , 
    653 N.W.2d 266
    ).             As a result, any departure from
    stare decisis requires "special justification."                            
    Id.
          Simple
    disagreement     with    a    prior    court's       rationale       is    not     such   a
    "special justification."           Progressive N. Ins. Co. v. Romanshek,
    
    2005 WI 67
    , ¶46, 
    281 Wis. 2d 300
    , 
    697 N.W.2d 417
    .
    Justice
    4         Kelly's  opinion  would  overrule   Sundby,   
    71 Wis. 2d 118
    ; Kleczka, 
    82 Wis. 2d 679
    ; Wisconsin Senate, 
    144 Wis. 2d 429
    ; Citizens Utility Board, 
    194 Wis. 2d 484
    ; and Risser
    v. Klauser, 
    207 Wis. 2d 176
    , 
    558 N.W.2d 108
     (1997); and would
    modify    Henry,     
    218 Wis. 302
    .       Justice     Kelly's
    concurrence/dissent, ¶230 n.14.
    13
    No.    2019AP1376-OA.awb
    ¶146 Third, the interpretation advanced by Justice Kelly's
    opinion     would     render   constitutional        language          superfluous.
    Specifically, the language added to the partial veto provision
    by constitutional amendments in 1990 and 2008 would have no
    effect under the position the opinion takes.
    ¶147 The people of Wisconsin have twice limited the partial
    veto power by constitutional amendment.                   Enacted in 1990 and
    2008, the sum total of these amendments is provided in Article
    V, Section 10(1)(c) of the state constitution:                   "In approving an
    appropriation bill in part, the governor may not create a new
    word   by   rejecting    individual       letters    in    the     words     of   the
    enrolled bill, and may not create a new sentence by combining
    parts of 2 or more sentences of the enrolled bill."
    ¶148 Under the interpretation espoused by Justice Kelly's
    opinion,    this    language   has   no    effect.        If,     as   the   opinion
    posits, the part of the bill not approved must "be one of the
    proposed laws in the bill's collection," then what would be the
    need to proscribe the creation of new words or new sentences as
    set forth in Article V, Section 10(1)(c)?             If Article V, Section
    10(1)(b)    already     prohibits    the    vetoes    described        in    section
    10(1)(c), the language of section 10(1)(c) is mere surplusage.
    ¶149 We are to construe constitutional provisions "to give
    effect to each and every word, clause and sentence" and to avoid
    rendering any language superfluous.              Wagner v. Milwaukee Cty.
    Election     Comm'n,    
    2003 WI 103
    ,    ¶33,    
    263 Wis. 2d 709
    ,      
    666 N.W.2d 816
        (internal    quotation       and   citation        omitted).        The
    14
    No.   2019AP1376-OA.awb
    interpretation advanced in the opinion runs directly counter to
    this established mode of constitutional interpretation.
    ¶150 Finally, Justice Kelly's opinion posits that the court
    has gone astray by "compar[ing] our partial veto to the 'line-
    item' vetoes adopted by some of our sister states and, assuming
    the different words meant Wisconsin must have done something
    very    much   different     from    the     others,     we   consulted         them   no
    further."      Justice Kelly's concurrence/dissent, ¶182.                  Yet there
    is a difference between a "partial" and an "item" veto, as our
    precedent      recognizes.         Wis.    Senate,      
    144 Wis. 2d at
            439-40
    (citing Henry, 218 Wis. at 313).                 The opinion does not account
    for the difference and would, as a practical matter, result in
    an "item" veto in spite of Wisconsin's unique constitutional
    language.
    IV
    ¶151 Next, I turn to address Justice Hagedorn's opinion,
    which     concludes    that    three       of    the    vetoes      at    issue        are
    unconstitutional and that one, the vehicle fee schedule veto,
    passes    constitutional       muster.          After    disavowing       each     test
    proposed by both the parties and members of this court, the
    opinion    states     that    "[w]hile     future       litigation       will    surely
    provide    opportunities      to    refine      the   analysis,     the    principles
    derived from our constitutional text, structure, and early cases
    draw sufficient lines to decide this case."                   Justice Hagedorn's
    concurrence, ¶264.
    ¶152 Those principles lead Justice Hagedorn's opinion to
    this essential inquiry:            "whether the governor vetoed a policy
    15
    No.   2019AP1376-OA.awb
    the legislature proposed and passed, which is permissible, or
    created a new policy the legislature did not propose or pass,
    which is not."         Id., ¶263.      "[W]hat the governor may not do is
    selectively edit parts of a bill to create a new policy that was
    not   proposed    by    the    legislature.          He    may    negate       separable
    proposals actually made, but he may not create new proposals not
    presented in the bill."             Id., ¶264.      In the opinion's view, all
    of the subject vetoes with the exception of the vehicle fee
    schedule veto fail this inquiry.
    ¶153 Justice       Hagedorn's       writing         suffers      from      several
    analytical      shortcomings.          First,       like    both       Chief    Justice
    Roggensack's opinion and Justice Kelly's opinion, it advances a
    theory    not   specifically        argued     by   any    party.        Indeed,      the
    opinion explicitly disavows each test proposed by the parties in
    this case.       Id., ¶¶259-63.          Thus, the parties are deprived of
    the opportunity to analyze and offer comment on this proposed
    theory.
    ¶154 Second, although the opinion appears reticent to say
    so, it would discard a significant amount of our precedent.
    Justice Hagedorn's opinion would keep Henry intact, but would
    "revisit" our "later cases . . . insofar as they abandoned the
    core principles undergirding the way laws are made pursuant to
    our constitution."          Id., ¶266.
    ¶155 Which       of     the     court's       "later       cases"        must    be
    "revisited?"      In a footnote, the opinion reveals that Kleczka is
    one of these cases, and that it must be overruled rather than
    merely "revisited."          Id., ¶266 n.11.        But the opinion also calls
    16
    No.   2019AP1376-OA.awb
    into question the entirety of our partial veto jurisprudence.
    It asserts that "[i]nsofar as our later decisions have treated
    Kleczka as pronouncing that a veto shall stand simply if it
    leaves a complete, entire, and workable law, these statements
    too must be withdrawn."            Id.
    ¶156 Yet, our court has never applied any test other than
    the "complete, entire, and workable law" test.                           Thus, although
    obscured in a footnote, Justice Hagedorn's opinion would tear
    down a substantial amount of our precedent.                      As explained above,
    such a position disregards the principle of stare decisis, which
    is    essential   to      the   rule     of      law.      Johnson        Controls,          
    264 Wis. 2d 60
    , ¶94.
    ¶157 Third,       Justice    Hagedorn's          proposed       test    injects       an
    element of subjectivity into partial veto decisions.                                In the
    opinion's view, the essential inquiry is "whether the governor
    vetoed a policy the legislature proposed and passed, which is
    permissible, or created a new policy the legislature did not
    propose or pass, which is not."                  Justice Hagedorn's concurrence,
    ¶263.    As with the test proposed in Chief Justice Roggensack's
    writing, such an inquiry is susceptible to manipulation and to
    the    subjective    preferences         of      judges.         The     "policy"       of    a
    proposed bill is just as amorphous as the "topic or subject
    matter" of the proposed bill.
    ¶158 For     example,        with         regard    to      the        school     bus
    modernization     veto,     Justice      Hagedorn's        opinion        suggests      that
    "[t]he     legislature's           budget        bill     did      not        propose        an
    appropriation       in    whole     or     in     part     for     alternative         fuels
    17
    No.    2019AP1376-OA.awb
    generally.      Instead, the legislature proposed an appropriation
    for the replacement of school buses."                  Id., ¶271.        Again, what
    the "policy proposal" is depends on the lens through which the
    bill is viewed.         See supra, ¶¶134-35.
    ¶159 By     asserting      that     "future     litigation         will    surely
    provide opportunities to refine the analysis," Justice Hagedorn
    acknowledges      the    instability     in     the   rule   of    law    that    these
    separate writings generated.                  Justice Hagedorn's concurrence,
    ¶264.    Without a clear rule, how will future courts know how to
    apply this law?         They won't.      How can governors be assured that
    the partial veto they are crafting is constitutional?                              They
    can't.    What is to happen if money has been paid or contracts
    signed based on the statutory language as it currently exists?
    Those who would strike down the vetoes provide no guidance.
    ¶160 Indeed, there will be future cases needed to iron out
    the wrinkled mess we leave to the people of this state as this
    court's partial veto jurisprudence.
    V
    ¶161 Instead of Chief Justice Roggensack's approach that
    would    inject    subjectivity         into    an    objective     test,       Justice
    Kelly's approach that would discard decades of case law, or
    Justice Hagedorn's approach that would do both, I would apply
    the time-honored test informed by our precedent.                         That is, we
    ask   whether     "the    part   of     the    bill   remaining     constitutes       a
    'complete, entire, and workable law.'"                  Risser, 
    207 Wis. 2d at
    183 (citing Henry, 218 Wis. at 314; State ex rel. Martin v.
    18
    No.   2019AP1376-OA.awb
    Zimmerman, 
    233 Wis. 442
    , 450, 
    289 N.W. 662
     (1940)); see Wis.
    Senate, 
    144 Wis. 2d at 453
    .
    ¶162 The resulting law after the school bus modernization
    veto       is    clearly       complete,      entire,    and     workable.5        As    Chief
    Justice Roggensack's opinion sets forth, the law after the veto
    states:              "The department shall establish a program to award
    grants          of    settlement      funds    from     the    appropriation       under      s.
    20.855(4)(h) for alternative fuels."                         Chief Justice Roggensack's
    concurrence/dissent, ¶16.                  This resulting sentence is complete
    and    workable          on     its     face,    providing          clear    direction        on
    administration of the subject grants.
    ¶163 Likewise, the local road improvement fund veto leaves
    a complete, entire, and workable law.                               After the local road
    improvement            veto,    § 126   of    the     budget    bill     states:        "Local
    supplement . . . 75,000,000."                    Id.,        ¶19.      Relatedly,       § 184s
    provides:             "Local    supplement.           From    the    general    fund,    as    a
    continuing appropriation, the amounts in the schedule for local
    grant."          Id., ¶20.        Although this law does not get high marks
    for grammar, that does not mean it is not complete and workable.
    "Awkward phrasing, twisted syntax, alleged incomprehensibility
    and vagueness are matters to be resolved only on a case-by-case
    basis in which specific challenges to discrete applications of
    the new provisions are raised in a complete factual setting."
    Wis. Senate, 
    144 Wis. 2d at 463
    .
    The vehicle fee schedule veto also results in a complete,
    5
    entire, and workable law, a premise that Petitioners do not
    dispute.    See Chief Justice Roggensack's concurrence/dissent,
    ¶90.
    19
    No.    2019AP1376-OA.awb
    ¶164 Similarly, the vapor products tax veto results in a
    complete, entire, and workable law.            After the Governor's veto,
    the   definition    of     "vapor      product"    is   set        forth    as     "a
    noncombustible     product     that    produces     vapor     or     aerosol     for
    inhalation from the application of a heating element, regardless
    of whether the liquid or other substance contains nicotine."
    Chief Justice Roggensack's concurrence/dissent, ¶24.                   Again, the
    veto leaves a coherent sentence that is complete, entire, and
    workable on its face.
    ¶165 Rather than embrace the novel and untested approaches
    advanced by each of the other separate opinions, this court
    should tread lightly and act with restraint.                      Such approaches
    foment confusion and inevitably will lead to more litigation.
    ¶166 The   majority      of   the   court    likewise       engenders      more
    litigation with the relief it affords.              The petitioners suggest
    that if this court finds the vetoes unconstitutional, then we
    consider as possible relief "remanding to the Governor to allow
    him to reconsider the relevant sections and either approve them
    in whole, veto them in whole, or veto them in part consistent
    with this Court's opinion."            Such a suggestion for this court,
    however, proves to be much too restrained.
    ¶167 Instead,      the   court      grants   an   alternative        relief,
    choosing to do an end run around the Governor.                    The per curiam
    opinion announces that the school bus modernization fund, local
    roads improvement fund, and vapor products tax are "in full
    force and effect as drafted by the legislature."                      Per curiam,
    ¶9.
    20
    No.    2019AP1376-OA.awb
    ¶168 Arguably, the constitution requires a remand to the
    Governor.    The Wisconsin Constitution provides for only two ways
    for a bill to become law:                if the governor approves and signs
    the bill, Wis. Const. art. V, § 10(1)(b), or if the legislature
    overrides the governor's veto.                  Wis. Const. art. V, § 10(2).
    Neither occurred here.
    ¶169 Citing Sundby, 
    71 Wis. 2d at 125
    , the per curiam seeks
    support for the action it takes.                Specifically, the Sundby court
    set forth:        "If, in fact, the partial vetoes are invalid, the
    secretary    of    state      has    a   mandatory       duty    to     publish       those
    sections of the enactment as if they had not been vetoed."                                
    Id.
    However,    the    statement        in   Sundby   is     not    accompanied         by    any
    constitutional analysis and comes in the context of deciding
    whether the secretary of state was a proper party.                                 That's a
    pretty slim reed to use as support for the constitutionally
    questionable relief the majority grants.
    ¶170 The people of this state deserve stability in the law
    and   clarity     in    our   opinions.         This   court     should          uphold   and
    follow our well-established precedent.                   Based on that precedent
    and the test it establishes, I determine that all four vetoes at
    issue   should     be    upheld      because      they    result       in    objectively
    complete, entire, and workable laws.
    ¶171 For     the    foregoing       reasons,       I     concur       in    part    and
    dissent in part.
    ¶172 I am authorized to state that Justice REBECCA FRANK
    DALLET joins this concurrence/dissent.
    21
    No.    2019AP1376-OA.dk
    ¶173 DANIEL KELLY, J.               (concurring in part, dissenting in
    part).         What a vexatious thing the word "part" can be, and
    indeed it has vexed us from the day we encountered it in Article
    V of our constitution.             When we first considered what it means
    for a governor to approve an appropriation bill "in part," we
    supposed the people of Wisconsin had adopted something very much
    unlike    the        "line-item   veto"       many     of    our    sister       states   have
    adopted.        Our supposing caused us to dress up the governor as
    the people's legislative agent (with respect to appropriations
    bills)    and        the   legislature     as      the      owner   of      an   exceedingly
    difficult to deploy veto.                 So now appropriation "bills" may
    originate with the governor, and they must surely become law
    unless     a     super-majority          of       both      legislative          houses    say
    otherwise.           Not because the constitution says this is how an
    appropriative law may come to be, but because we have said so.
    And this we have done in obeisance to a single word, a word of
    merely serviceable merit in the ordinary affairs of life, but on
    which we have conferred the gigantic power to swap the governor
    for      the     legislature        when          an     appropriation            is      under
    consideration.
    ¶174 The balance of my discourse, I trust, will accomplish
    three things.           First, I mean to describe the mechanism provided
    by the constitution for the enactment of laws.                              Second, I will
    recount        how     our   partial-veto          jurisprudence         has      completely
    disassembled that mechanism and reconstructed it with the parts
    all out of place.              And third, I will propose we retire our
    suppositions and instead consult the constitution's actual text
    1
    No.    2019AP1376-OA.dk
    to    learn       what       it     means   for       a     governor     to       approve    an
    appropriation bill "in part."
    I.   SCHOOLHOUSE ROCK
    ¶175 A law begins as someone's idea.                         Somewhere, for some
    often-unknown reason, it strikes someone that something within
    the government's purview ought to be required, or prohibited, or
    changed.      Through whatever pathways the idea might travel, it
    eventually comes to the attention of a legislator.                                  And if the
    idea finds there a receptive audience, the legislator engages
    the constitutional mechanism for turning the idea into a law.
    It    must   be    a    legislator        (as   opposed       to,   say,      the    governor)
    because the power to make the law is legislative.                                   Schmidt v.
    Dep't of Res. Dev., 
    39 Wis. 2d 46
    , 59, 
    158 N.W.2d 306
     (1968)
    (The legislative power is the power "'to declare whether or not
    there shall be a law; to determine the general purpose or policy
    to be achieved by the law; [and] to fix the limits within which
    the    law    shall          operate[.]'"       (quoting        State        ex     rel.    Wis.
    Inspection Bureau v. Whitman, 
    196 Wis. 472
    , 505, 
    220 N.W. 929
    (1928))); see also Wis. Legislature v. Palm, 
    2020 WI 42
    , ¶92,
    
