Johnson v. Watts Regulator Co. ( 1995 )

                      UNITED STATES COURT OF APPEALS
                          FOR THE FIRST CIRCUIT
    No. 95-1002
                              JAMES JOHNSON,
                           Plaintiff, Appellee,
                     WATTS REGULATOR COMPANY, ET AL.,
                         Defendants, Appellants.
           [Hon. Joseph A. DiClerico, Jr., U.S. District Judge]
                          Selya, Circuit Judge,
                     Campbell, Senior Circuit Judge,
                         and Cyr, Circuit Judge.
         Eleanor  H. MacLellan,  with  whom Sean  M.  Dunne, Ross  M.
    Weisman, and Sulloway & Hollis were on brief, for appellants.
         Christopher  J.  Seufert,  with  whom  Seufert  Professional
    Association was on brief, for appellee.
                             August 23, 1995
              SELYA,  Circuit  Judge.   This  appeal  requires us  to
                        SELYA,  Circuit  Judge.
    address,  for   the  first  time,  a   "safe  harbor"  regulation
    promulgated  by the Secretary of Labor (the Secretary) as a means
    of exempting certain group insurance programs from the strictures
    of the Employee Retirement Income  Security Act of 1974  (ERISA),
    29 U.S.C.    1001-1461.  Determining, as we do, that the district
    court  appropriately applied  the regulation,  and  discerning no
    clear  error in the court's  factual findings on  other issues in
    the case, we affirm the judgment below.
              I.  BACKGROUND
              Plaintiff-appellee  James Johnson worked  as a forklift
    operator  at the  Webster  Valve division  of defendant-appellant
    Watts Regulator Co. (Watts) in Franklin, New Hampshire.  While so
    employed, plaintiff  elected to participate in  a group insurance
    program made available to Watts' employees by defendant-appellant
    CIGNA Employee  Benefit Company  d/b/a Life Insurance  Company of
    North  America (CIGNA).    Under the  program plaintiff  received
    insurance protection against accidental death, dismemberment, and
    permanent disability.   He  paid  the premium  through a  payroll
    deduction plan.  Watts, in turn, remitted the premium payments to
              On  June 15, 1990, while  a participant in the program,
    plaintiff  sustained  a  severe   head  injury  in  a  motorcycle
    accident.  He remained disabled for the ensuing year, and, having
    crossed the policy's temporal  threshold, he applied for benefits
    on  July 17,  1991.   CIGNA  turned  him down,  claiming  that he
    retained the residual capacity  to do some work.   Plaintiff then
    sued Watts and CIGNA in a New Hampshire state court.  Postulating
    the  existence  of   an  ERISA-related   federal  question,   the
    defendants removed the action to the district court.
              Following  an evidentiary  hearing, the  district court
    ruled that ERISA did not pertain.  See Johnson v. Watts Regulator
    Co.,  No.  92-508-JD,  
    1994 WL 258788
      (D.N.H.  May  3,  1994).
    Nevertheless,  the court  denied  plaintiff's  motion to  remand,
    noting diverse citizenship and the  existence of a controversy in
    the  requisite amount.   See  28 U.S.C.    1332(a).   The parties
    subsequently tried the  case to the bench.   The judge heard  the
    evidence, perused  the group  policy, applied New  Hampshire law,
    found  plaintiff  to be  totally  and  permanently disabled,  and
    awarded the  maximum benefit,  together with attorneys'  fees and
    costs.  See Johnson  v. Watts Regulator Co., No.  92-508-JD, 
    1994 WL 587801
     (D.N.H. Oct. 26, 1994).  This appeal ensued.
              II.  THE ERISA ISSUE
              The  curtain-raiser  question  in  this  case  involves
    whether  the  program  under  which Johnson  sought  benefits  is
    subject to Title  I of  ERISA.  Confronting  this issue  requires
    that  we   interpret  and  apply  the   Secretary's  safe  harbor
    regulation,  29 C.F.R.   2510.3-1(j) (1994).  We divide this part
    of our analysis into  four segments.   First, we explain why  the
    curtain-raiser question matters.   Second, we limn the applicable
    standard  of review.  Third, we discuss the regulation itself and
    how it fits into the statutory and regulatory scheme.  Fourth, we
    scrutinize the  record and  test the district  court's conclusion
    that the program is within the safe harbor.
                        A.  The ERISA Difference.
                                  A.  The ERISA Difference.
              From   the  earliest  stages   of  the   litigation,  a
    controversy  has raged  over  the relationship,  if any,  between
    ERISA  and the  group  insurance program  underwritten by  CIGNA.