    391 Wis. 2d 497
    ,               
    942 N.W.2d 900
            (Kelly,      J.,       concurring)
    (describing the legislative power as the ability to determine
    and declare what the laws and policy of the state will be).                                  And
    according     to       the    unambiguous       and       unqualified    command       of   our
    constitution, "[t]he legislative power [is] vested in a senate
    and assembly."           Wis. Const. art. IV, § 1.
    ¶176 The legislative process must begin with the drafting
    of a bill to contain the championed idea because "[n]o law shall
    2
    No.    2019AP1376-OA.dk
    be enacted except by bill."                     Wis. Const. art. IV, §17(2).                    When
    the drafting is done, the bill contains a complete and workable
    potential       law,      which     is    then    introduced         to    the     legislature:
    "Any      bill        may         originate           in     either        house         of     the
    legislature . . . ."                     Wis.     Const.       art.        IV,      § 19        (the
    "origination         clause").            There       is,    obviously,          correspondence
    between the houses because a bill cannot become a law until
    approved       by    both:        "Every    bill       which       shall    have     passed      the
    legislature shall, before it becomes a law, be presented to the
    governor."          Wis. Const. art. V, § 10(1)(a) (the first clause is
    the     "legislative         passage        clause,"         and     the     second        is    the
    "presentment clause").                And in that correspondence, each house
    may    modify       the   proposed        law    considered         by    the     other.        Wis.
    Const. art. IV, § 19 ("[A] bill passed by one house may be
    amended by the other.") (the "amendment clause").
    ¶177 Once both houses have agreed upon a bill, it comes
    under the governor's scrutiny as it passes from the legislative
    branch to the executive branch.                       Wis. Const. art. V, § 10(1)(a).
    The bill becomes a law "[i]f the governor approves and signs the
    bill . . . ."          Wis. Const. art. V, § 10(1)(b).                       The process for
    appropriation bills (which is our particular topic of interest
    here)    is,     however,     a     little       different.           Such       bills   "may     be
    approved in whole or in part by the governor, and the part
    approved shall become law."                 Id.        But the governor's disapproval
    of    some   part      of    an    appropriation            bill    does     not    necessarily
    identify its terminus.                   Instead, the rejected part returns to
    the legislative branch for further consideration.                                 If two-thirds
    3
    No.    2019AP1376-OA.dk
    of the members of both houses approve, the rejected part becomes
    law notwithstanding the governor's disapproval.1
    ¶178 I beg forgiveness for this pedantry, but I find that
    our     partial   veto       jurisprudence      requires      recourse         to    these
    fundamental principles so that we may recover the law-making
    process provided by our constitution.                     We have before us two
    potential     understandings          of   what    it     means     to     approve        an
    appropriations bill "in part."                 One is extraordinarily broad,
    and in consequence of its broadness it rejects almost every
    other     piece   of    the    legislative        machinery    described            in   our
    constitution.          The    other   is   much    more    modest,       but    has      the
    benefit of leaving the pieces of the legislative machinery where
    the constitution put them, and in its operation it precisely
    answers the problem it was meant to solve.
    ¶179 I believe we should adopt the latter understanding in
    no small part because one of the fundamental rules of textual
    interpretation is that, when given a choice, we do not read one
    1   Wis. Const. art. V, § 10(2)(b):
    The rejected part of an appropriation bill,
    together with the governor’s objections in writing,
    shall be returned to the house in which the bill
    originated.     The house of origin shall enter the
    objections at large upon the journal and proceed to
    reconsider the rejected part of the appropriation
    bill.    If, after such reconsideration, two−thirds of
    the members present agree to approve the rejected part
    notwithstanding the objections of the governor, it
    shall be sent, together with the objections, to the
    other    house,   by  which  it   shall   likewise  be
    reconsidered, and if approved by two−thirds of the
    members present the rejected part shall become law.
    4
    No.   2019AP1376-OA.dk
    constitutional provision to conflict with others.                         See Thomas M.
    Cooley, A Treatise on the Constitutional Limitations Which Rest
    upon the Legislative Power of the States of the American Union,
    58 (1868) ("[O]ne part is not to be allowed to defeat, if by any
    reasonable       construction     the       two        can    be     made     to    stand
    together."); Antonin Scalia & Bryan A. Garner, Reading Law:                            The
    Interpretation of Legal Texts 180 (2012) ("The provisions of a
    text    should     be   interpreted         in     a    way       that    renders    them
    compatible, not contradictory.").                 So we construe constitutional
    provisions with the assumption that they are all supposed to
    function together in concert.               When faced with two permissible
    constructions of the word "part," we must choose the one that
    harmonizes with other relevant text.                   A reading that introduces
    dissonance is a powerful hint that we're doing it wrong.
    ¶180 The tuning fork by which I will test for harmony and
    dissonance     comprises     three      interrelated          propositions          called
    forth by our constitution's text.                 The first proposition is that
    the    most   elemental    part   of    a       bill   is    an    idea   (that     is,   a
    proposal for a complete, entire, and workable law).                          The second
    is that the powers of amending and vetoing are different things,
    the respective exercise of which our constitution commits to
    different branches of government.                And the third is that an idea
    may not become law without the legislature having first voted
    for it.       It seems remarkable to me that I should be offering
    these as propositions rather than as settled descriptions of
    constitutional principles, but our partial-veto jurisprudence is
    at odds with each of them.              And that means all I can do is
    5
    No.    2019AP1376-OA.dk
    recommend them to the attention of future courts who may be
    called upon to consider the meaning of Wis. Const. art. V, § 10.
    II.     ON THE JUMBLING OF THE LEGISLATIVE MECHANISM
    ¶181 Great variances often begin as minor imprecisions, and
    such is the case with the path we traveled over the years as we
    addressed the partial veto.                 I will detail only enough of that
    journey to describe how we disassembled some of the key pieces
    of    the    legislative       mechanism     and    then    reassembled       them    into
    something that is constitutionally unrecognizable.
    A.    The Disassembly
    ¶182 We first entertained a claim that the governor had
    improperly employed his partial veto power in                            State ex rel.
    Wisconsin Tel. Co. v. Henry, 
    218 Wis. 302
    , 260 N.W 486 (1935).
    Having never encountered such a veto before, we sensibly looked
    about for tools to help us understand its telos.                         Our first step
    was    to    compare     our    partial      veto   to     the   "line-item"        vetoes
    adopted by some of our sister states and, assuming the different
    words       meant    Wisconsin       must    have    done    something       very     much
    different from the others, we consulted them no further.                            It was
    certainly fair to observe that a partial veto must differ in
    some measure from a line-item veto——the word-choice suggests as
    much.       But it was a mistake to suppose the measure of difference
    was so great that other states' experience with vetoes of less
    than an entire bill could tell us nothing about their impact on
    the overall law-making mechanism.                   So we missed out on what we
    might have learned about whether such vetoes have any effect on
    the    vesting      of   legislative        authority,      or   the     origination    of
    6
    No.    2019AP1376-OA.dk
    bills, or the difference between amendments and vetoes, or the
    need for the legislature to vote on a proposed law.                            Finding no
    pedagogical     value       in   the   partial      veto's       cousin,       we   instead
    consulted a dictionary wherein, unknowingly, we found mischief.
    ¶183 We learned from Webster's New International Dictionary
    that "part" means
    one of the portions, equal or unequal, into which
    anything is divided, or regarded as divided; something
    less than a whole; a number, quantity, mass, or the
    like, regarded as going to make up, with others or
    another, a large number, quantity, mass, etc., whether
    actually separate or not; a piece, fragment, fraction,
    member, or constituent.
    Henry, 218 Wis. at 313 (quoting Part Webster's New International
    Dictionary 1781       (2d    ed.     1934)).        This     provided      a   reasonably
    adequate    etymological         meaning;2        but    what      we     needed     was    a
    constitutionally contextualized meaning.                     Antonin Scalia & Bryan
    A.   Garner,    supra       at 427     (We   consider        a   word's     meaning        "in
    context according to a fair reading.").                      That is, we needed to
    discover    the      most    elemental       part       of   a     bill,    the     further
    subdivision of which leaves something no longer identifiable as
    a part of a bill.           If we had done this work then, it would have
    saved us from concluding in subsequent cases (which I address
    below) that the most elemental part of a bill is not an idea,
    but instead a letter or a digit.
    ¶184 But we did not know then what would be urged upon us
    later,    and   so    our    analysis        in   Henry      was    adequate        for    our
    2A common, contemporaneous dictionary may provide a word's
    generally understood meaning. State v. Sample, 
    215 Wis. 2d 487
    ,
    499–500, 
    573 N.W.2d 187
     (1998).
    7
    No.   2019AP1376-OA.dk
    immediate needs, if not for future cases.                All we needed to do
    there was decide whether the partial veto empowered the governor
    to unbundle what the legislature had bundled——a practice then
    known as "logrolling."            A case we decided a few years later
    neatly summed up the relationship between the problem and the
    solution provided by the partial veto:
    Its purpose [the partial veto] was to prevent, if
    possible, the adoption of omnibus appropriation bills,
    log–rolling, the practice of jumbling together in one
    act inconsistent subjects in order to force a passage
    by uniting minorities with different interests when
    the particular provisions could not pass on their
    separate   merits,   with   riders   of   objectionable
    legislation attached to general appropriation bills in
    order to force the governor to veto the entire bill
    and thus stop the wheels of government or approve the
    obnoxious act. Very definite evils were inherent in
    the   law–making    processes   in    connection   with
    appropriation measures. Both the legislature and the
    people deemed it advisable to confer power upon the
    governor to approve appropriation bills in whole or in
    part . . . .
    State ex rel. Martin v. Zimmerman, 
    233 Wis. 442
    , 447-48, 
    289 N.W. 662
     (1940).       We foreshadowed this conclusion in Henry where
    we observed that "there is nothing in that provision [art. V, §
    10]   which    warrants     the    inference    or     conclusion       that      the
    Governor's    power    of   partial   veto   was   not    intended      to   be    as
    coextensive    as     the   Legislature's      power     to    join    and     enact
    separable pieces of legislation in an appropriation bill."                        218
    Wis. at 315.     The rule we developed in Henry was sufficient to
    meet the problem of logrolling.            It required that the parts of
    the bill remaining after the partial veto "constitute, in and by
    themselves, a complete, entire, and workable law . . . ."                         Id.
    at 314.   Applied in this context, it was a workable rule because
    8
    No.    2019AP1376-OA.dk
    its operation reflected the partial veto's purpose——separating
    ideas the legislature had joined.                               Unfortunately, embedded in
    this rule is an intrinsic deficiency:                             We had neglected to say
    that the "complete, entire, and workable law" remaining after
    the   veto       must    be   one       on    which       the    legislature     had    actually
    voted.       The deficiency was not apparent in Henry because the
    parts of the bill remaining after the veto were the same as they
    had   been       when    transmitted           to    the    governor.         What     we    didn't
    foresee at the time was that a future governor might so employ
    the partial veto that the remaining parts would comprise a law
    the legislature had never seen.
    ¶185 The         rule's     deficiency             bore   fruit   in     State   ex        rel.
    Sundby      v.    Adamany,         
    71 Wis. 2d 118
    ,         
    237 N.W.2d 910
             (1976).
    There, the bill in question gave local taxpayers the option of
    calling for a public referendum before a municipality increased
    its tax levy.            
    Id. at 121-22
    .              But the governor vetoed part of
    one sentence in such a way that the remaining language made the
    referendum mandatory.               The legislature, of course, had neither
    proposed nor approved such a thing.                                The idea had not been
    drafted      as    a    bill,      it    did        not    originate     in     the   senate       or
    assembly, it was not subject to amendment in the corresponding
    legislative house, and no one in the legislature had ever voted
    on    it.     And       yet   we    said       the    gubernatorial-authored                law    was
    constitutionally permissible.                       Why?     Because, apparently, a veto
    has affirmative policy-making powers:
    Some argument is advanced that in the exercise of
    the item veto the governor can negative what the
    legislature   has  done   but  not   bring  about   an
    affirmative change in the result intended by the
    9
    No.   2019AP1376-OA.dk
    legislature. We are not impressed by this argued
    distinction. Every veto has both a negative and
    affirmative ring about it.    There is always a change
    of policy involved.      We think the constitutional
    requisites of art. V, sec. 10, fully anticipate that
    the governor's action may alter the policy as written
    in the bill sent to the governor by the legislature.
    Sundby, 
    71 Wis. 2d at 134
     (emphasis added).                 Every veto has an
    affirmative "ring" about it?              Well, I suppose so, but only in
    the sense that declining a marriage proposal has the "ring" of a
    wedding about it.        A veto cannot be the genesis of a new policy
    any more than telling an amorous suitor "no" means there is a
    reception to plan.        Vetoes and "noes" are for stopping things,
    not creating them.        See Federalist No. 73, 440-41 (Hamilton) (C.
    Rossiter ed. 1961) (The veto power "is the qualified negative of
    the [executive] upon the acts or resolutions of the two houses
    of the legislature; or, in other words, his power of returning
    all bills with objections, to have the effect of preventing
    their becoming laws[.]").
    ¶186 To Chief Justice Roggensack, however, a veto is an
    invitation to participate in law making rather than just law
    stopping.      She says:      "Furthermore, our jurisprudence is not
    unique in describing a quasi-legislative role for the governor.
    A   veto     power,    regardless     of       its   contours,     is    inherently
    legislative."         Chief Justice Roggensack's concurrence/dissent,
    ¶84.       The second sentence is certainly true, but it has no
    connection to what she means by a "quasi-legislative role" in
    the first sentence.        The veto is simply one of the instances in
    which   our    framers    broke     off    a    small   piece    of     power   that
    naturally belongs in one branch and put it in another.                    So, yes,
    10
    No.    2019AP1376-OA.dk
    it is quite obviously legislative in nature.                   But there are no
    penumbras    emanating     from   the   veto       power;    it     authorizes   the
    executive to do nothing more than what it says——stop a law from
    coming into being.         In the Chief Justice's hands, however, the
    veto is a clandestine vehicle for smuggling the legislature's
    law-authoring function into the executive branch where, through
    the power of the word "part," it turns the governor into a
    quasi-legislator (whatever that might be).                    If we are to be
    constrained by the words of the constitution, this operation is
    simply impossible.        So the first sentence of the quote above is
    incorrect.         Our   misguided   jurisprudence          might    describe    the
    governor    as   having    a   "quasi-legislative       role"       beyond   merely
    stopping a proposed law, but literally no other authority in
    these United States does.3
    ¶187 And that brings us back to Henry's unfinished work——
    defining the "thing" that a partial veto may stop.                     The rule we
    adopted in that case assumed, but never stated, that it was a
    bundled    piece    of   legislation.        But    without    a    contextualized
    3 The Chief Justice buttresses the executive's claim to
    legislative powers with reference to its rule-making authority
    (which it borrows from the legislature).      See Chief Justice
    Roggensack's concurrence/dissent, ¶86 ("First, as demonstrated
    by rulemaking, and as we have long concluded, the Legislature
    may delegate its power to make law to the executive."); Koschkee
    v. Taylor, 
    2019 WI 76
    , ¶34, 
    387 Wis. 2d 552
    , 
    929 N.W.2d 600
    ("The source for rulemaking is legislative delegation."). The
    nature, scope, effect, and validity of administrative rule-
    making are subjects of a continually growing body of literature
    that is enormous both in terms of its volume and potential
    constitutional implications.   So this probably isn't the best
    reference if the goal is to show that executive law-making is a
    settled and universally accepted phenomenon.
    11
    No.    2019AP1376-OA.dk
    definitional anchor point for "part," we concluded in Sundby
    that the most elemental part of a bill can be something smaller
    than one of the proposed laws bundled into an appropriation
    bill; we said it could be part of a sentence in one of the
    bundled   proposals,   so    long   as    the   resulting     document   still
    comprised   a   "complete,    entire,     and   workable    law."        So   we
    accepted the veto of a part of a part of an idea even though the
    result expressed an idea not contained in the bill presented to
    the governor.
    ¶188 But wait, there's more.          We've said the most elemental
    part of a bill a veto can stop isn't a sentence, or even part of
    a sentence——it's a letter or a digit:
    Thus, in this opinion, we break no new ground except
    as we now, on the facts before us, have the obligation
    to clarify that the governor may, in the exercise of
    his partial veto authority over appropriation bills,
    veto individual words, letters and digits, and also
    may reduce appropriations by striking digits, as long
    as what remains after veto is a complete, entire, and
    workable law.
    State ex rel. Wis. Senate v. Thompson, 
    144 Wis. 2d 429
    , 437, 
    424 N.W.2d 385
     (1988).     In what came to be known as the Vanna White
    veto, a governor would strike individual letters or numbers to
    create words, sentences, and ideas that appeared nowhere in the
    bill passed by the legislature.4
    4 See, e.g., State ex rel. Wisconsin Senate v. Thompson, 
    144 Wis. 2d 429
    , 460 n.15, 
    424 N.W.2d 385
     (1988):
    12
    No.    2019AP1376-OA.dk
    ¶189 We approved this practice in large part because we
    considered    it     all     part   of    the      governor's       "quasi-legislative"
    role.   
    Id. at 446
    .           Warming to our theme a few pages later, we
    dropped both the "quasi" and any remaining pretense that the
    legislature is the exclusive legislative branch of government:
    "This   broad      and     expansive      interpretation            of   the   governor's
    partial veto authority as mandated by the constitution has, in
    effect, impelled this court's rejection of any separation of
    powers-type     argument       that      the       governor    cannot       affirmatively
    legislate by the use of the partial veto power."                           
    Id. at 453
    .
    ¶190 After releasing our Wisconsin Senate opinion in 1988,
    the   court-approved         method      of   enacting    appropriation          bills    no
    longer bore any resemblance to the mechanism described by our
    constitution.         The three propositions I introduced above, and
    which   I    now      address,      demonstrate         that    our        experience     in
    reconstructing the dismantled legislative process left several
    of the key pieces in the wrong place.
    B.    The Reassembled Legislative Mechanism
    ¶191 The       first    proposition          traduced    by    our     partial    veto
    jurisprudence is that the irreducible part of a bill is an idea—
    —that is, a proposal for a complete, entire, and workable law.
    Governor Lee Sherman Dreyfus used a digit veto to cut
    $8.9 million appropriated for state school aids in the
    1979–81 budget bill. He accomplished this by vetoing
    the decimal point and number 9 from the percentage
    "96.9%", thereby decreasing the percentage used for
    calculating a portion of such school aids. That veto
    was not challenged, and the legislature subsequently
    failed to override it.
    13
    No.   2019AP1376-OA.dk
    This     is     the     first       because        it     necessarily          informs     our
    understanding of the entire legislative mechanism——specifically,
    it   identifies        the    required     entry         point     to    the    legislative
    process,      where    and    how    the   idea         may   be   changed,       and    whose
    approval is needed before the idea may become a law.                                However,
    by treating a bill as a potpourri of letters and digits, rather
    than an expression of one or more complete and comprehensible
    ideas,        our     reconstruction          of        the   legislative          mechanism
    dramatically changed the legislative process.
    ¶192 Our       refutation     of    this     proposition          started    when    we
    looked to a dictionary to learn what "part" means.                                      We had
    recourse       to   that     venerable     source        because,       surprisingly,       we
    didn't think the context in which the constitution used the word
    was significant:
    As the meaning of that word, as used in section 10,
    art. 5, Wis. Const., is not . . . rendered doubtful by
    reason of context, or uncertainty as to application to
    a particular subject–matter, or otherwise, there is
    nothing because of which that word, as used in that
    section, is not to be given its usual, customary, and
    accepted meaning . . . .
    Henry, 218 Wis. at 313.             But it's one thing to understand that a
    "part" is something less than the whole, as the dictionary says;
    it's an entirely different thing to understand what a part of a
    bill might be.
    ¶193 As we learned in Schoolhouse Rock, a bill encompasses
    someone's idea.            The purpose of the bill, of course, is to
    introduce the idea it contains to the legislature, where the
    14
    No.       2019AP1376-OA.dk
    legislators evaluate its merits as a potential law.5                              The fate of
    a   bill   in   each     legislative         house,      therefore,         is     to    be    the
    subject    of     debate.        See       Wis.    Const.       art.    IV,        § 16       (Our
    constitution anticipates a vigorous debate:                           "No member of the
    legislature shall be liable in any civil action, or criminal
    prosecution       whatever,          for      words       spoken        in         debate.");
    Legislature——Public            Officers——Secretary              of     State——Wisconsin
    Statutes, 10 Wis. Op. Att'y Gen. 613 (1921) (Broadly describing
    the legislative process as an introduction of a bill in one
    house, potential amendment in the other, and ultimate agreement
    between    them    before      presentation         to   the    governor).               Debates
    (proper ones, at least) involve reasoning——the setting forth of
    intelligible        arguments          for        or      against           a      rationally
    comprehensible proposal.             Dividing a bill into anything smaller
    completely      destroys       its     distinctive            nature——that              is,    the
    expression of a proposed law susceptible of debate and adoption.
    This is why the basic part of a bill cannot be a letter or a
    digit.     Neither the letter "y" nor the number "5" (nor any of
    their relations) can be, in isolation, a bill because such a
    thing would be incomprehensible in debate or as a law.                                    So the
    irreducible       part    of     any       bill,       even     the     simplest,             most
    uncomplicated,       inconsequential              bill    one    can        imagine,          must
    necessarily be, at a minimum, an idea expressing a potential
    5See, e.g., Follow the Process:    The Legislative Process,
    Wisconsin State Legislature (Last Accessed Jun. 13, 2020),
    https://legis.wisconsin.gov/about/follow   ("When   a  legislator
    gets an idea or is prompted by their constituency to make a
    change, they have a drafting lawyer prepare a draft of a bill to
    see what laws will need to change.").
    15
    No.   2019AP1376-OA.dk
    complete, entire, and workable law.                  This is why Justice Hansen
    said the partial veto "is not a power to reduce a bill to its
    single phrases, words, letters, digits and punctuation marks."
    State      ex    rel.    Kleczka    v.    Conta,     
    82 Wis. 2d 679
    ,        726,    
    264 N.W.2d 539
     (1978) (Hansen, J., concurring in part, dissenting in
    part).
    ¶194 The second proposition is that the powers of amending
    and vetoing are different things, the respective exercise of
    which      our    constitution        commits       to    different        branches    of
    government.           Amending belongs to the legislative houses:                     "[A]
    bill passed by one house may be amended by the other."                                Wis.
    Const. art. IV, § 19.               The power to amend a bill comprehends
    changing its meaning:              "When a change is made in a bill, it is
    said       to    be     amended.         There     are    simple     and     substitute
    amendments."6           See also Amend, Black's Law Dictionary (11th ed.
    2019) ("To change the wording of; specif., to formally alter (a
    statute, constitution, motion, etc.) by striking out, inserting,
    or substituting words.")             An amendment may accomplish something
    as minor as subtracting a penny from an appropriation, as major
    as introducing an entirely new idea, or quite literally anything
    in between.           Our constitution commits the power to amend to the
    assembly or senate; it contains no suggestion that the governor
    might be able to partake of it.                  This should have given us pause
    as we were developing our theory of partial vetoes, but instead
    How a Bill Becomes Law, Wisconsin State Legislature 14
    6
    (available                                                     at
    http://legis.wisconsin.gov/assembly/acc/media/1106/howabillbecom
    eslaw.pdf) (May 2016).
    16
    No.    2019AP1376-OA.dk
    we rejected the idea that "the governor cannot affirmatively
    legislate by the use of the partial veto power."                          Wis. Senate,
    
    144 Wis. 2d at 453
    .            This is patent error because it draws the
    amending power into the executive branch in direct and express
    contradiction to the constitution.                    If we say the governor's
    "veto"     may    change      a     bill's    idea,    then     there's       really     no
    cognizable       difference        between   the     concepts    of     amendments      and
    partial    vetoes.          Because     we   failed     to    keep      these      concepts
    distinct, our reconstructed legislative mechanism now allows for
    amendments       in   the    assembly,       the   senate,      and     the   governor's
    mansion.     Obviously, we put the power to amend in the wrong
    place as we were reconstructing the legislative mechanism.
    ¶195 The third (and perhaps most important) proposition is
    that an idea may not become a law without the legislature having
    voted for it.         But when we finished reassembling the legislative
    mechanism,       this       proposition       was,     disturbingly,          no     longer
    categorically true.               If a bill contains an appropriation, our
    reconstruction allows a new idea to originate not as a bill but
    as a partial veto.            It further allows the idea to originate in
    the executive branch instead of the legislative branch.                                And,
    finally, it allows this new idea to become law so long as the
    legislature does not reject it by a two-thirds vote in both
    houses.     So our reconstruction put more legislative pieces in
    the wrong place——we made the governor the author of the law
    (instead of the legislature), and we reduced the legislature to
    wielding a very difficult to deploy veto over the governor's
    edict.     The net effect is that the governor may create a law
    17
    No.   2019AP1376-OA.dk
    without ever having to obtain legislative approval.               In fact, a
    majority of both houses' members may affirmatively reject the
    governor's law, yet it is law nonetheless unless that majority
    is super-sized.
    ¶196 This     reconstructed      mechanism   violates      four   specific
    constitutional requirements.         The first is that all bills must
    originate in one of the two legislative houses, the second is
    that they must be subject to amendment in the corresponding
    house.   Wis. Const. art. IV, § 19 ("Any bill may originate in
    either house of the legislature, and a bill passed by one house
    may be amended by the other.").           The third is that "[n]o law
    shall be enacted except by bill," and the fourth is that the
    bill must be approved by both houses of the legislature.                  Wis.
    Const. art. IV, § 17(2); Wis. Const. art. V, § 10(1)(a) ("Every
    bill which shall have passed the legislature shall, before it
    becomes a law, be presented to the governor." (emphasis added)).
    To the extent a governor's partial veto introduces an idea not
    previously present in the bill, its origin is in the executive
    branch, not the legislature.         And because the new idea did not
    originate in the assembly or senate, it was never subject to
    amendment in the corresponding house.            Finally——and this should
    definitively    dispose   of   our     partial    veto   jurisprudence——it
    allows an idea to become a law even though it has not "passed
    the legislature."
    ¶197 Now, to be sure, the judicially-engineered executive
    legislative power (how's that for a tri-lateral oxymoron?) is
    not as comprehensive as that belonging to the legislature.                  We
    18
    No.   2019AP1376-OA.dk
    have left some limitations in place, which is encouraging even
    if they have nothing to do with the constitution.                      For example,
    when the governor addresses himself to a dollar figure, we allow
    him to make it smaller, not larger.                    Citizens Util. Bd. v.
    Klauser,     
    194 Wis. 2d 484
    ,      488,     
    534 N.W.2d 608
             (1995).
    Presumably, this limit derives from the mathematical principle
    that $10 is a part of $100.          But it still allows introduction of
    an   idea   different      from   the    one    to     which    the     legislature
    assented.      Another      limitation       relates    to     the     letters     the
    governor may use in the creation of new words and ideas:                            We
    have never said he may add letters not already present in the
    bill.      I suppose this is an etymological limit based on the
    proposition    that    a   letter    (as     opposed    to     an    idea)   is    the
    indivisible part of a bill, and so a new letter cannot be said
    to be a part of the existing potpourri.                 Speaking of which, we
    have not said (at least not yet) that he may change the order of
    letters in the potpourri.           This limit almost certainly survives
    because we haven't turned our attention to it.                      If the governor
    may create new words and ideas not already present in the bill,
    it seems like scrupling at a trifle to insist that the letters
    he uses to create them remain in the order presented.                             If a
    letter really is the most elemental part of a bill, it is just
    as much a "part" if it appears before rather than after any of
    the bill's other "parts."           Nothing in the dictionary definition
    of "part" suggests that sequencing has anything to do with it.
    In any event, aside from these few limitations, our cases say
    the governor is free to draft new ideas and we will pretend the
    19
    No.   2019AP1376-OA.dk
    resulting     document     is   still   a    bill    that   has       "passed   the
    legislature" when, quite obviously, it isn't.               As a consequence,
    our cases refute the proposition that no idea shall become a law
    without legislative approval.
    ¶198 So, as far as the Wisconsin Supreme Court is concerned
    (at   least      until    we    were    contradicted        by    a     brace   of
    constitutional amendments),7 because the most elemental part of
    an appropriations bill is a letter, a bill may originate with
    the governor, it is not subject to legislative amendment, and it
    may become the law of Wisconsin even if the legislature has not
    approved    it   (or,    more   shockingly,    has   actually     affirmatively
    voted against it, albeit by less than a supermajority).                         As
    Justice Hansen said,
    [i]t appears that we have now arrived at a stage
    where one person can design his own legislation from
    the appropriation bills submitted to him after they
    have been approved by the majority of the legislature.
    The laws thus designed by one person become the law of
    the sovereign State of Wisconsin unless disapproved by
    two-thirds of the legislators. I am not persuaded that
    7The people of Wisconsin amended their constitution in 1990
    to prevent a veto from "creat[ing] a new word by rejecting
    individual letters in the words of the enrolled bill[.]"     Wis.
    Const. art. V, § 10(1)(c). They amended it again in 2008, this
    time to prevent a veto from creating "a new sentence by
    combining parts of 2 or more sentences of the enrolled bill."
    Id.
    The Chief Justice and Justice Ann Walsh Bradley treat these
    amendments as though they have something to say about the
    meaning of the original partial veto power.         Chief Justice
    Roggensack's   concurrence/dissent,   ¶73;   Justice   Ann    Walsh
    Bradley's concurrence/dissent, ¶146.        They don't.       These
    amendments were directed at us; they were meant to rein in our
    jurisprudential   excesses,  not   limit   the  meaning    of   the
    constitution's actual text.
    20
    No.   2019AP1376-OA.dk
    art. V, sec. 10, was ever intended to produce such a
    result.
    Kleczka,      
    82 Wis. 2d at 727
       (Hansen,     J.,    concurring       in   part,
    dissenting in part).             I agree.        All of this upending of the
    constitutional order we have done because of the word "part," a
    word so meek and mild that it should be entirely incapable of
    wreaking such havoc on our constitutional order.                            This case
    presents an opportunity to return the disordered pieces of the
    law-making machinery to their proper places, and I think we
    should take it.           In fact, I think we are required to take it.
    III.    ON THE DUTY TO RETURN TO THE CONSTITUTIONAL TEXT
    ¶199 The       majority     of   the      court's    members       base    their
    analyses on two propositions.                 The first is that our decision
    here must follow what we have done in our prior cases, even if
    we were wrong before.             And the second is that we must respect
    the governor and legislature's historical practice of allowing
    partial vetoes so long as the resulting legislation is either on
    the same topic as the bill passed by the legislature (according
    to the Chief Justice), or is a "complete, entire, and workable
    law" (according to other members of our court).                             I disagree
    because I believe our obligation to the Wisconsin Constitution
    supersedes         both   of   them.    I    appreciate     the     Chief    Justice's
    opinion because she attempts to cabin in the governor's use of
    the partial veto so that the resulting law is at least on the
    same topic, and in doing so she moves at least part of the way
    back   to   the      constitutional     limitations        on   the     partial   veto.
    Other members of the court would not even attempt that much.                         To
    the extent my opinion responds to others, it focuses primarily
    21
    No.    2019AP1376-OA.dk
    on the Chief Justice's opinion——not because I disagree with her
    the   most   (I     don't),       but    because    in    moving     closer          to   the
    constitution, her opinion helpfully illustrates the remaining
    distance     we     need     to     go    before     we     can      call        ourselves
    constitutionally orthodox.
    A.     What we have done before
    ¶200 "We       cannot        rehash       original         meaning——and             its
    interaction with stare decisis——every time a partial veto comes
    before     us[,]"     the     Chief       Justice        says.           Chief       Justice
    Roggensack's concurrence/dissent, ¶75.                    Maybe.     But if we were
    to address ourselves to the original meaning of the relevant
    constitutional text in this case, we wouldn't be rehashing it,
    we would be analyzing it for the first time.                     In our 85 years of
    experience with the partial veto, we have not once asked how it
    fits with the origination clause, the amendment clause, or the
    legislative passage clause.
    ¶201 Standing between us and the constitution's original
    meaning, however, is a string of cases stretching back over
    those 85 years.       Stare decisis counsels that we tread carefully
    here, and that we not upset what has been settled without a good
    reason.      This principle rests on the premise that we do not
    begin every analysis ab initio mundi; our work builds on the
    accomplishments       of    our     capable     predecessors.              If    a    court
    disregards this premise, there is a risk that "deciding cases
    becomes a mere exercise of judicial will, with arbitrary and
    unpredictable results."             State v. Roberson, 
    2019 WI 102
    , ¶49,
    
    389 Wis. 2d 190
    ,     
    935 N.W.2d 813
            (quoting     Schultz       v.     Natwick,
    22
    No.    2019AP1376-OA.dk
    