    This controversy  stems from  perceived self-interest:   if ERISA
    applies, preemption is  triggered, see 29 U.S.C.    1144(a), and,
    in many situations, the substitution of ERISA principles (whether
    derived from the statute  itself or from federal common  law) for
    state-law  principles  can make  a  pronounced  difference.   For
    example, ERISA preemption may cause potential state-law  remedies
    to vanish, see,  e.g., Carlo  v. Reed Rolled  Thread Die Co.,  
    49 F.3d 790
    , 794 (1st Cir.  1995);  McCoy v. Massachusetts Inst.  of
    950 F.2d 13
    , 18  (1st Cir. 1991),  cert. denied, 
    504 U.S. 910
     (1992), or may change the standard of review, see, e.g.,
    Firestone  Tire & Rubber Co. v.  Bruch, 
    489 U.S. 101
    , 115 (1988),
    or may affect the  admissibility of evidence, see, e.g.,  Taft v.
    Equitable Life Ins. Co., 
    9 F.3d 1469
    , 1471-72 (9th Cir. 1993), or
    may determine whether a jury trial is available, see, e.g., Blake
    v. Unionmutual Stock  Life Ins.  Co., 
    906 F.2d 1525
    , 1526  (11th
    Cir. 1990).
              We are  uncertain which of these  boggarts has captured
    the  minds of the protagonists in this  case.  But exploring that
    question does not strike us  as a prudent use of scarce  judicial
    resources.   Given the  marshalled realities    the parties agree
    that the ERISA difference is of potential significance here; they
    successfully persuaded the district court to that view; and it is
    entirely plausible under the circumstances of this case  that the
    applicability vel non of ERISA makes a meaningful difference   we
    refrain from  speculation about  the parties' tactical  goals and
    proceed directly  to a determination  of whether the  court below
    correctly concluded that state law provides the rule of decision.
                         B.  Standard of Review.
                                   B.  Standard of Review.
              The question  of whether ERISA applies  to a particular
    plan or program requires an evaluation of the facts combined with
    an elucidation of the law.  See, e.g., Kulinski v. Medtronic Bio-
    Medicus,  Inc., 
    21 F.3d 254
    , 256 (8th Cir. 1994) (explaining that
    the existence  of an ERISA plan  is a mixed question  of fact and
    law);  Peckham v. Gem  State Mut., 
    964 F.2d 1043
    , 1047 n.5 (10th
    Cir. 1992)  (similar).  For  purposes of appellate  review, mixed
    questions  of  fact and  law ordinarily  fall along  a degree-of-
    deference  continuum,  ranging  from  plenary   review  for  law-
    dominated  questions  to  clear-error review  for  fact-dominated
    questions.  See In re Extradition of Howard, 
    996 F.2d 1320
    , 1327-
    28   (1st  Cir.   1993).     Plenary   review   is,  of   course,
    nondeferential, whereas clear-error  review is quite deferential.
    See id.  Both standards are in play here.
              The interpretation  of a  regulation presents  a purely
    legal question, sparking de  novo review.  See, e.g.,  Strickland
    v.  Commissioner, Me. Dep't  of Human Serv., 
    48 F.3d 12
    , 16 (1st
    Cir. 1994); Liberty Mut.  Ins. Co. v. Commercial Union  Ins. Co.,
    978 F.2d 750
    ,  757 (1st  Cir. 1992).   Once  the meaning  of the
    regulation has been clarified, however, the "mixed" question that
    remains    the regulation's applicability  in a given  case   may
    require factfinding, and if it does, that factfinding is reviewed
    only for  clear error.  To that extent, the existence of an ERISA
    plan  becomes primarily  a  question of  fact.   See  Wickman  v.
    Northwestern  Nat'l  Ins. Co.,  
    908 F.2d 1077
    , 1082  (1st Cir.),
    cert. denied,  
    498 U.S. 1013
     (1990); Kanne  v. Connecticut  Gen.
    Life Ins.  Co., 
    867 F.2d 489
    , 492  (9th Cir. 1988), cert. denied,
    492 U.S. 906
                  C.  Statutory and Regulatory Context.
                            C.  Statutory and Regulatory Context.
              Congress  enacted ERISA  to  protect the  interests  of
    participants in employee  benefit plans (including  the interests
    of  participants' beneficiaries).  See 29 U.S.C.   1001(a) & (b);