    2002 WI 125
    , ¶37, 
    257 Wis. 2d 19
    , 
    653 N.W.2d 266
     (citations and
    quotations omitted)).                  Embedded within our commitment to stare
    decisis is our recognition that "reliance interests are real,
    prior    generations         of    judges       did    their     job    with    wisdom,       and
    efficiency in dispute resolution is important."                               Daniel R. Suhr
    &    Kevin    LeRoy,    The       Past    and    the     Present:       Stare       Decisis   in
    Wisconsin Law, 
    102 Marq. L. Rev. 839
    , 859 (2019).                                   It is also
    conducive to what others legitimately expect of their judicial
    servants:         "Litigants        and    the       public    at    large     need    to   know
    courts    function      as    neutral          decision       makers,    delivering         equal
    justice under law."               
    Id.
         All of this explains why we must be
    "'respectful of the doctrine of stare decisis.'"                               Chief Justice
    Roggensack's        concurrence/dissent,               ¶66     (quoting       Roberson,       
    389 Wis. 2d 190
    , ¶49).
    ¶202 But we mustn't let this principle capture us, for it
    contains dangers of its own.                     To err is human, and judges are
    nothing      if   not   human——especially              when    the     mellifluousness         of
    "your honor" makes the humility necessary to recognize mistakes
    harder to maintain.               See generally Marah Stith McLeod, A Humble
    Justice, The Yale L.J. Forum (Aug. 2, 2017).                            And the potential
    for mistakes is constantly at hand, because it is tempting for a
    creative court to reach a decision "by extorting from precedents
    something which they do not contain."                          Robert Rantoul, Oration
    in    Scituate      (July    4,        1836)    in    Antonin       Scalia,     A    Matter    of
    Interpretation 39 (1991).                 Once embarked on this path, it is too
    easy    for   the    court        to    "extend       [its]    precedents,          which   were
    themselves the extensions of others, till, by this accommodating
    23
    No.    2019AP1376-OA.dk
    principle,    a    whole     system       of   law   is    built     up    without     the
    authority or interference of the [people]."                      
    Id.
           In this way,
    it is possible for us to "'do more damage to the rule of law by
    obstinately       refusing    to    admit       errors,       thereby      perpetuating
    injustice,     than     by     overturning           an    erroneous        decision.'"
    Roberson, 
    389 Wis. 2d 190
    , ¶49 (quoting Johnson Controls, Inc.
    v. Emp'rs Ins. of Wausau, 
    2003 WI 108
    , ¶100, 
    264 Wis. 2d 60
    , 
    665 N.W.2d 257
    ).
    ¶203 We risk this doctrine becoming a mechanism for error-
    perpetuation if we don't respect its purpose:                      To remind us that
    those who came before were diligent and capable in their work,
    and that in doubtful matters it is best to leave settled things
    settled   unless      there    is     a    clear     and    present        need   to    do
    otherwise.
    In the matter of reforming things, as distinct from
    deforming them, there is one plain and simple
    principle; a principle which will probably be called a
    paradox.    There exists in such a case a certain
    institution or law; let us say, for the sake of
    simplicity, a fence or gate erected across a road.
    The more modern type of reformer goes [happily] up to
    it and says, "I don't see the use of this; let us
    clear it away." To which the more intelligent type of
    reformer will do well to answer:    "If you don't see
    the use of it, I certainly won't let you clear it
    away.   Go away and think.    Then, when you can come
    back and tell me that you do see the use of it, I may
    allow you to destroy it.
    G.K. Chesterton, The Thing:                Why I am Catholic 27 (Dodd, Mead
    and Co., Inc. 1930).
    ¶204 Most of the members of this court would turn this
    prudential    lesson    into    a     permanent       fence    that       would   deprive
    Chesterton's reformer of the ability to bring change even after
    24
    No.   2019AP1376-OA.dk
    he had gained the necessary wisdom.                   To fortify this fence, the
    Chief Justice turns to Justice Scalia, who once said:
    "In [originalism's] undiluted form, at least, it is
    medicine that seems too strong to swallow. Thus,
    almost every originalist would adulterate it with the
    doctrine stare decisis——so that Marbury v. Madison
    would stand even if [a prominent legal scholar] should
    demonstrate unassailably that it got the meaning of
    the Constitution wrong."
    Chief      Justice    Roggensack's      concurrence/dissent,               ¶67    (quoting
    Antonin       Scalia,           Originalism:         The       Lesser           Evil,     57
    U. Cinn. L. Rev. 849, 861 (1989) (alteration in original)).                              But
    if   the    Chief     Justice     believes     Justice       Scalia       thought       stare
    decisis should unalterably privilege precedent over text, she is
    mistaken.          Both    Chesterton        and     Justice     Scalia         were    both
    consciously addressing something that could be described as a
    paradox, and this quote captures only one of its sides.                                  The
    other is on display in Justice Scalia's many opinions in which
    he sets the doctrine aside in favor of the text.                                   So, for
    example,      he     disregarded     precedent        when     it    was    "wrong       and
    unworkable,"         or   its    rationale     had    no     support       in    "history,
    precedent, or common sense."            See, e.g., Witte v. United States,
    
    515 U.S. 389
    , 406 (1995) (Scalia, J., concurring) ("This is one
    of those areas in which I believe our jurisprudence is not only
    wrong but unworkable as well, and so persist in my refusal to
    give that jurisprudence stare decisis effect."); Dickerson v.
    United States, 
    530 U.S. 428
    , 450, 461-65 (2000) (Scalia, J.,
    dissenting) (urging the Court to disregard                      Miranda v. Arizona,
    25
    No.   2019AP1376-OA.dk
    
    384 U.S. 436
     (1966), because its underlying rationale had no
    support in "history, precedent, or common sense.").8
    ¶205 So when precedent unavoidably collides with the law——
    that is, when it is wrong and its rationale has no support in
    history,     precedent,   or    common    sense——there     must   be   no    doubt
    about which will prevail.         I agree with Justice Clarence Thomas,
    who   said    that    "[w]hen   faced     with     a   demonstrably    erroneous
    precedent, my rule is simple:             We should not follow it.            This
    view of stare decisis follows directly from the Constitution's
    supremacy     over     other    sources       of   law——including      our     own
    precedents."         Gamble v. United States, 
    139 S. Ct. 1960
    , 1984
    (2019) (Thomas, J., concurring).              It also follows from the fact
    that no amount of judicial error can change the constitution,
    for "[t]he meaning of the constitutional provision having been
    once firmly established as of the time of its adoption, such
    meaning continues forever, unless it is changed or modified by
    8The Chief Justice isn't quite as wed to stare decisis as
    her opinion would seem to suggest. For an abbreviated sample of
    cases in which she wrote an opinion overturning one or more
    precedents, see State v. Roberson, 
    2019 WI 102
    , 
    389 Wis. 2d 190
    ,
    
    935 N.W.2d 813
    , abrogating State v. Dubose, 
    2005 WI 126
    , 
    285 Wis. 2d 143
    , 
    699 N.W.2d 582
    ; Koschkee v. Taylor, 
    2019 WI 76
    , ¶1,
    
    387 Wis. 2d 552
    , 
    929 N.W.2d 600
    , overruling Coyne v. Walker,
    
    2016 WI 38
    , 
    368 Wis. 2d 444
    , 
    879 N.W.2d 520
    ; Megal v. Green Bay
    Area Visitor & Convention Bureau, Inc., 
    2004 WI 98
    , 
    274 Wis. 2d 162
    , 
    682 N.W.2d 857
    , abrogating Balas v. St. Sebastian's
    Congregation, 
    66 Wis.2d 421
    , 
    225 N.W.2d 428
     (1975) and Lealiou
    v. Quatsoe, 
    15 Wis. 2d 128
    , 
    112 N.W.2d 193
     (1961); State v.
    Ferguson,   
    2009 WI 50
    ,   
    317 Wis. 2d 586
    ,  
    767 N.W.2d 187
    ,
    overruling State v. Mikkelson, 
    2002 WI App 152
    , 
    256 Wis. 2d 132
    ,
    
    647 N.W.2d 421
    ; State v. Sykes, 
    2005 WI 48
    , 
    279 Wis. 2d 742
    , 
    695 N.W.2d 277
    , overruling State v. Hart, 
    2001 WI App 283
    , 
    249 Wis. 2d 329
    , 
    639 N.W.2d 213
    .
    26
    No.    2019AP1376-OA.dk
    the Constitution."            State v. Schinz, 
    194 Wis. 397
    ,                    403, 
    216 N.W. 509
     (1927).
    ¶206 Justice Thomas's formulation also respects the fact
    that the judiciary's authority to decide cases is dependent upon
    an oath in which we swear to uphold the constitution——an oath
    that makes no reference to our precedents.                         "[T]he Constitution
    does not mandate that judicial officers swear to uphold judicial
    precedents.      And the Court has long recognized the supremacy of
    the    Constitution          with     respect     to        executive        action    and
    'legislative act[s] repugnant to' it."                       Gamble, 139 S. Ct. at
    1985   (Thomas,    J.,       concurring)      (quoted       source    omitted;     second
    alteration in original)); see also Mayo v. Wis. Injured Patients
    &   Families    Comp.    Fund,      
    2018 WI 78
    ,      ¶91,    
    383 Wis. 2d 1
    ,   
    914 N.W.2d 678
         (Rebecca       Grassl    Bradley,       J.,    concurring)        ("'[T]he
    Constitution      is    to    be    considered        in    court     as    a   paramount
    law[.]'" (quoting Marbury v. Madison, 5 U.S. (1 Cranch) 137, 178
    (1803)).      This supreme law is the very source of the authority
    we exercise.      If we used it in a manner repugnant to its source,
    we would break faith with those who are the stewards of the
    document from which that authority arises.                         This we must avoid
    at all cost, even should it mean abandoning our wrongly decided
    cases.     We have been equal to the task when called upon to do so
    before, and we must not shrink from it now.
    B.    Of the provenance and operation of "topicality"
    ¶207 Today's decision expressly carries forward our partial
    veto   jurisprudence,         along    with     all    of    its    errors,     with   the
    unremarkable consequence that, when we finished our work, pieces
    27
    No.    2019AP1376-OA.dk
    of the legislative mechanism were still in the wrong place.                             The
    Chief Justice says "[h]aving broken no new ground, I employ our
    decisions and continue the constitutional analysis of 'part' in
    the     four      vetoes    that       were     challenged."             Chief     Justice
    Roggensack's concurrence/dissent, ¶90.9                      The undisturbed ground
    on    which       the    Chief     Justice         builds    her   analysis       is    the
    germaneness test we adopted in Wisconsin Senate:
    [F]or the first time in this case [we] give explicit
    judicial recognition to[] the long-standing practical
    and administrative interpretation or modus vivendi
    between   governors   and   legislatures,   that   the
    consequences of any partial veto must be a law that is
    germane to the topic or subject matter of the vetoed
    provisions.
    Wis. Senate, 
    144 Wis. 2d at 437
    .                    The Chief Justice's statement
    of the rule is almost identical:                     "A veto that does not alter
    legislative control of the topic or subject matter of enrolled
    bills       has   been   referred       to    as     'germane.'"         Chief     Justice
    Roggensack's concurrence/dissent, ¶91.                       Whether we call this a
    "germaneness" test (as we did in Wisconsin Senate) or a "topic
    or    subject     matter"       test   (as    the    Chief    Justice     does)    it   has
    nothing to do with the constitution, as the Wisconsin Senate
    quote makes clear.              It is, instead, merely descriptive of how
    the        executive      and     legislative          branches     have         conducted
    themselves.        As I will explain below, while this may helpfully
    The "continu[ing] constitutional analysis of 'part[,]'"
    9
    unfortunately, did not extend beyond reciting the partial veto
    language and noting that "part" is something less than the
    whole. Neither the Chief Justice nor Justice Ann Walsh Bradley
    mention any of the constitutional provisions that must be
    ignored to operationalize our historical understanding of
    "part."
    28
    No.    2019AP1376-OA.dk
    guide us to a starting point for our analysis, it can never
    authoritatively establish what the judiciary must consider to be
    constitutionally      orthodox.    Consequently,     our     analysis   ended
    where we should have just been starting, which means we are no
    closer to a constitutional understanding of our subject than we
    were in Wisconsin Senate.          I'll say a brief word about the
    inadequacy of the topicality test first, and then address why we
    shouldn't be in the business of blessing the other branches'
    modi vivendi, as Wisconsin Senate says.
    1.     Why "topicality" is an inadequate rule
    ¶208 The Chief Justice says a partial veto is appropriate
    so long as it does "not alter the topic or subject matter of the
    'whole' bill before the veto . . . .          [S]uch a veto does not
    alter the stated legislative idea that initiated the enrolled
    bill."      Chief    Justice   Roggensack's   concurrence/dissent,        ¶11
    (footnote omitted).        It then repeats the proposition at greater
    length, but without any additional explanatory power:
    When the part approved by the governor does not alter
    the topic or subject matter of the whole bill
    presented to him for signature, the part approved
    maintains the legislature's choice of topic or subject
    matter that underlies     the   "whole"  bill. Stated
    otherwise, when the legislative topic or subject
    matter is maintained, the "part" approved and the
    "part" that was not approved remain portions of the
    same "whole" bill, consistent with the constitutional
    text of § 10(1)(b).
    Id., ¶91.
    ¶209 The problem with the topicality rule is that it does
    nothing to repatriate the law-authoring piece of the legislative
    mechanism    to      the   legislature.       From    a      constitutional
    29
    No.   2019AP1376-OA.dk
    perspective,         it    really    doesn't         matter       whether             the    remaining
    parts of the bill speak to the same topic or subject as the bill
    passed     by    the       legislature.              It     matters          whether         they     are
    different from what the legislature passed.                                       The legislature
    does not pass a topic on which the governor may riff, it passes
    one or more proposed laws that he may accept or reject.                                        And, as
    the Chief Justice's opinion very capably explains, id., ¶29, we
    understand that the partial veto power arose in response to the
    legislature's practice of bundling several proposed laws into
    one appropriations bill, and that its telos was to give the
    governor the option of severally treating each of the proposed
    laws.     But a bundle of proposed laws is not an invitation to
    bebop.          The        topicality       rule           may        keep        the       governor's
    improvisations            attached    to   the       neighborhood                of    the    original
    bill,    but    it     still     allows     him       to     change             the   legislatively
    proposed    law       into     something        on    which       the           legislature         never
    voted.     So the topicality test still leaves law-authoring power
    where it does not belong.
    2.    Why we cannot accede to the other branches' modus vivendi
    ¶210 Not only is the topicality rule insufficient to put
    the pieces of the legislative mechanism back where they belong,
    the     rationale         on   which       it    rests           is        at     odds       with     our
    responsibility to ensure the branches of government don't barter
    their powers.          Part of the undisturbed ground on which the Chief
    Justice    bases       her     analysis     is       the    executive             and    legislative
    branches'       "historical         practice,"            which       we    said       in    Wisconsin
    Senate was a "modus vivendi" that had "achieved the force of
    30
    No.   2019AP1376-OA.dk
    law."     Wis. Senate, 
    144 Wis. 2d at 453
    .                    But when it comes to
    the allocation of powers amongst the branches, there is no force
    of law capable of reallocating them, save only a constitutional
    amendment.
    ¶211 I     have     addressed      elsewhere       the    nature         and        rough
    contours of how the constitution allocates power amongst the
    branches of government, so I won't belabor them here.                                       See,
    e.g.,     Wis.    Legislature,        
    391 Wis. 2d 497
    ,        ¶92    (Kelly,           J.,
    concurring) ("Powers constitutionally vested in the legislature
    include the powers: "'to declare whether or not there shall be a
    law; to determine the general purpose or policy to be achieved
    by the law; [and] to fix the limits within which the law shall
    operate.'" quoting Schmidt v. Dep't of Res. Dev., 
    39 Wis. 2d 46
    ,
    59, 
    158 N.W.2d 306
     (1968) (alterations in original)); State ex
    rel. Wisconsin Dev. Auth. v. Dammann, 
    228 Wis. 147
    , 159, 
    277 N.W. 278
    , on reh'g, 
    228 Wis. 147
    , 
    280 N.W. 698
     (1938) ("It is
    fundamental        that      under        our     constitutional           system            the
    governmental       power     to    execute       the   laws     is    vested          in     the
    executive     department      of    the     state[.]");       and    Gabler      v.        Crime
    Victims    Rights     Bd.,     
    2017 WI 67
    ,      ¶37,    
    376 Wis. 2d 147
    ,            
    897 N.W.2d 384
     ("No aspect of the judicial power is more fundamental
    than    the      judiciary's       exclusive       responsibility          to     exercise
    judgment in cases and controversies arising under the law.").
    ¶212 The piece of the doctrine that bears some emphasis in
    this case is that the location of the boundaries between the
    branches is a structural limitation that is beyond the branches'
    power to move, no matter the length of their practice to the
    31
    No.   2019AP1376-OA.dk
    contrary.       Even if two coordinate branches of government should
    agree that the boundary might lie more comfortably elsewhere,
    they     are    powerless          to    affect      its    actual     location.           The
    importance of constitutional limitations, Chief Justice Marshall
    once said, is that they compel restraint when restraint is not
    desired:        "To       what     purpose    are    powers    limited,      and    to    what
    purpose is that limitation committed to writing, if these limits
    may, at any time, be passed by those intended to be restrained?"
    Marbury, 5 U.S. (1 Cranch) at 176.                     This forbids the voluntary
    transfer of core powers to another branch just as much as it
    protects       one        branch      from    encroachment       by     another.          "It
    is . . . fundamental and undeniable that no one of the three
    branches       of    government         can   effectively      delegate       any    of   the
    powers     which          peculiarly      and     intrinsically        belong       to    that
    branch."       Rules of Court Case, 
    204 Wis. 501
    , 503, 
    236 N.W. 717
    (1931); see also 
    id.
     (stating that "'any attempt to abdicate [a
    core power] in any particular field, though valid in form, must,
    necessarily, be held void'" (quoting State ex rel. Mueller v.
    Thompson, 
    149 Wis. 488
    , 491, 
    137 N.W. 20
     (1912))).                                  Even the
    abandonment         of     a     branch's     own    authority       cannot     justify      a
    coordinate branch taking it up and using it as its own.                              "'As to
    these    areas       of    authority, . . . any            exercise    of   authority       by
    another branch of government is unconstitutional.'"                           Gabler, 
    376 Wis. 2d 147
    , ¶31 (quoting State ex rel. Fiedler v. Wis. Senate,
    
    155 Wis. 2d 94
    ,            100,     
    454 N.W.2d 770
           (1990)       (ellipses       in
    original)).
    32
    No.    2019AP1376-OA.dk
    ¶213 The operative principle here is not that the branches
    should      not     delegate    their    core     authority,       it    is    that       they
    cannot.          This principle is a matter of power, not of prudence:
    the     constitution's         progenitors        did    not     grant     the      various
    branches         permission    to   shuffle     their     distinct       powers     amongst
    themselves.         Justice Neil Gorsuch, commenting on this principle
    in    the    federal     context,       consulted       John    Locke    ("one       of    the
    thinkers who most influenced the framers' understanding of the
    separation of powers") for its animating rationale:
    "The legislative cannot transfer the power of making
    laws to any other hands; for it being but a delegated
    power from the people, they who have it cannot pass it
    over to others. The people alone can appoint the form
    of the commonwealth, which is by constituting the
    legislative, and appointing in whose hands that shall
    be.   And when the people have said we will submit to
    rules, and be governed by laws made by such men, and
    in such forms, nobody else can say other men shall
    make laws for them; nor can the people be bound by any
    laws but such as are enacted by those whom they have
    chosen and authorised to make laws for them."
    Gundy       v.     United     States,     
    139 S. Ct. 2116
    ,          2133–34      (2019)
    (Gorsuch,         J.,   dissenting)      (quoting        John    Locke,       The    Second
    Treatise of Civil Government and a Letter Concerning Toleration
    § 41, p. 71 (1947)).
    ¶214 It is for that reason that the several branches of
    government         cannot     alienate    their    core    powers,       even       if    they
    consciously intend that end.              Not because it would be unwise, or
    imprudent, but because those who created them gave them no power
    to do so.          Therefore, prohibiting the legislature and executive
    from swapping their powers "isn't about protecting institutional
    prerogatives or governmental turf."                     Gundy, 
    139 S. Ct. at
    2135
    33
    No.    2019AP1376-OA.dk
    (Gorsuch, J., dissenting).                Instead, "[i]t's about respecting
    the people's sovereign choice to vest the legislative power in
    [the    legislature]          alone.      And           it's    about       safeguarding       a
    structure designed to protect their liberties, minority rights,
    fair notice, and the rule of law."                      
    Id.
        In the constellation of
    constitutional         doctrines,      this    serves          as   one     of    the    central
    organizing principles.            Without it, our constitution would be an
    incomprehensible jumble:               "If [the Legislature] could pass off
    its legislative power to the executive branch, the '[v]esting
    [c]lauses, and indeed the entire structure of the Constitution,'
    would 'make no sense.'"             
    Id. at 2134-35
     (quoted source omitted;
    second and third alterations in original).
    ¶215 But        just    because     the          legislative             and    executive
    branches shouldn't pass their powers around doesn't mean they
    won't    sometimes      try.      Indeed,          Wisconsin        Senate's          recognition
    that the legislative and executive branches have arrived at a
    "modus vivendi" in the allocation of their powers proves not
    only    that    they    are    willing    to       try,       but   that        they   sometimes
    succeed.       This would not necessarily come as a surprise to the
    constitution's         authors.        They        structured        it     to    prevent    the
    shifting of boundaries through its internal system of checks and
    balances, and by arraying ambition against ambition, yet they
    knew    these    structures       wouldn't         be    sufficient         to    prevent    all
    attempted incursions.             "The framers knew . . . that the job of
    keeping the legislative power confined to the legislative branch
    couldn't be trusted to self-policing by Congress; often enough,
    legislators will face rational incentives to pass problems to
    34
    No.   2019AP1376-OA.dk
    the   executive   branch."     
    Id. at 2135
    .      When   an     attempted
    incursion   comes   before   us,   we     do    not    have    the     luxury   of
    shrugging off our duty to repulse it.
    [T]he Constitution does not permit judges to look the
    other way; we must call foul when the constitutional
    lines are crossed.    Indeed, the framers afforded us
    independence from the political branches in large part
    to encourage exactly this kind of "fortitude . . . to
    do   [our]   duty  as   faithful   guardians  of   the
    Constitution."
    
    Id.
     (quoting The Federalist No. 78, at 468-469 (C. Rossiter ed.
    1961) (ellipsis in original)).
    ¶216 The Chief Justice may very well be right that the
    legislative   and    executive       branches      have       fallen     into    a
    comfortable partial veto routine in which the legislature allows
    the governor to unilaterally create law so long as it's on the
    same topic as the bill he is reviewing.            But basing our analysis
    on that practice is quite literally the definition of "begging
    the question."      We should not base our analysis on a logical
    35
    No.       2019AP1376-OA.dk
    fallacy,    especially      when       the    assumed      conclusion           is   one    our
    constitution so thoroughly rejects.10
    IV.    WHAT WE SHOULD DO
    ¶217 I     propose     that         we      respect       the      constitution's
    structural      limitations       on    what      it    means     for     a     bill   to    be
    approved    "in   part."          As   I     explained         above,     the    law-making
    mechanism described by our constitution contemplates that the
    most elemental part of a bill can be no less than an idea——that
    is to say, a proposal for a complete, entire, and workable law.
    This, of course, makes perfect sense in light of the partial
    veto    power's    purpose,       which      the       Chief    Justice        persuasively
    described as answering the legislative practice of bundling many
    proposed laws into one bill.11                    Therefore, because the partial
    The Chief Justice finds this constitutional analysis
    10
    faulty because it "does not account for the text of the
    Wisconsin Constitution," and it "ignore[s] that Wisconsinites
    are free to assign powers traditional to one branch of
    government to another branch by constitutional amendment."
    Chief Justice Roggensack's concurrence/dissent, ¶87.   Well, the
    people of Wisconsin certainly are free to reassign the
    traditional powers of one branch to another.     But whether the
    people did so by making the governor into a one-man legislature
    requires accounting for all of the constitutional provisions
    relevant to the legislative process.    Might I remind the Chief
    Justice that her conclusion that the people of Wisconsin did
    this novel and radical thing is based on a single word?      And
    that her opinion did not even refer to the constitutional
    provisions that define the legislative process even once?    The
    word "part" simply isn't powerful enough to countermand all the
    constitutional text necessary to make the Chief Justice's
    understanding of the partial veto viable.
    We have understood this as the rationale for the partial
    11
    veto from the very beginning:
    36
    No.   2019AP1376-OA.dk
    veto power cannot act against any division less than the most
    elemental part, the governor must take the bill as he finds it:
    as a collection of proposed laws.            So the smallest part of a
    bill    against   which   the   partial   veto   may   act   is   one   of   the
    proposed laws in that collection.          Consequently, the applicable
    rule guiding the application of the partial veto is as follows:
    After exercising the partial veto, the remaining part of the
    bill must not only be a "complete, entire, and workable law," it
    must also be a law on which the legislature actually voted; and
    the part of the bill not approved must be one of the proposed
    laws in the bill's collection.12           Nothing less than this will
    [T]he Legislature may, if it pleases, unite as many
    subjects in one bill as it chooses. Therefore, in
    order to check or prevent the evil consequences of
    improper joinder, so far, at least, as appropriation
    bills are concerned, it may well have been deemed
    necessary, in the interest of good government, to
    confer upon the Governor, as was done by the amendment
    in 1930 of section 10, art. 5, Wis. Const., the right
    to pass independently on every separable piece of
    legislation in an appropriation bill.
    State ex rel. Wis. Tel. Co. v. Henry, 
    218 Wis. 302
    , 315,
    
    260 N.W. 486
     (1935).
    This, of course, is very close to the rule stated in
    12
    Henry.   Indeed, the rule, in the main, simply makes Henry's
    unstated assumption explicit in that it requires the remaining
    parts of the bill to contain ideas on which the legislature
    actually voted.
    37
    No.   2019AP1376-OA.dk
    restore the pieces of the legislative machinery to their proper
    places.13
    13Justice Ann Walsh Bradley says we should not return to
    our constitution's structural limitations on the partial veto
    because it "embraces a test neither advanced by any party nor
    ever applied in any case."        Justice Ann Walsh Bradley's
    concurrence/dissent, ¶113. I disagree, of course. But I think
    a few words on the nature of this objection would be
    appropriate, starting with the latter clause.         It is an
    embarrassment, not a source of authority, that our court has
    never honored the constitution's limitations on the partial
    veto. Perpetuating an embarrassment is not a judicial doctrine
    to which I subscribe.      Nor is the novelty of applying the
    constitution's terms to this case an argument against doing so.
    There is a first time for everything that happens——including the
    "topicality/germaneness" test, which had never been applied in
    any case in Wisconsin's history until the day it was.    Because
    everything has its genesis, a proscription against doing
    something for the first time——if we were to take it seriously——
    would be a condemnation of everything that has ever been done.
    That is not a workable standard.
    But   even  more   interesting   to  me,   because   of  its
    curiousness, is the objection that we should not interpret the
    law in a manner not advanced by one of the parties.          That
    sentiment compasses an understanding of the court that is
    entirely foreign to me. The work of the judiciary is not some
    glorified form of "baseball arbitration" in which we are
    constrained to choosing one of the proposals offered by the
    competing parties. The attorneys who appear before us are there
    to help us discover what the law requires, not to control us.
    It is our job, not theirs, to "say what the law is." Marbury v.
    Madison, 5 U.S. (1 Cranch) 137, 177 (1803).         If we should
    discover, in the course of our research, that the parties both
    mistook it for something other than it is, it would be an
    abdication of our sworn duty to simply adopt whichever argument
    seemed   closest  to   what   the   law  actually   says.     Our
    responsibility is to determine for ourselves——in every single
    case, without exception——what the law requires. And there is no
    one to whom we can delegate that responsibility.      So even if
    neither of the parties' arguments were correct, our duty would
    remain the same——to discover and say what the law says, not what
    a party says.
    38
    No.   2019AP1376-OA.dk
    ¶218 Justice       Ann      Walsh     Bradley      is   concerned       that    my
    analysis    would     collapse      our   constitution's      partial      veto     into
    something indistinguishable from other states' line-item vetoes.
    "[T]here is a difference," she says, "between a 'partial' and an
    'item' veto . . . [;] [Justice Kelly's opinion] does not account
    for the difference and would, as a practical matter, result in
    an 'item' veto in spite of Wisconsin's unique constitutional
    language."       Justice      Ann    Walsh     Bradley's     concurrence/dissent,
    ¶150.     I do not think that is so.             There is no mandatory, single
    definition for what a "line-item veto" might comprise, so its
    content    and   operation     could      vary   from    state    to    state.      But
    generally speaking, line-item vetoes operate only on the fiscal
    elements    of   an    appropriation         bill.      Corpus     Juris     Secundum
    contains the following description of such a veto:
    The purposes of an appropriations item or line-item
    veto are to give the executive, who is elected
    statewide rather than from a particular district, the
    power to achieve fiscal constraint and to advance
    statewide rather than parochial fiscal interests by
    excising unneeded "pork barrel" programs or projects
    from an appropriations bill so as to restrain public
    expenditures and to permit the governor to disentangle
    issues so they will be considered on their individual
    merits . . . .
    . . . Specific   allocations   within   a   general
    appropriation are subject to separate veto, either
    leaving the general appropriation intact in its full
    Happily, Justice Ann Walsh Bradley's concern about whether
    I grounded my analysis in a party's argument is unwarranted
    here.    The petitioner's brief and the legislature's amicus
    brief, in combination, either directly or obliquely advance most
    of the analysis in my opinion. And if the concern is that part
    of the analysis appears in an amicus brief rather than a party
    brief, then I wonder why we allow amici at all.
    39
    No.    2019AP1376-OA.dk
    and original amount or reduced by a sum less than the
    aggregate of the specific items vetoed.
    82 C.J.S. Statutes § 68 (2020) (footnotes omitted).
    ¶219 Currently,            43        states       have     some         form   of     the
    item/partial veto.           Most limit the vetoes to the fiscal elements
    of an appropriation bill.                  So Wisconsin's partial veto would not
    be the same as a line-item veto inasmuch as ours could be used
    as    against     any       of   the       legislative         ideas    bundled      into    an
    appropriations         bill,     even       if    the    vetoed        part    contained    no
    appropriation.
    V.    APPLICATION
    ¶220 2019 Assembly Bill 56 (which became 2019 Wis. Act 9,
    as amended by the governor's "veto") contained a multitude of
    proposed laws, amongst which were a school bus modernization
    fund,     a    local    roads      improvement           fund,     a    modified      vehicle
    registration fee schedule, and a tax on vapor products.                               Here is
    how   a   constitutionally-grounded                   partial     veto        analysis    would
    address the governor's actions.
    A.    School Bus Modernization Fund
    ¶221 The first partial veto at issue in this case changed a
    school bus modernization fund into an alternative fuel fund.
    Section 55c established a grant for the replacement of school
    buses.        And § 9101(2i) identified the monies to be used to fund
    the replacement program.                The governor's partial "veto" amended
    § 55c as follows:
    16.047(4s) of the statutes is created to read: 16.047
    (4s) SCHOOL BUS REPLACEMENT GRANTS. (a) In this
    subsection: 1. "School board" has the meaning given in
    s. 115.001(7).2. "School bus" has the meaning given in
    s. 121.51(4).(b) The department shall establish a
    40
    No.    2019AP1376-OA.dk
    program to award grants of settlement funds from the
    appropriation under s. 20.855(4)(h) to school boards
    for the replacement of school buses owned and operated
    by the school boards with school buses that are energy
    efficient, including school buses that use alternative
    fuels. Any school board may apply for a grant under
    the program. (c) As a condition of receiving a grant
    under this subsection, the school board shall provide
    matching funds equal to the amount of the grant award.
    (d) A school board may use settlement funds awarded
    under this subsection only for the payment of costs
    incurred by the school board to replace school buses
    in accordance with the settlement guidelines.
    The governor entirely struck § 9101(2i):
    (2i) VOLKSWAGEN SETTLEMENT FUNDS. Of the settlement
    funds in s. 20.855(4)(h), during the 2019–21 fiscal
    biennium, the department of administration shall
    allocate $3,000,000 for grants under s. 16.047 (4s)
    for the payment of school buses.
    The surviving language reads:             "16.047(4s) of the statutes is
    created to read:        16.047 (4s) GRANTS.                The department shall
    establish a program to award grants of settlement funds from the
    appropriation under s. 20.855(4)(h) for alternative fuels."
    ¶222 The Chief Justice says the result is not on the same
    topic as the original bill.               But "topicality" is an elastic
    measuring tape, as even the Chief Justice recognizes.                               Chief
    Justice    Roggensack's      concurrence/dissent,              ¶91    ("Clearly,       the
    evaluation of 'part' and 'whole' in § 10(1)(b) depends on how
    broadly you define the topic or subject matter.").                        Both before
    and   after   the   veto,   this   part       of   the    bill       created    a    grant
    program.      And the funding would still come from the Volkswagen
    dispute settlement.         The Chief Justice says the "topic" of the
    provision     was   replacement      of       buses,      not        limiting       carbon
    emissions.      Actually,     it   was    both.          The    legislature         wanted
    41
    No.    2019AP1376-OA.dk
    school boards to replace current school buses not with just any
    buses,    but      "with    school      buses       that    are     energy          efficient,
    including school buses that use alternative fuels."                               So it seems
    that     under      the     Chief       Justice's         "topicality"            test,     the
    constitutionality of a partial veto depends on which topic we
    figure is more important.
    ¶223 The     resolution       called         for    by    the    constitution         is
    considerably        more    straightforward.                Here,       the       legislature
    bundled the creation of a school bus replacement fund into a
    bill    with    many     other   proposed      laws.        As    relevant          here,   the
    school bus replacement fund is the proposed law, the legislative
    idea.    The governor could approve that part of the bill or he
    could reject it.           What he may not do is turn it into something
    other than what passed the legislature.                     This partial "veto" was
    inappropriate because it violated the origination clause, the
    amendment clause, and the legislative passage clause.
    B.     The Local Road Improvement Fund
    ¶224 In      another      part    of     2019       Assembly          Bill    56,    the
    legislature proposed the creation of a local road improvement
    fund.        The   governor      amended      the    proposed       law      by     using   his
    partial "veto" on §§ 126, 184s, and 1085m:
         Section 126:      "(fc) Local roads improvement
    discretionary supplement . . . 90,000,000 [the
    governor replaced it with 75,000,000]."
         Section 184s:  "20.395(2)(fc) of the statutes is
    created to read: 20.395(2) (fc) Local roads
    improvement discretionary supplement. From the
    general fund, as a continuing appropriation, the
    amounts in the schedule for the local roads
    42
    No.   2019AP1376-OA.dk
    improvement   discretionary   supplemental                   grant
    program under s. 86.31 (3s)."
        Section 1085m:    "86.31 (3s) of the statutes is
    created   to   read:   86.31   (3s)  DISCRETIONARY
    SUPPLEMENT GRANTS. (a) Funds provided under s.
    20.395 (2) (fc) shall be distributed under this
    subsection as discretionary grants to reimburse
    political subdivisions for improvements. The
    department     shall     solicit    and     provide
    discretionary grants under this subsection until
    all funds appropriated under s. 20.395 (2) (fc)
    have been expended. (b) 1. From the appropriation
    under s. 20.395 (2) (fc), the department shall
    allocate $32,003,200 in fiscal year 2019–20, to
    fund county truck highway improvements. 2. From
    the appropriation under s. 20.395 (2) (fc), the
    department shall allocate $35,149,400 in fiscal
    year 2019–20, to fund town road improvements. 3.
    From the appropriation under s. 20.395 (2) (fc),
    the department shall allocate $22,847,000 in
    fiscal year 2019–20, to fund municipal street
    improvement projects. (c) Notwithstanding sub.
    (4), a political subdivision may apply to the
    department     under     this    subsection     for
    reimbursement of not more than 90 percent of
    eligible costs of an improvement."
    The surviving language reads:               "20.395(2)(fc) of the statutes is
    created       to   read:    20.395(2)   (fc)       Local      supplement.    From    the
    general fund, as a continuing appropriation, the amounts in the
    schedule for local grant."
    ¶225 These, of course, were amendments just as much as the
    partial       "veto"   of    the   school    bus       modernization    fund   was    an
    amendment, not a veto.             The result of these amendments is that
    the new idea introduced by the amendment passed into law without
    the   legislature           ever   voting        for    it.      This     "veto"     was
    inappropriate for the same reasons the partial "veto" of the
    school bus modernization fund was inappropriate.
    43
    No.   2019AP1376-OA.dk
    C.     Vehicle Fee Schedule
    ¶226 Section         1988b     of    the    bill    would       have    made     the
    registration fee for four truck weight classes identical.                            The
    governor   amended   this         section      with    his   partial       "veto"    as
    follows:
    341.25(2)(a) to (cm) of the statutes are amended to
    read: 341.25 (2)(a) Not more than 4,500 $ 75.00 100.00
    (b) Not more than 6,000 . . . . . . . . . . 84.00
    100.00 (c) Not more than 8,000 . . . . . . . . . .
    106.00 100.00 (cm) Not more than 10,000 . . . . . . .
    . . . 155.00 100.00
    Prior to the "veto," all registration fees were $100, but what
    remained   afterwards       was     a    graduated      schedule        according    to
    vehicle size.     This might be good policy, but it's not a veto.
    It's an amendment, and it fails for the same reason as the
    others.
    D.     Vapor Products Tax
    ¶227 Section     1754        addresses      taxation     of    vapor     products.
    The governor amended it with his partial "veto" as follows:
    139.75 (14) of the statutes is created to read: 139.75
    (14) "Vapor product" means a noncombustible product
    that produces vapor or aerosol for inhalation from the
    application of a heating element to a liquid or other
    substance that is depleted as the product is used,
    regardless of whether the liquid or other substance
    contains nicotine.
    The surviving language reads:               "139.75 (14) of the statutes is
    created    to    read:     139.75        (14)    'Vapor      product'        means     a
    noncombustible    product         that   produces      vapor       or    aerosol     for
    inhalation from the application of a heating element regardless
    of whether the liquid or other substance contains nicotine."
    44
    No.    2019AP1376-OA.dk
    ¶228 In this part of the bill, the legislature proposed a
    law that would tax "vaping" equipment, but not the liquids used
    in the equipment.             The governor's partial "veto" expanded the
    tax to include the liquids as well, which made it an amendment,
    not a veto.       For anyone even vaguely familiar with our country's
    history and the revolution that brought it into existence, this
    should make you sit up and take notice:                       The governor, all by
    himself,       imposed    a    tax    on     a     product      without     legislative
    approval.       Taxation without representation was once a powerful
    rallying cry.         See Declaration of Independence (U.S. 1776) (One
    of our grievances with the King of England was his habit of
    "imposing taxes on us without our consent[.]"; John Dickinson,
    Letter's From a Farmer in Pennsylvania reprinted in Tracts of
    the    American      Revolution      141    (1763-1776)       (Merrill    Jensen      ed.,
    Hackett Pub. Co. 2003) (1768) ("That it is inseparably essential
    to     the    freedom    of    a   people,        and   the     undoubted     right    of
    Englishmen, that NO TAX be imposed on them, but with their own
    consent, given personally, or by their representatives.").                             As
    with    all    the   other    partial       "vetoes"    in    this   case,    this    one
    violated the origination clause, the amendment clause, and the
    legislative passage clause.                It also violated the unwritten, but
    only    slightly      less    well-known,         "don't   do    revolution-inciting
    things" clause.
    VI.     CONCLUSION
    ¶229 Because a majority of this court does not favor this
    analysis, our partial veto jurisprudence leaves key pieces of
    the legislative machinery in places where they do not belong.
    45
    No.    2019AP1376-OA.dk
    As a direct and unavoidable result, our cases (including this
    one) condone violations of the origination clause, the amendment
    clause, and the legislative passage clause.
    ¶230 The proper role of the partial veto is to separate the
    several      proposed    laws       the      legislature          bundled     into    one
    appropriations     bill.        After     exercising        this    veto     power,   the
    remaining document must comprise one or more "complete, entire,
    and   workable     laws,"      all      of        which    must    have     passed    the
    legislature.     The corollary to this is that the part or parts of
    the bill the governor did not approve must also comprise one or
    more "complete, entire, and workable laws" that had passed the
    legislature.      This symmetry guarantees that the partial veto
    does nothing but unbundle the proposed laws the legislature had
    bundled.14       Because      the    majority         of    this    court     does    not
    accurately     apply    the    legislative           mechanism      the     constitution
    created, I cannot join it.            However, I concur with that part of
    the court's judgment that strikes three of the vetoes at issue
    I would overrule State ex rel. Sundby v. Adamany, 71
    