    see also  Curtiss-Wright Corp. v. Schoonejongen, 
    115 S. Ct. 1223
    1230 (1995); Fort Halifax Packing Co. v. Coyne, 
    482 U.S. 1
    , 15-16
    (1987).  ERISA safeguards  these interests in a variety  of ways,
    e.g.,   by  creating   comprehensive  reporting   and  disclosure
    requirements, see 29 U.S.C.    1021-1031, by setting standards of
    conduct   for  fiduciaries,   see  id.       1101-1114,   and  by
    establishing an appropriate remedial  framework, see id.    1131-
    1145.  An  integral part  of the  statutory scheme  is a  broadly
    worded  preemption clause  that, in  respect to  covered employee
    benefit plans,  sets to  one side  "all  laws, decisions,  rules,
    regulations, or other State  action having the effect of  law, of
    any State."  Id.   1144(a).  The purpose of the preemption clause
    is to enhance the  efficient operation of the federal  statute by
    encouraging  uniformity  of  regulatory  treatment   through  the
    elimination of state and local supervision over ERISA plans.  See
    Ingersoll-Rand Co. v. McClendon, 
    498 U.S. 133
    , 142 (1990); McCoy,
    950 F.2d at 18.
              For an employee welfare benefit plan or program to come
    within ERISA's sphere of influence, it  must, among other things,
    be  "established  or maintained"  by  an  employer,1 an  employee
    organization,  or  both.   See  29  U.S.C.     1002(1); see  also
    Wickman, 908 F.2d at 1082 (enumerating necessary components of an
    ERISA plan);  Donovan v.  Dillingham, 
    688 F.2d 1367
    ,  1370 (11th
    Cir. 1982) (en  banc)(same).   The parties agree  that the  group
    insurance program that CIGNA wrote for Watts' employees, covering
    accidental  death,  dismemberment,   and  permanent   disability,
    qualifies  as a  "program" of  employee welfare benefits  as that
    term is used in the statute.  See generally 29  U.S.C.   1002(1).
    Hence, the ERISA question  reduces to whether the program  is one
    "established or maintained" by an employer.  
              To  address this  very  requirement, the  Secretary  of
    Labor, pursuant to 29 U.S.C.   1135 (authorizing the Secretary to
    promulgate  interpretive   regulations  in  the   ERISA  milieu),
    promulgated a safe harbor regulation describing when (and to what
    extent) an  employer or  a trade union  may be  involved with  an
         1The statute also requires that the  employer be "engaged in
    commerce" or in an  industry or activity affecting commerce.   29
    U.S.C.     1003(2).   It  is  undisputed  that  Watts meets  this
    employee  welfare benefit  program without  being deemed  to have
    "established or maintained" it.   See 40 Fed. Reg.  34,527 (1975)
    (explaining the rationale underlying the safe harbor regulation);
    see also Silvera v. Mutual Life Ins. Co., 
    884 F.2d 423
    , 426 (9th
    Cir.  1989); see generally  Ronald J.  Cooke, ERISA  Practice and
    Procedure    2.06  (1994).   The regulation provides  in relevant
    part that the term 
              "employee  welfare  benefit  plan" shall  not
              include  a  group  or   group-type  insurance
              program offered by an insurer to employees or
              members  of  an employee  organization, under
              (1) No contributions are made by  an employer
              or employee organization;
              (2)   Participation   [in]  the   program  is
              completely   voluntary   for   employees   or
              (3) The  sole  functions of  the employer  or
              employee  organization  with  respect to  the
              program are, without  endorsing the  program,
              to   permit  the  insurer  to  publicize  the
              program to  employees or members,  to collect
              premiums through payroll  deductions or  dues
              checkoffs and to remit  them to the  insurer;
              (4)  The  employer  or employee  organization
              receives no consideration in the form of cash
              or otherwise in connection with  the program,
              other than reasonable compensation, excluding
              any   profit,  for   administrative  services
              actually rendered in connection  with payroll
              deductions or dues checkoffs.
    29  C.F.R.     2510.3-1(j).     A  program  that  satisfies   the
    regulation's  standards   will  be   deemed  not  to   have  been
    "established  or  maintained" by  the  employer.   The  converse,
    however, is not necessarily true; a program that fails to satisfy
    the regulation's  standards is  not automatically deemed  to have
    been "established or maintained" by the employer, but, rather, is
    subject to further evaluation under the conventional tests.   See
    Hansenv. Continental Ins. Co., 
    940 F.2d 971
    , 976 (5th Cir. 1991).
              Here, we need not proceed beyond the regulation itself.
    The safe harbor dredged by the regulation operates on the premise
    that the  absence of employer involvement  vitiates the necessity
    for ERISA safeguards.  In theory, an employer can assist its work
    force by  arranging for  the provision  of desirable  coverage at
    attractive rates,  but, by complying with  the regulation, assure
    itself that, if  it acts  only as  an honest  broker and  remains
    neutral vis-a-vis the plan's operation, it will not be put to the
    trouble  and expense  that meeting ERISA's  requirements entails.
    Failure  to fulfill any  one of the  four criteria  listed in the
    regulation, however, closes the  safe harbor and exposes a  group
    insurance program, if it otherwise qualifies as an ERISA program,
    to the strictures of the Act.  See  Qualls v. Blue Cross of Cal.,
    22 F.3d 839
    , 843 (9th Cir. 1994); Fugarino v. Hartford Life
    & Accident  Ins. Co., 
    969 F.2d 178
    ,  184 (6th Cir.  1992), cert.
    113 S. Ct. 1401
     (1993); Memorial Hosp. Sys. v. Northbrook
    Life Ins. Co., 
    904 F.2d 236
    ,  241 n.6 (5th Cir. 1990); Kanne, 867
    F.2d at 492.