    14 Wis. 2d 118
    , 
    237 N.W.2d 910
     (1976); State ex rel. Kleczka v.
    Conta, 
    82 Wis. 2d 679
    , 
    264 N.W.2d 539
     (1978); Citizens Util. Bd.
    v. Klauser, 
    194 Wis. 2d 484
    , 
    534 N.W.2d 608
     (1995); Risser v.
    Klauser, 
    207 Wis. 2d 176
    , 
    558 N.W.2d 108
     (1997); and State ex
    rel. Wisconsin Senate v. Thompson, 
    144 Wis. 2d 429
    , N.W.2d 385
    (1988). Each of these decisions depends on the unconstitutional
    transfer of law-making power to the governor through the use of
    a partial veto.
    I would not, however, overrule State ex rel. Wisconsin Tel.
    Co. v. Henry, 
    218 Wis. 302
    , 
    260 N.W. 486
     (1935).      Instead, I
    would modify its holding to make its assumption explicit:    The
    parts of the bill remaining after exercise of the partial veto
    must comprise "a complete, entire, and workable law" that was
    actually voted on by the legislature.
    46
    No.   2019AP1376-OA.dk
    in this case, and respectfully dissent from the court's judgment
    upholding the other one.
    ¶231 I am authorized to state that Justice REBECCA GRASSL
    BRADLEY joins this opinion.
    47
    No.    2019AP1376-OA.bh
    ¶232 BRIAN        HAGEDORN,         J.        (concurring).          In    1930,   the
    people   of    Wisconsin         amended      our     constitution       and    gave   the
    governor      power        to    veto        parts     of   appropriation          bills.
    Nonetheless,     the       constitution           retains   the    basic       structural
    principle that legislating is the job of the legislature.                              The
    question in this case is whether the judiciary will sanction the
    former swallowing the latter.
    ¶233 The         partial      veto       power     grants     the     governor     the
    authority to disapprove appropriations bills in part——a power
    that no doubt allows the governor to alter the legislature's
    global policy objectives.               The partial veto power in this sense
    is quasi-legislative in nature.                      But a bill presented to the
    governor is not sand on a seashore from which a governor can
    construct any sandcastle his ingenuity conceives.                         A bill is not
    merely a collection of words, letters, and numbers that can be
    repurposed; it is a set of legislatively chosen policies.                                A
    partial veto is the power to negate some proposed policies and
    accept others, not the power to unilaterally create new policies
    never passed by the legislature.
    ¶234 While the governor's partial veto power is incredibly
    broad, it should not be read to fundamentally upend the overall
    structure of our government embedded in our constitution.                              The
    constitution's placement of law-creation in the hands of the
    legislature means we cannot permit a practice that turns the
    governor      into     a        one-person         legislature.           Because      the
    constitution contains these substantive limitations, we should
    1
    No.       2019AP1376-OA.bh
    enforce     them,   even    acknowledging        the    potential          difficulty     of
    that project.
    ¶235 In this case, the petitioners challenge four sets of
    vetoes in the state's 2019-21 biennial budget bill.                           I conclude
    that with three of the challenges——the school bus modernization
    fund, the local road improvement fund, and the vapor products
    tax——the     governor's      vetoes    went      beyond      negating        legislative
    policy proposals; they created brand new ones.                             These are in
    excess of the governor's constitutional veto authority.                                  The
    fourth challenge to the vehicle fee schedule vetoes was properly
    within    constitutional         boundaries.       Therefore,          I    respectfully
    concur.
    I.    LEGAL PRINCIPLES
    ¶236 Something         is    amiss     in    our       jurisprudence         when    a
    constitutional provision allowing the governor to strike parts
    of an appropriation bill has, through creativity and judicial
    acquiescence, turned into a license for an enterprising governor
    to   create    brand   new       policies     from      a    proposed        package      of
    statutory words.           This is a bipartisan affair, of course, as
    governors for decades have been working within the Wild West
    framework this court has established.                   But no one conducting a
    reasonable reading of the partial veto provision in its greater
    constitutional context would see it as a fundamental reshaping
    of    our     constitutional          order.            See      Justice          Kelly's
    concurrence/dissent, ¶198.            We are here because this court has
    allowed it to be so.         As one former justice aptly prophesied, "I
    2
    No.   2019AP1376-OA.bh
    fear that the court may now have painted itself into a corner,
    and that a time may come when we regret having done so."                             State
    ex rel. Kleczka v. Conta, 
    82 Wis. 2d 679
    , 724, 
    264 N.W.2d 539
    (1978) (Hansen, J., dissenting).                 For me, that time is now.
    ¶237 So where do we go from here?                  The petitioners candidly
    ask us to start from scratch.              They ask us to overturn or modify
    language    in    every      case     that       we've    ever    decided       on    this
    significant and repeatedly litigated provision.                          That's a big
    ask.    But the petitioners come with the right question:                        What is
    the original public meaning of the constitutional text?                                   Our
    starting    point    in   constitutional           interpretation        must    be       the
    original public meaning of the constitution's language because
    this is the law the people have enacted.                     Attorney Gen. ex rel.
    Bashford    v.    Barstow,    4     Wis. *567,         *757–58   (1855)    (explaining
    that because the people "made this constitution, and adopted it
    as their primary law," constitutional interpretation rests not
    in generic theories of governance, but on the "true intent and
    meaning"    of    the   "authoritative           and     mandatory"     words    of       the
    document itself).         But our analysis is informed by, and gives
    proper deference to, the reasoned decisions of those who have
    come before us.
    A.     Lawmaking in the Wisconsin Constitution
    ¶238 Three types of government power are described in the
    Wisconsin    Constitution,           and     each       power    is     vested       in     a
    corresponding branch of government.                      Gabler v. Crime Victims
    Rights Bd., 
    2017 WI 67
    , ¶11, 
    376 Wis. 2d 147
    , 
    897 N.W.2d 384
    .
    3
    No.     2019AP1376-OA.bh
    The senate and assembly are vested with the power to legislate,
    the governor is vested with the power to execute the laws, and
    the judiciary is vested with the power to decide cases based on
    the law.       Wis. Const. art. IV, § 1; id. art. V, § 1; id. art.
    VII, § 2.
    ¶239 The mechanism for exercising legislative power under
    the constitution is the enactment of laws; the legislature is
    the chief lawmaker.             League of Women Voters of Wis. v. Evers,
    
    2019 WI 75
    ,   ¶35,    
    387 Wis. 2d 511
    ,         
    929 N.W.2d 209
    ;      Justice
    Kelly's concurrence/dissent, ¶175.                  A law begins with a proposed
    bill, which can originate in either house of the legislature.
    Wis.       Const.    art.        IV,    § 17(2),        § 19;      Justice      Kelly's
    concurrence/dissent, ¶176.               Bills may be amended during this
    process,      and   when    a    bill   is   passed      by    both    houses   of   the
    legislature, it is presented to the governor.                         Wis. Const. art.
    IV,        § 19;    id.     art.        V,       § 10(1)(a);      Justice       Kelly's
    concurrence/dissent,            ¶¶176-77.         The   governor       then   has    four
    potential options:          (1) sign the whole bill into law; (2) do
    nothing and allow the bill to become law on its own after six
    days (Sundays excluded); (3) veto the whole bill; or (4) if the
    bill contains an appropriation, sign the bill into law while
    vetoing part of it.1            Wis. Const. art. V, § 10(1)(b), § 10(2)(a),
    § 10(3).
    If rejected in whole or in part, the bill is returned,
    1
    with objections, to the originating house, and that which was
    rejected may nevertheless become law if it garners approval of
    two-thirds of the members of both houses.  Wis. Const. art. V,
    § 10(2).
    4
    No.   2019AP1376-OA.bh
    ¶240 The fourth option, the partial veto, was added to the
    constitution    in   1930.         The   relevant    constitutional         language
    today provides:       "If the governor approves and signs the bill,
    the bill shall become law.           Appropriation bills may be approved
    in whole or in part by the governor, and the part approved shall
    become law."       Id. art. V, § 10(1)(b).2          Appropriation bills are
    required to pay money out of the treasury.                     Id. art. VIII, § 2
    ("No money shall be paid out of the treasury except in pursuance
    of an appropriation by law.").
    ¶241 This framework deserves a few reflections.                      First, the
    constitutional meaning of a "bill" must be rooted in the concept
    of what the legislature is producing when a bill is passed.                         A
    bill presented to the governor is not a potpourri of words,
    letters, and numbers that the governor may do with as he wishes.
    See State ex rel. Wis. Senate v. Thompson, 
    144 Wis. 2d 429
    , 473,
    