              In  the instant  case,  the first,  second, and  fourth
    criteria  are not in dispute.  Plaintiff paid the premium without
    the employer's financial assistance; the decision to purchase the
    coverage was his and  his alone; and Watts received  no forbidden
    consideration.   We concentrate,  therefore, on  the regulation's
    third facet.  This is a fitting focus, as the Department of Labor
    has called  the employer neutrality  that the third  facet evokes
    "the  key to the rationale for not  treating such a program as an
    employee benefit plan . . . ."  40 Fed. Reg. 34,526.
              In dealing with the  regulation, courts have echoed the
    agency's view of  the importance  of employer  neutrality.   See,
    e.g., Hensley v. Philadelphia  Life Ins. Co., 878 F.  Supp. 1465,
    1471 (N.D. Ala. 1995); du Mortier v. Massachusetts Gen. Life Ins.
    805 F. Supp. 816
    ,  821  (C.D. Cal.  1992).    But as  the
    regulation  itself indicates, remaining  neutral does not require
    an  employer to  build a  moat  around a  program or  to separate
    itself from all aspects of program administration.  Thus, as long
    as the employer merely advises  employees of the availability  of
    group insurance,  accepts payroll  deductions, passes them  on to
    the insurer, and performs other ministerial tasks that assist the
    insurer in publicizing the program, it will not be deemed to have
    endorsed the  program under  section 2510.3-1(j)(3).   See Kanne,
    867 F.2d  at 492; du  Mortier, 805 F. Supp.  at 821.   It is only
    when an employer purposes to do more, and takes substantial steps
    in  that  direction,  that  it  offends  the  ideal  of  employer
    neutrality  and brings ERISA into the picture.  See, e.g., Kanne,
    867  F.2d at 492-93 (holding  that an employer  group crossed the
    line when it established  a trust entity in its name for purposes
    of  plan  administration); Brundage-Peterson  v.  Compcare Health
    Servs.  Ins. Corp., 
    877 F.2d 509
    , 510-11 (7th Cir. 1989) (finding
    that   an  employer   who  determined   eligibility,  contributed
    premiums, and collected and remitted premiums paid for dependents
    did  not qualify  for  the safe  harbor  exemption); Shiffler  v.
    Equitable Life Assur. Soc.  of U.S., 
    663 F. Supp. 155
    , 161 (E.D.
    Pa. 1986) (finding that an employer that touted a group policy to
    employees  as part  of its  customary benefits package,  and that
    specifically endorsed the  policy, did not  qualify for the  safe
    harbor exemption), aff'd, 
    838 F.2d 78
      (3d Cir. 1988).  This case
    falls  between these  extremes, and  requires us  to clarify  the
    standard for endorsement under section 2510.3-1(j)(3).  
              The  Department of  Labor has  linked endorsement  of a
    program on the part of an employee organization to its engagement
    "in activities that  would lead a  member reasonably to  conclude
    that  the program is part of a benefit arrangement established or
    maintained by the employee organization."  Dep't of Labor Op. No.
    94-26A (1994).2   What is sauce  for the goose  is sauce for  the
    gander.  Thus, we believe that the agency, in a proper case, will
    link  endorsement on  an  employer's part  to  its engagement  in
    activities that would lead a worker reasonably to conclude that a
    particular  group   insurance  program  is  part   of  a  benefit
    arrangement backed by the company.
              This conclusion is bolstered by the Department's stated
    rationale  to  the  effect  that  a  communication  to  employees
         2Opinion letters  issued by the  Secretary of Labor  are not
    controlling even in the cases  for which they are authored.   See
    Reich v. Newspapers  of New Eng.,  Inc., 
    44 F.3d 1060
    ,  1070 (1st
    Cir. 1995).   Nonetheless, courts may derive guidance  from them.
    See id.
    indicating  that an employer has  arranged for a  group or group-
    type insurance program would constitute an endorsement within the
    meaning of  section 2510.3-1(j)(3) if, taken  together with other
    employer  activities, it leads  employees reasonably  to conclude
    that  the  program  is  one  established  or  maintained  by  the
    communicator.   See id.; see also 40 Fed. Reg. 34,526 (explaining
    that the current phrasing  of the safe harbor  provision replaced
    an  earlier   version  requiring   that  the  employer   make  no
    representation to its  employees that the insurance  program is a
    benefit of  employment because critics found  the earlier version
    "too vague and  difficult to apply").   In short, the agency  has
    suggested  that the  employees'  viewpoint should  constitute the
    principal frame of  reference in determining  whether endorsement
              The interpretation of the safe harbor regulation by the
    agency charged with administering and enforcing ERISA is entitled
    to  substantial deference.  See Berkshire Scenic Ry. Museum, Inc.
    v. ICC, 
    52 F.3d 378
    ,  381-82 (1st Cir. 1995); Keyes  v. Secretary
    of  the  Navy, 
    853 F.2d 1016
    ,  1021  (1st  Cir. 1988).    Here,
    moreover, the respect usually accorded an agency's interpretation
    of  a statute is magnified  since the agency  is interpreting its
    own  regulation.   See  Arkansas v.  Oklahoma,  
    503 U.S. 91
    , 112
    (1992);  Puerto Rico Aqueduct & Sewer Auth. v. United States EPA,
    35 F.3d 600
    ,  604 (1st Cir. 1994), cert. denied,  
    115 S. Ct. 1096
    (1995).    So long  as the  agency's  interpretation does  not do
    violence to  the  purpose  and  wording  of  the  regulation,  or
    otherwise cross into forbidden terrain, courts should defer.  See
    Martin v. OSHRC, 
    499 U.S. 144
    , 150 (1991);  see also Stinson  v.
    United  States, 
    113 S. Ct. 1913
    , 1919  (1993) (holding  that an
    agency's  interpretation of  its  own regulations  must be  given
    controlling weight unless plainly erroneous, inconsistent with  a
    federal  statute, or  unconstitutional); Kelly v.  United States,
    924 F.2d 355
    , 361 (1st Cir. 1991) (similar).
              In  this instance,  we believe  that deference  is due.
    The Secretary's sense of the  safe harbor regulation is consonant
    with  both the regulation's text and the overlying statute.  And,
    moreover, looking  at the employer's conduct  from the employees'
    place of vantage best ensures that employer neutrality  remains a
    reality rather  than  a  mere illusion.    Phrased  another  way,
    judging  endorsement  from  the   viewpoint  of  an   objectively
    reasonable employee most efficaciously serves ERISA's fundamental
    objective:  the protection  of employee benefit plan participants
    and their beneficiaries.
              We rule,  therefore, that an  employer will be  said to
    have  endorsed a  program within the  purview of  the Secretary's
    safe  harbor regulation if, in light of all the surrounding facts
    and  circumstances,  an  objectively  reasonable  employee  would
    conclude on the basis of the employer's actions that the employer
    had  not merely  facilitated the  program's availability  but had
    exercised control over it or made it appear to be part and parcel
    of the company's own benefit package.
                              D.  Analysis.
                                        D.  Analysis.
              Here,  the  district court  interpreted  the regulation
    correctly and  concluded that  the company  had not  endorsed the
    group insurance  program.   This conclusion is  fact-driven, and,
    thus,  reviewable only for clear  error.3  See  Cumpiano v. Banco
    Santander P.R., 
    902 F.2d 148
    , 152  (1st Cir. 1990); see also Fed.
    R. Civ. P.  52(a).  Thus, the trier's findings  of fact cannot be
    set aside unless,  on reviewing  all the evidence,  the court  of
    appeals is left  with an  abiding conviction that  a mistake  has
    been  committed.  See Dedham Water Co. v. Cumberland Farms Dairy,
    972 F.2d 453
    , 457 (1st Cir.  1992); Cumpiano, 902  F.2d at
    152-153.  Applying this deferential  standard, we cannot say that
    the trial court's "no endorsement" finding is clearly erroneous.
              The   anatomy   of   the   court's   determination   is
    instructive.  Based primarily on  the testimony of two  corporate
    officials    Watts'  benefits administrator  and Webster  Valve's
    employee relations manager   the court found that the company had
         3The  question of endorsement vel non is a mixed question of
    fact  and law.  In some cases  the evidence will point unerringly
    in one direction so that a  rational factfinder can reach but one
    conclusion.  In those cases, endorsement becomes a matter of law.
    Cf.  Griffin v.  United  States,  
    502 U.S. 46
    ,  55  n.1  (1991)
    (discussing  "adequacy  on  the proof  as  made"  as meaning  not
    whether the evidence  sufficed to  enable an alleged  fact to  be
    found, but, rather,  whether the facts adduced at  trial sufficed
    in  law to support a  verdict); Anderson v.  Liberty Lobby, Inc.,
    477 U.S. 242
    ,  251-52 (1986) (describing the  appropriate mode of
    inquiry for directed verdicts and  summary judgments).  In  other
    cases, the legal significance  of the facts is less  certain, and
    the  outcome will  depend on  the inferences that  the factfinder
    chooses to draw.  See, e.g.,  TSC Indus., Inc. v. Northway, Inc.,
    426 U.S. 438
    , 450 (1976); In  re Varasso, 
    37 F.3d 760
    , 763 (1st
    Cir.  1994).  In those  cases, endorsement becomes  a question of
    fact.  This case is of the latter type.
    made its  employees aware of the opportunity  to obtain coverage,
    but  had stopped short of  endorsing the program.   CIGNA drafted
    the  policy and,  presumably,  set the  premium rates.   Although
    Watts  distributed the sales brochure, waiver-of-insurance cards,
    and enrollment cards, those efforts were undertaken to help CIGNA
    publicize the program; the documents themselves were prepared and
    printed  by CIGNA, and delivered by it to Watts for distribution.