    424 N.W.2d 385
     (1988) (Bablitch, J., dissenting).                        As Justice
    Kelly explains, a bill is composed of policy proposals (or as
    Justice    Kelly     calls   them,       ideas).         See     Justice     Kelly's
    concurrence/dissent,       ¶¶175-76,      180.      It   is    the    legislature's
    province   to   exercise     the    legislative     power       to    determine   and
    declare what the policies of the state shall be.                        Wis. Const.
    art. IV, § 1.      And this is done by passing bills composed of its
    policy choices.       Borgnis v. Falk Co., 
    147 Wis. 327
    , 351, 133
    2  The    amendment   as    initially   adopted    provided:
    "Appropriation bills may be approved in whole or in part by the
    governor, and the part approved shall become law, and the part
    objected to shall be returned in the same manner as provided for
    other bills." 1927 S.J. Res. 35.
    5
    No.    2019AP1376-OA.bh
    N.W. 209 (1911) ("When acting within constitutional limitations,
    the Legislature settles and declares the public policy of a
    state . . . .").
    ¶242 Second, the veto power is a bit of an aberration from
    the general distribution of constitutional power.                        That is, the
    power to veto, whether in whole or in part, is legislative in
    nature; it is a participation in lawmaking.                      Edwards v. United
    States,    
    286 U.S. 482
    ,         490-91       (1932)      (characterizing          the
    President's      ability       to     approve         or     disapprove     bills       as
    "legislative" in character); Rateree v. Rockett, 
    852 F.2d 946
    ,
    951 (7th Cir. 1988) ("[T]he President acts legislatively when he
    approves or vetoes bills passed by Congress."); Chief Justice
    Roggensack's concurrence/dissent, ¶84.                     And while a partial veto
    places more quasi-legislative power in the hands of the governor
    than a whole-bill veto, we cannot lose sight of the nature of a
    veto.     A veto is, by definition, the ability to negate, not
    create.    This is the plain meaning of the word "veto."                             Veto,
    Black's   Law    Dictionary         (11th       ed.   2019)     ("A     power   of     one
    governmental branch to prohibit an action by another branch."
    (emphasis added)); The Federalist No. 73 (Hamilton) (describing
    the veto as "the qualified negative of the President upon the
    acts or resolutions of the two houses of the legislature").
    ¶243 Finally, the partial veto power must be read in the
    context   of     the   whole        constitutional          structure     and   design.
    Namely, any policy proposal that becomes law must be a policy
    proposed by the legislature——one that originates as a bill that
    eventually passes both houses of the legislature.                         Wis. Const.
    6
    No.    2019AP1376-OA.bh
    art. IV, § 17(2), § 19; id. art. V, § 10(1)(a).                                 Partial veto or
    not, the legislature is still the constitutional branch charged
    with     making      law,     not        the    governor.               See     Justice          Kelly's
    concurrence/dissent, ¶175.
    ¶244 We        must        hold     all           of     these     lessons          from      the
    constitution together.               A blind focus on the partial veto power
    alone at the expense of the rest of constitutional text is not
    constitutional faithfulness.                        State v. City of Oak Creek, 
    2000 WI 9
    ,    ¶18,       
    232 Wis. 2d 612
    ,             
    605 N.W.2d 526
          (we     discern        the
    meaning of the constitutional text based on the context in which
    it is used).          This means any reading of the partial veto power
    that enables the governor to take the raw materials of a bill
    (words, letters, and numbers) and recast them to create a new
    policy not proposed and passed by the legislature contradicts
    the constitutional design for how a bill becomes a law.                                          And the
    core negating, not creating, concept of a veto must be true if
    the     legislature          is     still           the       branch     authorized          by      the
    constitution to make law and appropriate funds.                                           Wis. Const.
    art.     IV,       § 1,    § 17(2),        § 19;          id.     art.     VIII,          § 2.       The
    legislature must be the primary policymaker, and the governor
    cannot       usurp    that    role        by     creating         new     policies          from     the
    reworked language of enacted bills.
    ¶245 With this broader constitutional framework in view, we
    turn    to     a   brief     overview          of    how       this     court       has    previously
    handled the partial veto power in particular.
    7
    No.    2019AP1376-OA.bh
    B.     The Partial Veto and the Wisconsin Supreme Court
    ¶246 Alfred            North    Whitehead           famously    said    that       Western
    philosophy consists of a series of footnotes to Plato.3                                   In the
    same       way,   this    court's          decisions       interpreting       the    governor's
    partial veto power consist largely of a series of footnotes to
    our first case on that matter, State ex rel. Wisconsin Telephone
    Co. v. Henry, 
    218 Wis. 302
    , 
    260 N.W. 486
     (1935).                                    Handed down
    just five years after the ratification of the 1930 amendment,
    Henry       presented     two       questions:             whether    the     governor     could
    "disapprove           parts    of     an    appropriation       bill    that        are   not   an
    appropriation" and whether he could "disapprove a proviso or
    condition inseparably connected to the appropriation."                                     Id. at
    309.            The    court     engaged       in      a     considered       plain       meaning
    examination of the text and reached several conclusions that
    establish the framework for the partial veto power.
    ¶247 Of        primary       importance,        the    court    reasoned       that      the
    choice of constitutional language——using "part" and not "item"——
    was intentional and must be given meaning.                            Id. at 313-14.4           The
    amendment, the court concluded, was not an item veto, but a part
    veto that authorized gubernatorial disapproval of something less
    than an entire legislative policy proposal.                             Id.         A governor,
    Alfred North Whitehead, Process and Reality 39 (The Free
    3
    Press 1978) (1929).
    The court in Henry surveyed constitutions of other states
    4
    that permitted some form of partial veto.    State ex rel. Wis.
    Tel. Co. v. Henry, 
    218 Wis. 302
    , 310-15, 
    260 N.W. 486
     (1935).
    Noting that many states used "items" or "any item or items or
    part or parts," the court concluded that our constitution's use
    of the word "part" but not the word "item" was significant and
    must be given meaning. Id. at 310-11.
    8
    No.    2019AP1376-OA.bh
    then,    could    veto     non-appropriation                   language       in    appropriation
    bills.     Id.     He could also strike portions of a broader policy
    proposal     that        did     not           constitute         provisos         or     conditions
    inseparably connected to the appropriation.                                  Id.         As long as
    what remained was a complete, entire, and workable law, vetoing
    portions of the proposed law that were not "essential, integral,
    and   interdependent            parts          of    those     which      were     approved"        was
    consistent with the constitution.                             Id. at 314, 317.              Applying
    this to the facts before it, the court concluded that the vetoed
    language declaring the purpose for a new appropriation and the
    proposed     creation           of        a     new       administrative           apparatus        for
    distribution        of     that           appropriation           were       not        provisos    or
    conditions inseparably connected to the remainder.                                       Id. at 317.
    The   governor's         veto    was           therefore      within        his    constitutional
    authority.       Id.
    ¶248 From        this,         we       observe      that      Henry    identified           both
    procedural       and     substantive                limitations      on     the     partial        veto
    power.     Procedurally, what is left must be a complete, entire,
    and workable law.          Id. at 314.                This is obviously correct if the
    part approved is actually to become law as the constitution
    specifies.       Wis. Const. art. V, § 10(1)(b).                            But the court also
    recognized       substantive              limitations,         unsubtly       suggesting           that
    provisos    and     conditions                that    could    not     be    separated       from    a
    policy proposal could not be stricken.                                Henry, 218 Wis. 2d at
    309-10.    The court labelled the veto power coextensive with the
    legislature's power to assemble.                          Id. at 315.         But this is just
    as much a limitation on the power's reach as it is a recognition
    9
    No.      2019AP1376-OA.bh
    of the power's breadth.                    Id. at 315.        The court also discussed
    how severability principles——which include at least some focus
    on   legislative          intent——were            relevant    to    an    inquiry        into   the
    scope of the partial veto power.                        Id. at 314-15.
    ¶249 The petitioners ask us to overturn Henry.                                  They argue
    this court misconstrued the original public meaning from the
    beginning, and that the partial veto was intended to be an item
    veto.       Some evidence, including newspaper stories reflecting the
    sponsor's goals and other public discussion on the proposed 1930
    amendment, certainly supports this view.                           But plenty of evidence
    goes       the    other     way       too.         See    Chief     Justice          Roggensack's
    concurrence/dissent,              ¶¶31-36         (summarizing       the       evidence        which
    supports both an item veto and a part veto).                             Notably, one draft
    amendment         in     1925        would      have     permitted        the        governor     to
    disapprove         "items       or     parts       of     items."5            This     shows     the
    legislature understood the difference between "part" and "item,"
    and that the choice to use this language is reasonably read to
    mean       something.          And    it     is   not    insignificant          that     Henry,    a
    decision         close    in    time       to     the    enactment       of     the    amendment,
    unanimously rejected the petitioners' view.                          I accept Henry as a
    fair,       considered,        and     likely       correct    effort         to      discern   the
    original public meaning of our constitutional text.                                   At the very
    least, the petitioners have not demonstrated that the original
    public meaning is clearly otherwise.
    See 1925 S.J. Res. 23 (proposing to amend Article V,
    5
    Section 10 to allow the governor to "disapprove or reduce items
    or parts of items in any bill appropriating money").
    10
    No.    2019AP1376-OA.bh
    ¶250 For the first 45 years of the partial veto power's
    history,      the     principles       announced     in     Henry,        including     a
    recognition      that    the    broader     constitutional        context      requires
    both       procedural         and     substantive        limitations,          remained
    substantially in place.               Our veto cases that abided by these
    principles are, in my view, unproblematic and consistent with
    the    constitution's         meaning.      See    State   ex     rel.    Finnegan     v.
    Dammann, 
    220 Wis. 143
    , 
    264 N.W. 622
     (1936); State ex rel. Martin
    v. Zimmerman, 
    233 Wis. 442
    , 
    289 N.W. 662
     (1940).
    ¶251 The broadly accepted legal framework, however, started
    to drift in the 1970s.                Around that time, governors began to
    take their partial veto power to new artistic heights.6                          Rather
    than       maintain     the    twin      pillars    of     both     procedural        and
    substantive limitations on that power, this court started to
    jettison      its     commitment       to   any    standard       other      than     the
    requirement that after a partial veto the part approved must be
    a complete, entire, and workable law.
    ¶252 This change was explicitly undertaken in Kleczka, 
    82 Wis. 2d 679
    .          There, the legislature had proposed allowing a
    taxpayer to effectively increase her tax liability such that $1
    would be deposited into the Wisconsin Election Campaign Fund.
    
    Id. at 685
    .         As partially vetoed by the governor, the published
    law enabled the taxpayer to designate that the campaign fund was
    Among other novelties, governors started removing words
    6
    such as "not" from sentences to reverse the policy enacted by
    the legislature (i.e., an "editing veto").      See Richard A.
    Champagne,   Legislative   Reference    Bureau, The   Wisconsin
    Governor's Partial Veto, at 14-15 (2019).
    11
    No.     2019AP1376-OA.bh
    to receive $1 from the state's general funds.                                 
    Id.
         We upheld
    the veto, and expressly dispensed with Henry's discussion of
    inseparable        provisos       or     conditions.              
    Id. at 711-15
    .          By
    sanctioning       this     action,       we    allowed      the     governor         to   take    a
    policy proposal from the legislature, edit the words, and create
    a     different     policy        that        had    not     been       proposed          by   the
    legislature.
    ¶253 Justice        Hansen      vigorously          dissented       on       the   grounds
    that abandoning any substantive limitations on the partial veto
    authority      could        not        possibly       be        consistent           with      the
    constitutional design.              "It appears," Justice Hansen observed,
    "that we have now arrived at a stage where one person can design
    his own legislation from the appropriation bills submitted to
    him    after   they      have     been        approved     by     the    majority         of   the
    legislature."       
    Id. at 727
     (Hansen, J., dissenting).                            Indeed.      As
    Justice Hansen explained:
    Only the limitations on one's imagination fix the
    outer limits of the exercise of the partial veto power
    by incision or deletion by a creative person. At some
    point this creative negative constitutes the enacting
    of legislation by one person, and at precisely that
    point the governor invades the exclusive power of the
    legislature to make laws.
    
    Id. at 720
    .
    ¶254 Justice Hansen's prescience did not stop this court
    from proceeding further down this path, but we have continued to
    wrestle     with     the     implications            of    our     jurisprudence.                In
    Wisconsin Senate, while upholding the most creative uses yet of
    the partial veto power, we recognized as having obtained the
    12
    No.    2019AP1376-OA.bh
    "force of law" the notion that vetoes cannot change a policy
    proposal's    topic    or     subject     matter    into       something       unrelated.
    