    Watts recommended enrollment via a cover letter (reproduced as an
    appendix hereto)  written on  the letterhead of  Watts Industries
    and signed by  its vice-president for  financial matters.   CIGNA
    typeset the letter and incorporated it into the cover page of the
    brochure.   The letter explicitly informed  Watts' employees that
    the enrollment  decision  was  theirs to  make.    Watts  nowhere
    suggested that it had any  control over, or proprietary  interest
    in,  the group  insurance  program.   And,  finally, neither  the
    letter nor any other passage in the brochure mentioned ERISA.
              The  district   court   also  examined   Watts'   other
    activities concerning  the  program.   Watts  collected  premiums
    through  payroll  deductions,  remitted  the  premiums  to CIGNA,
    issued   certificates  to   enrolled  employees   confirming  the
    commencement of  coverage, maintained  a list of  insured persons
    for  its own records, and  assisted CIGNA in securing appropriate
    documentation when claims eventuated.   Watts' activities in this
    respect consisted principally of filling out the employer portion
    of  the claim form,  inserting statistical information maintained
    in Watts' personnel  files (such as the insured's  name, address,
    age,  classification, and  date  of hire),  making various  forms
    available to  employees (e.g.,  claim forms),4 and  keeping track
    of  employee eligibility.   Watts would follow  up on  a claim to
    determine  its status, if CIGNA  requested that Watts  do so, and
    would occasionally answer a broker's questions about a claim.  In
    sum, Watts  performed only  administrative  tasks, eschewing  any
    role in the substantive aspects of  program design and operation.
    It had no  hand in drafting the plan, working  out its structural
    components,  determining  eligibility for  coverage, interpreting
    policy  language, investigating, allowing and disallowing claims,
    handling litigation, or negotiating settlements.
              In  the last  analysis, the  district court  found that
    Watts' cover  letter fell  short of constituting  an endorsement.
    The  court pointed out that  neither the letter  nor the brochure
    expressly stated that the  employer endorsed the program.   Apart
    from the letter,  the court  concluded that  Watts had  performed
    only ministerial activities,  and that these activities  (whether
    viewed alone or  in conjunction  with the cover  letter) did  not
    rise to the level of an endorsement.
              We believe that  this finding deserves  our allegiance.
    Drawing permissible inferences from the evidence, the trial court
    could  plausibly  conclude  on   this  scumbled  record  that  an
    objectively reasonable employee would not have thought that Watts
    endorsed  the  group insurance  program.   Several considerations
         4CIGNA prepared  and  printed all  such  forms, and  sent  a
    supply of forms to Watts.
    lead us in this direction.  We offer a representative sampling.
              First, we think that  endorsement of a program requires
    more  than  merely  recommending  it.    An  employer's  publicly
    expressed opinion as to  the quality, utility and/or value  of an
    insurance  plan, without  more,  while relevant  to (and  perhaps
    probative of) endorsement, will  most often not indicate employer
    control of the plan.   Second, the administrative  functions that
    Watts   undertook   fit   comfortably   within   the  Secretary's
    regulation.  Activities such  as issuing certificates of coverage
    and  maintaining a list of  enrollees are plainly  ancillary to a
    permitted function (implementing payroll deductions).  Activities
    such  as answering brokers' questions similarly  can be viewed as
    assisting the insurer  in publicizing the plan.  Other activities
    that  arguably fall closer  to the line, such  as the tracking of
    eligibility   status,   are   completely  compatible   with   the
    regulation's aims.   Under the circumstances,  the court lawfully
    could  find that the employer's activities, in the aggregate, did
    not take the case out of  the safe harbor.5  See, e.g., Brundage-
         5Appellants stress the fact that Watts unilaterally prepared
    and filed a Form 5500 with the Internal Revenue Service.  This is
    an  example of the mountain laboring, but bringing forth a mouse.
    Such forms are informational in nature and are designed to comply
    with  various reporting  requirements  that ERISA  imposes.   See
    Cooke,  supra,   3.10, at  3-34.   But, there  is no  evidence to
    suggest that Watts' employees knew of this protective filing, and
    it  is surpassingly  difficult for  us to  fathom how  the filing
    makes  a dispositive  difference.   Although  the inference  that
    compiling the tax form  demonstrated Watts' intent to provide  an
    ERISA  plan does not  escape us, but  cf. Kanne, 867  F.2d at 493
    (explaining  that a brochure describing  a plan as  an ERISA plan
    evidences the intent of the employer to create an ERISA plan, but
    the same may  not be said of the  filing of a tax return),  it is
    entirely  possible, as the plaintiff  suggests, that the form was
    Peterson,  877   F.2d  at  510  (assuming  that   steps  such  as
    "distributing advertising brochures from insurance  providers, or
    answering  questions  of its  employees concerning  insurance, or
    even  deducting  the  insurance  premiums  from   its  employees'
    paychecks and  remitting  them to  the  insurers," do  not  force
    employers out of the  safe harbor provision); du Mortier,  805 F.
    Supp.  at 821 (holding that activities such as maintaining a file
    of informational materials, distributing  forms to employees, and
    submitting completed forms  to the insurer, do  not transcend the
    boundaries of the safe harbor).
              In arguing  for reversal, appellants rely  on Hansen v.
    Continental  Ins. Co., a case that  involved a similar situation.
    In  Hansen, as here, participation in the plan was voluntary, and
    premiums were paid by  the employees via payroll deduction.   See
    Hansen,  940 F.2d at 973.   The employer  collected the premiums,
    remitted them to the  insurer, and employed an  administrator who
    accepted  claim forms and transmitted  them to the  carrier.  See
    id.  at 974.    In addition,  the  employees received  a  booklet
    embossed with  the employer's  corporate logo that  described the
    plan and encouraged employee participation.  The court found that
    the company had endorsed the plan.  See id.
              Despite the resemblances, there  are two critical facts
    that distinguish Hansen  from the case at bar.   First, in Hansen
    the corporate logo was  embossed on the booklet itself,  see id.,
    filed merely as  a precaution.  In any event,  this case turns on
    the employer's activities, not its intentions.
    making  it  appear  that  the  employer  vouched  for  the entire
    brochure (and for the  plan).  Here, however, only  Watts' letter
    bore its  imprimatur.    Second,  and perhaps  more  cogent,  the
    booklet  at issue in Hansen described the policy as the company's
    plan, see id. ("our  plan"), while here, the letter  typeset onto
    the booklet describes  the policy  as a plan  offered by  another
    organization.6   Though the  appellants decry the  distinction as
    merely  a matter  of semantics,  words  are often  significant in
    determining legal  rights and  obligations.  See  generally Felix
    Frankfurter,  Some  Reflections on  the  Reading  of Statutes  29
    (1947)  ("Exactness  in the  use  of words  is the  basis  of all
    serious thinking.").
              In the difference  between "our plan" and "a plan" lies
    the  quintessential meaning of endorsement.  If a plan or program
    is  the  employer's plan  or program,  the  safe harbor  does not
    beckon.  See, e.g., Sorel v. CIGNA, 
    1994 WL 605726
    , at *2 (D.N.H.
    Nov.  1,  1994)  (holding  that statement  describing  policy  as
    employer's  plan  on first  page  of  plan description  indicates
    endorsement); Cockey  v. Life Ins.  Co. of  N. Am., 
    804 F. Supp. 1571
    , 1575 (S.D. Ga. 1992) (finding that when employer presents a
    program to its  employees as an integral part of its own benefits
         6There  may  also  be  a  critical  difference  between  our
    approach to  the  question of  endorsement  and that  adopted  in
    Hansen.  Although the  Hansen court did not articulate  its ratio
    decidendi, at least one district court has come to the conclusion
    that Hansen  analyzed the  situation from  the standpoint  of the
    employer  rather than the employee.  See Barrett v. Insurance Co.
    of  N.  Am., 813  F.  Supp.  798, 800  (N.D.  Ala.  1993).   This
    possibility  renders appellants'  reliance  on Hansen  even  more
    package, the safe harbor is unavailable); Shiffler, 663 F.  Supp.
    at 161  (finding endorsement  because policy  had been  hawked to
    employees  as a part of the company's benefits package); see also
    Dep't of Labor Op.  No. 94-26A, supra (advising that  safe harbor
    is  unavailable  when  a  union, inter  alia,  describes  a group
    insurance program as its  program).  When, however, the  employer
    separates  itself from  the program,  making it  reasonably clear
    that the program is a third  party's offering, not subject to the
    employer's  control, then the safe harbor may be accessible.  See
    Hansen, 940 F.2d at 977; Kanne, 867 F.2d at 493;  Hensley, 878 F.
    Supp. at 1471.
              This  distinction is  sensible.   When  an  objectively
    reasonable  employee reads  a  brochure describing  a program  as
    belonging to his employer, he  is likely to conclude that, if  he
    participates,  he will be dealing  with the employer  and that he
    will therefore enjoy the  prophylaxis that ERISA ensures  in such
    matters.  When the possessive pronoun is eliminated in favor of a
    neutral article, however, the  employee's perception is much more
    likely  to  be  that, if  he  participates,  he  will be  dealing
    directly with  a third party    the  insurer   and  that he  will
    therefore be beyond the scope of ERISA's protections.