    144 Wis. 2d at 452-53
    .         This    "germaneness"        limitation         was   a
    clear     attempt     to     acknowledge         that    the     constitution         must
    countenance     some        kind    of      substantive        limitation        of    the
    governor's    partial       veto    power.        
    Id.
          While        we    have    since
    reaffirmed    the    germaneness         requirement,      this    court       has    never
    fleshed out what it means or how it operates in practice.                               See
    Citizens     Util.    Bd.     v.    Klauser,      
    194 Wis. 2d 484
    ,          505,    
    534 N.W.2d 608
     (1995); Risser v. Klauser, 
    207 Wis. 2d 176
    , 183, 
    558 N.W.2d 108
     (1997).
    ¶255 Finally, it is worth noting that in direct response to
    gubernatorial       practice       and   the     outer   reaches        of     our    later
    decisions, the people have twice amended the partial veto power
    to prevent the governor from using a partial veto to combine
    sentences or strike letters to make new words.                     Wis. Const. art.
    V, § 10(1)(c).7        These amendments should be given substantive
    effect, but they should not be read as green-lighting everything
    less than the limitations they impose.                     While the amendments
    represent the people's effort to rein in certain excesses, these
    constitutionally           prescribed       procedural         limitations           aren't
    particularly     instructive         regarding      whether       the        constitution
    still contains other substantive limitations on the partial veto
    power.
    7"In approving an appropriation bill in part, the governor
    may not create a new word by rejecting individual letters in the
    words of the enrolled bill, and may not create a new sentence by
    combining parts of 2 or more sentences of the enrolled bill."
    Wis. Const. art. V, § 10(1)(c).
    13
    No.    2019AP1376-OA.bh
    C.    Implementing Doctrine
    ¶256 The core question presented in this case is whether
    and how this court will enforce substantive limitations on the
    scope of the governor's partial veto power moving forward.                                   As
    reflected in the multiplicity of writings in today's decision
    and in the tests put forward by the litigants, it is not always
    easy to discern the line between negating some proposed policies
    in a bill and a veto that strategically edits statutory language
    to create a policy that was not in the legislatively passed
    bill.
    ¶257 One response to this dilemma could be to declare that
    line-drawing        is   too     difficult      and     to   surrender         that    project
    altogether.           Our     more     recent       cases    have    trended          in   this
    direction, enforcing only procedural limitations and offering at
    best       a    tip-of-the-cap        to    future     enforcement        of    substantive
    limitations.         In effect, this leaves the policing of substantive
    limitations to politics rather than constitutional law.                                Such an
    approach is not without merit.                  Not all constitutional questions
    need a judicial referee.                   We must acknowledge that increased
    judicial         patrolling      of    these        constitutional     borderlands           is
    fraught with some danger.8                   Engaging in this line-drawing may
    lead       to   uncertainty      for       political     actors     and    entangle         the
    See, e.g., State ex rel. Friedrich v. Circuit Court for
    8
    Dane Cty., 
    192 Wis. 2d 1
    , 14, 
    531 N.W.2d 32
     (1995) (per curiam)
    ("In these borderlands it is neither possible nor practical to
    categorize governmental action as exclusively legislative,
    executive or judicial.").
    14
    No.    2019AP1376-OA.bh
    judiciary in more political and policy fights.                        And sometimes we
    make       things    worse,      not   better,     when       we    attempt      to   make
    distinctions        that    are——let's     be   honest        here——awfully      hard    to
    delineate with precision from the constitutional text.
    ¶258 That         said,   giving    up     on     judicial      enforcement       of
    constitutional limits poses greater dangers, especially in an
    area so central to our constitutional design for how law is
    made.       We swear an oath to uphold the constitution, and it is
    incumbent on us to defend the separation of powers, even if it
    involves getting a little dirt under our nails.
    ¶259 If we are to retain judicially enforceable substantive
    limitations         on    the    partial   veto        power,      there    remains     the
    difficult task of identifying an implementing doctrine, or legal
    test,9 that gets us to the heart of the constitution's meaning.
    Several options are presented in this case.
    ¶260 The petitioners propose a standard severability test.
    Under this test, the inquiry is whether the legislature intended
    for provisions to be severable.                 Burlington N., Inc. v. City of
    Superior,       
    131 Wis. 2d 564
    ,      580,        
    388 N.W.2d 916
            (1986).
    See Ezell v. City of Chicago, 
    651 F.3d 684
    , 700-04 (7th
    9
    Cir. 2011) (devising an implementing doctrine for Second
    Amendment litigation based on the Supreme Court's original
    public meaning interpretation of that constitutional provision
    in District of Columbia v. Heller, 
    554 U.S. 570
     (2008)).   See
    generally Lawrence B. Solum, The Interpretation-Construction
    Distinction, 
    27 Const. Comment. 95
     (2010) (explaining how
    authoritative legal texts are applied in two stages:  one, the
    text is interpreted to discern its linguistic meaning and
    semantic context, and two, the text is given legal effect by
    translating that meaning and context into implementable legal
    doctrine).
    15
    No.    2019AP1376-OA.bh
    Essentially,      we'd     have    to     determine         whether       the    legislature
    would still have wanted the provisions as vetoed to become law.
    This has the virtue of being grounded in some of the discussion
    in    Henry,     and    theoretically          works       within    existing         judicial
    competence.        But    it     seems    difficult,          if    not     impossible,      to
    determine      the      legislature's          intent         and     preferences         when
    reviewing      discrete    proposals       in     omnibus        bills      reflecting       the
    whole of state government operations.                      This test also depends on
    the petitioners' request that we overrule Henry, which I do not
    believe is warranted.
    ¶261 The      legislature        proposes       a    separate        test     based   on
    Justice Hansen's dissent in Kleczka:                       the part rejected, as well
    as the part remaining, must be a complete, entire, and workable
    law   on   its    own.         Kleczka,    
    82 Wis. 2d at 726
         (Hansen,        J.,
    dissenting).       The petitioners also support this as an acceptable
    approach.        This    test     has    the     virtue      of     being       an   objective
    inquiry that does not entangle the judiciary in subjectively
    evaluating policy proposals.                But as the Chief Justice points
    out, there is no basis in the constitutional text to suggest
    that the rejected part must stand on its own as though it were
    itself     enacted        law.           See      Chief          Justice        Roggensack's
    concurrence/dissent, ¶89.               Justice Hansen's test is at best an
    indirect    way    of    getting     at    the    core       constitutional           line   of
    demarcation:           allowing    the    governor          to   create      something       the
    legislature has not proposed, rather than just approve or veto
    separable proposals.              In addition, the legislature's proposal
    appears to be a backdoor way to turn the part veto into an item
    16
    No.   2019AP1376-OA.bh
    veto, or very close to it.           And this too does not square with
    the proposition announced in Henry that the constitutional text
    allows governors to strike portions of proposals smaller than an
    item.
    ¶262 The Chief Justice adopts and attempts to breathe life
    into the germaneness requirement discussed in Wisconsin Senate.
    See    Chief    Justice   Roggensack's       concurrence/dissent,       ¶¶91-94.
    This test has the virtue of being grounded in our precedent.
    Moreover,      the   Wisconsin   Senate     court    adopted   the   germaneness
    requirement, which focuses on the topic or subject matter of a
    provision, as a nod to the need for some substantive limitation
    on    unadulterated    gubernatorial      creation     of   legislation.      See
    Wis. Senate, 
    144 Wis. 2d at 451-52
    .                 But this standard suffers
    from some flaws as well.         As to its foundation, the germaneness
    requirement has not been firmly rooted in the constitutional
    text, but instead in the historical practice of the legislative
    and executive branches.           
    Id. at 437, 452-53
    .             Second, while
    cited, none of our cases have done much to explain what this
    requirement actually means or how it would guide legal analysis
    going forward.         See Citizens Util. Bd., 
    194 Wis. 2d at 505
    ;
    Risser, 
    207 Wis. 2d at 183
    .         Finally, it does not seem to get to
    the core issue of policy creation by the governor.                    It is far
    too underinclusive.        A topicality approach would presumably let
    the governor rewrite laws to create new policy based on the same
    topic    as    the   legislature's     proposal,       thereby    allowing    the
    governor to usurp the role of the legislature in violation of
    the structural separation of powers.                  In other words, as an
    17
    No.   2019AP1376-OA.bh
    implementing doctrine, it does not do well in doing what any
    good legal test should do:             allowing the original public meaning
    of the constitutional text to come to life when applied to a new
    set of facts.
    ¶263 Justice Kelly proposes yet another way.                       His writing
    does     an    excellent       job     outlining     the        separation-of-powers
    problems with our current approach.                  Justice Kelly frames his
    proposed legal test as whether the legislature voted on the
    policy proposal.           At a high level, I agree the question is
    whether the governor vetoed a policy the legislature proposed
    and passed, which is permissible, or created a new policy the
    legislature did not propose or pass, which is not.                              But in
    application,      Justice      Kelly's     opinion     would     appear    to   require
    sweeping away much if not all of our cases, including Henry.                          I
    do     not    believe    the    constitutional       standard       we    agree    upon
    requires      going     this    far.       I    accept     Henry's       holding   that
    something less than a separate item may be vetoed, and this will
    necessarily      involve       some    modification        of    the     legislature's
    policy choice.          So while I agree with Justice Kelly on the core
    constitutional limits, I do not agree with his application of
    that standard.
    ¶264 While        future        litigation        will      surely       provide
    opportunities      to    refine      the   analysis,     the     principles     derived
    from our constitutional text, structure, and early cases draw
    sufficient lines to decide this case.                The partial veto power is
    broad and expansive.           When presented with an appropriation bill
    containing various legislative proposals, the governor can——as a
    18
    No.    2019AP1376-OA.bh
    general matter——negate some proposals and accept others.                                    This
    will necessarily effect a partial change in the policy soup
    reflected in the proposed bill.                      But what the governor may not
    do is selectively edit parts of a bill to create a new policy
    that    was       not    proposed       by    the    legislature.              He   may    negate
    separable proposals actually made, but he may not create new
    proposals not presented in the bill.
    ¶265 By          way   of   a    hypothetical,       imagine       the       legislature
    proposes that $500,000 be appropriated for the building of a
    house, which may be painted white or blue or brown.                                  Under the
    principles derived from the constitutional text and our early
    cases, the governor could strike the word "brown" so that the
    house may only be white or blue.                       But the governor could not
    strike words to create a law that simply appropriates $500,000
    to   the      general         fund.10        While   some       policy    modification         is
    inherent in striking parts of a proposal, a governor may not
    usurp       the   legislature's         lawmaking      role      by   creating        a    policy
    proposal that was not previously there.
    ¶266 Putting             this     together,          I     conclude          that      the
    petitioners' request that we overturn Henry and our early cases
    should be rejected based on the arguments presented in this
    case.       But I agree that later cases must be revisited insofar as
    they abandoned the core principles undergirding the way laws are
    As discussed further below, this type of gubernatorial
    10
    creation is similar to the local road improvement fund vetoes,
    which were an effective rewriting of specific provisions to
    create a generic appropriation for an undefined local grant.
    19
    No.    2019AP1376-OA.bh
    made    pursuant    to     our     constitution.11          Rather        than     simply
    approving or disapproving of proposed policies, the governor's
    partial veto power cannot be converted into a tool for wholesale
    policy creation.           By turning the governor into a one-person
    legislature      subject    only    to    a    two-thirds    override        vote,    our
    basic constitutional structure is turned on its head.
    II.    APPLICATION
    ¶267 Applying those principles to this case, three of the
    four sets of partial vetoes challenged by the petitioners go
    beyond what the constitution permits.
    ¶268 We begin with the sole veto challenge that survives in
    light of our constitutional framework.                     In 2019 Wis. Act 9,
    § 1988b, the legislature sought to amend the registration fees
    assessed    to     truck    owners       based    on    vehicle         weight.      The
    preexisting      fees    for    vehicles       weighing    not    more     than     4,500
    pounds,     6,000       pounds,     8,000       pounds,     and         10,000     pounds
    respectively     were    $75,     $84,    $106,   and     $155.         § 1988b.      The
    legislature proposed modifications to make each of them $100.
    Id.     The governor accepted the increased fee for the lighter
    weight classifications, but rejected the reduction of the fee
    for the heavier vehicles. Id.                 This rejection of the proposed
    Accordingly, I agree with petitioners that State ex rel.
    11
    Kleczka v. Conta, 
    82 Wis. 2d 679
    , 
    264 N.W.2d 539
     (1978) is
    "unsound in principle" and must be overruled. Johnson Controls,
    Inc. v. Emp'rs Ins. of Wausau, 
    2003 WI 108
    , ¶99, 
    264 Wis. 2d 60
    ,
    
    665 N.W.2d 257
    .    Insofar as our later decisions have treated
    Kleczka as pronouncing that a veto shall stand simply if it
    leaves a complete, entire, and workable law, these statements
    too must be withdrawn.
    20
    No.    2019AP1376-OA.bh
    decreases in two registration fees may not reflect the uniform
    schedule        the    legislature       was   apparently     intending.              But    the
    governor        here     chose    a      partially       uniform    fee        schedule       by
    accepting part of the proposed fee schedule and rejecting part
    of the new fee schedule.                 These partial vetoes served to negate
    parts      of    the    broader       policy    proposal.          In    rejecting          this
    proposal in part, the governor did not cobble together words or
    phrases to create a new policy or fee.                      Rather, he declined to
    adopt part of a policy change advanced by the legislature.                                   See
    
    Wis. Stat. § 341.25
    (2)(a)-(cm) (2017-18).12
    ¶269 The         other     three    sets      of   partial        vetoes,       however,
    cannot      be        upheld.          All     three      exceed         the     governor's
    constitutional power to partially veto appropriation bills.
    ¶270 First, faced with an appropriation for the replacement
    of school buses, the governor used multiple vetoes to create an
    appropriation for alternative fuels.                     Wisconsin is a beneficiary
    of   the    Environmental         Mitigation        Trust    created       by     a   partial
    consent decree in In re Volkswagen, 
    2016 WL 6442227
     (N.D. Cal.
    2016).          In Act 9, the legislature enacted two provisions to
    address the allocation of these funds, §§ 55c and 9101(2i).                                  The
    governor partially vetoed § 55c as follows:
    16.047(4s) of the statutes is created to read:
    16.047(4s) SCHOOL BUS REPLACEMENT GRANTS. (a) In this
    subsection:
    1. "School board" has the meaning given in s. 115.001
    (7).
    All subsequent references to the Wisconsin Statutes are
    12
    to the 2017-18 version.
    21
    No.   2019AP1376-OA.bh
    2. "School bus" has the meaning given in s. 121.51
    (4).
    (b) The department [of administration] shall establish
    a program to award grants of settlement funds from the
    appropriation under [Wis. Stat. §] 20.855(4)(h) to
    school boards for the replacement of school buses
    owned and operated by the school boards with school
    buses that are energy efficient, including school
    buses that use alternative fuels. Any school board may
    apply for a grant under the program.
    (c) As a condition of receiving a grant under this
    subsection, the school board shall provide matching
    funds equal to the amount of the grant award.
    (d) A school board may use settlement funds awarded
    under this subsection only for the payment of costs
    incurred by the school board to replace school buses
    in accordance with the settlement guidelines.
    2019 Wis. Act 9, § 55c.            Removing the vetoed words, 
    Wis. Stat. § 16.047
    (4s)    now      reads:       "The      department    shall       establish   a
    program   to       award     grants       of    settlement        funds     from     the
    appropriation under [Wis. Stat. §] 20.855(4)(h) for alternative
    fuels."      The    governor       also    vetoed      in   full    a     nonstatutory
    provision regarding the allocations of these funds.                         2019 Wis.
    Act 9, § 9101(2i).13
    ¶271 The      legislature's          budget    bill    did    not     propose    an
    appropriation       in     whole   or     in    part    for   alternative          fuels
    generally.     Instead, the legislature proposed an appropriation
    13The legislature's proposal stated:    "Of the settlement
    funds in [Wis. Stat. §] 20.855(4)(h), during the 2019-21 fiscal
    biennium, the department of administration shall allocate
    $3,000,000 for grants under [Wis. Stat. §] 16.047(4s) for the
    replacement of school buses."
    22
    No.    2019AP1376-OA.bh
    for the replacement of school buses.14                       While both proposals may
    have similar green energy goals, the governor's partial vetoes
    created an entirely new policy proposal that spends money in
    ways        not    proposed        in     the         legislature's           bill.       This
    gubernatorial-created              policy       sidestepped       the    constitutionally
    mandated procedures governing how a bill becomes a law.
    ¶272 Second, the governor used a trio of vetoes to rewrite
    an appropriation for local road funding into an appropriation
    for some other undefined local grant.                       The governor began with a
    partial       veto    of    Act     9,        § 126     (schedule       item     
    Wis. Stat. § 20.395
    (2)(fc))           as    follows:           "(fc)   Local       roads    improvement
    discretionary                             supplement . . . 90,000,000[inserting
    75,000,000]."         Next, the governor partially vetoed Act 9, § 184s
    as follows:          "20.395(2)(fc) of the statutes is created to read:
    20.395(2)(fc) Local roads improvement discretionary supplement.
    From    the       general       fund,    as    a     continuing     appropriation,         the
    amounts       in     the    schedule          for     the   local   roads        improvement
    discretionary supplemental grant program under s. 86.31 (3s)."
    Wisconsin Stat. § 20.395(2)(fc) now reads:                          "Local supplement.
    From    the       general       fund,    as    a     continuing     appropriation,         the
    amounts in the schedule for local grant."                       Finally, the governor
    vetoed in full Act 9, § 1095m, which detailed how the Department
    The governor's budget had proposed utilizing these funds
    14
    to allow for "the installation of charging stations for vehicles
    with an electric motor," which the legislature rejected in favor
    of creating a school bus modernization fund. See Chief Justice
    Roggensack's concurrence/dissent, ¶14 & n.6-7.    In effect, the
    governor's vetoes could allow for something the legislature
    considered but rejected in enacting its own policy proposal.
    23
    No.   2019AP1376-OA.bh
    of   Transportation    was    to     structure      and       allocate     the
    discretionary grants for local road improvements.15
    ¶273 The   legislature   did   not   propose   a     broad    and   vague
    appropriation for local grants in whole or in part.               Rather, the
    legislature detailed a grant program for the express purpose of
    improving local roads.    By clever editing, the governor created
    a new appropriation out of thin air.        But again, appropriations
    15Prior to the governor's veto of this provision in full,
    it provided:
    86.31(3s) of the statutes is created to read:
    86.31(3s) DISCRETIONARY SUPPLEMENTAL GRANTS. (a) Funds
    provided under [Wis. Stat. §] 20.395(2)(fc) shall be
    distributed under this subsection as discretionary
    grants   to   reimburse  political    subdivisions for
    improvements.     The department [of transportation]
    shall solicit and provide discretionary grants under
    this subsection until all funds appropriated under
    [§] 20.395(2)(fc) have been expended.
    (b)1. From the appropriation under [§] 20.395(2)(fc),
    the department shall allocate $32,003,200 in fiscal
    year   2019−20,  to   fund   county   trunk   highway
    improvements.
    2. From the appropriation under [§] 20.395(2)(fc), the
    department shall allocate $35,149,400 in fiscal year
    2019−20, to fund town road improvements.
    3. From the appropriation under [§] 20.395(2)(fc), the
    department shall allocate $22,847,400 in fiscal year
    2019−20,   to   fund   municipal  street   improvement
    projects.
    (c) Notwithstanding sub. (4), a political subdivision
    may apply to the department under this subsection for
    reimbursement of not more than 90 percent of eligible
    costs of an improvement.
    2019 Wis. Act 9, § 1095m.
    24
    No.   2019AP1376-OA.bh
    must originate in the legislature, which has the power to enact
    such laws in the first instance.           Wis. Const. art. IV, § 17(2),
    § 19; id. art. VIII, § 2.           While the governor may generally
    accept    or   reject   appropriations     proposed   to     him,    he   cannot
    through    creative     editing   author    a   new   appropriation        never
    proposed to him.
    ¶274 Finally, the governor created a new vaping-related tax
    not proposed by the legislature.          The vetoed provision reads:
    139.75 (14) of the statutes is created to read:
    139.75 (14) "Vapor product" means a noncombustible
    product that produces vapor or aerosol for inhalation
    from the application of a heating element to a liquid
    or other substance that is depleted as the product is
    used, regardless of whether the liquid or other
    substance contains nicotine.
    2019 Wis. Act 9, § 1754.          As enacted by the legislature, this
    section taxed the hardware that produces vapor as a result of
    applying the heating element to the liquid.            Through his vetoes
    the governor created a new tax on the liquid which goes inside
    the device, often sold separately.
    ¶275 Once more, a tax on the liquid inside a vaping device
    was not proposed to the governor.          His veto went beyond negating
    a proposal; he created a new tax on a product.                      Because the
    legislature did not propose this new tax, the governor did not
    have the power to rewrite language to create it.                 This kind of
    editing exceeds the governor's partial veto power.
    25
    No.    2019AP1376-OA.bh
    III.    CONCLUSION
    ¶276 Faithfulness        to     the      whole     constitution        and    the
    structure it establishes means our partial veto jurisprudence
    needs a partial reset.         We cannot myopically focus our attention
    on the words of the partial veto provisions in our constitution
    at the expense of the rest of the document's text.                       Early cases
    established principles outlining a broad and expansive partial
    veto power that is no doubt legislative in nature.                           I accept
    those cases and the basic framework they outlined.                           But more
    recent cases, in combination with gubernatorial creativity, have
    upset the constitutional order and allowed governors to invade
    the   lawmaking     powers     of   the     legislature.          It    is   time    to
    reestablish these core constitutional principles.                         I conclude
    that three sets of vetoes challenged here go beyond what the
    constitution      permits.16        For     these      reasons,    I    respectfully
    concur.
    ¶277 I   am     authorized       to      state     that     Justice     ANNETTE
    KINGSLAND ZIEGLER joins this concurrence.
    A
    16   compelling   case  can   be   made   that  prospective
    application of the new rule announced in this case is warranted
    here.   See State v. Beaver Dam Area Dev. Corp., 
    2008 WI 90
    ,
    ¶¶95-96, 
    312 Wis. 2d 84
    , 
    752 N.W.2d 295
     (explaining when
    prospective application is warranted).        However, under the
    circumstances, I join the court's mandate that grants the relief
    requested for all vetoes we determine are unconstitutional.
    26
    No.   2019AP1376-OA.bh
    1