              To sum up, we  are drawn to three conclusions.   First,
    the district court did not clearly err in finding that Watts  had
    not endorsed  the group insurance  program.  Second,  the court's
    fact-sensitive determination  that  the program  fits within  the
    parameters  of   the  Secretary's  safe   harbor  regulation   is
    sustainable.   Third, since ERISA does not apply, the court below
    did  not  blunder  in  scrutinizing  the  merits  of  plaintiff's
    contract claim through the prism of state law.
              Appellant asseverates  that, even if  New Hampshire law
    controls,  the judgment below is  insupportable.  We  turn now to
    this asseveration.
              The  starting point  for  virtually any  claim under  a
    policy of insurance is  the policy itself.  Here,  the applicable
    rider promises benefits  to an insured who has been injured in an
    accident,  whose ensuing disability  is "continuous"  and "total"
    for  a year, and who thereafter  remains "permanently and totally
    disabled."  The  rider defines "continuous total disability" as a
    disability  resulting  from  injuries sustained  in  an accident,
    "commencing within  180 days  after the  date  of the  accident,"
    lasting for at least  a year, and producing during  that interval
    "the  Insured's complete inability  to perform every  duty of his
              If an insured meets this benchmark, he must then  prove
    that  he is "permanently and totally disabled."  Under the policy
    definitions,  this  phrase   signifies  "the  Insured's  complete
    inability,  after one  year  of continuous  total disability,  to
    engage in an occupation or employment for which [he] is fitted by
    reason of education, training, or experience for the remainder of
    his  life."   It  is against  this  linguistic backdrop  that  we
    inspect  appellants'  assertion that  the  trial  court erred  in
    finding plaintiff to be totally and permanently disabled.
                         A.  Standard of Review.
                                   A.  Standard of Review.
              In actions  that are tried  to the  court, the  judge's
    findings  of fact  are to  be honored  unless  clearly erroneous,
    paying  due  respect  to  the judge's  right  to  draw reasonable
    inferences  and  to  gauge the  credibility  of  witnesses.   See
    Cumpiano,  902  F.2d  at 152  (citing  Fed.  R.  Civ. P.  52(a));
    Reliance Steel Prods.  Co. v.  National Fire Ins.  Co., 
    880 F.2d 575
    , 576 (1st  Cir. 1989).   A corollary of  this proposition  is
    that, when there are  two permissible views of the  evidence, the
    factfinder's choice  between  them cannot  be clearly  erroneous.
    See Anderson v. City of Bessemer City, 
    470 U.S. 564
    , 574  (1985);
    Cumpiano, 902 F.2d at 152.  In fine, when a case has been decided
    on the facts by  a judge sitting jury-waived, an  appellate court
    must refrain  from any  temptation to  retry  the factual  issues
              There  are, of  course, exceptions  to  the rule.   For
    example, de novo  review supplants clear-error review  if, and to
    the  extent that, findings of  fact are predicated  on a mistaken
    view of  the law.  See,  e.g., United States v.  Singer Mfg. Co.,
    374 U.S. 174
    ,  195 n.9  (1963); RCI  N.E. Servs. Div.  v. Boston
    Edison Co., 
    822 F.2d 199
    , 203  (1st Cir. 1987).   This does  not
    mean, however, that the clearly erroneous standard  can be eluded
    by the simple expedient  of creative relabelling.  See  Cumpiano,
    902 F.2d  at 154; Reliance Steel,  880 F.2d at 577.   For obvious
    reasons, we will not  allow a litigant to subvert  the mandate of
    Rule 52(a) by hosting a masquerade, "dressing factual disputes in
    `legal' costumery."  Reliance Steel, 880 F.2d at 577; accord Dopp
    v.  Pritzker, 
    38 F.3d 1239
    ,  1245 (1st Cir.  1994), cert. denied,
    115 S. Ct. 1959
                              B.  Analysis.
                                        B.  Analysis.
              Appellants   make  two  main  arguments  in  regard  to
    plaintiff's  disability claim.  First, in an effort to skirt Rule
    52(a),  they assert that the district court committed an error of
    law, mistaking the meaning of the phrase "permanently and totally
    disabled" as  that phrase is used  in the policy.   We reject the
    assertion as  comprising nothing  more than  a clumsy  attempt to
    recast a  clear-error challenge  as an issue  of law  in hope  of
    securing  a more welcoming standard of review.  The policy itself
    defines  the operative term,  and the record  makes pellucid that
    the district judge applied the term within the parameters of that
              Appellants' second contention posits that  the district
    court  misperceived  the  facts,   and  that  plaintiff  was  not
    sufficiently  disabled to  merit  an  award  of benefits.    This
    contention also  lacks force.   The  district court  had adequate
    grounds for  deciding that plaintiff was  totally and permanently
    disabled.    The  evidence  showed  that  plaintiff  sustained  a
    devastating brain injury, and that, throughout the year following
    his accident, a number  of physicians found his disability  to be
    continuous.  By and large, plaintiff's  condition did not improve
    significantly during that year  (or thereafter, for that matter).
    Without  exception, the  doctors  concluded that  he could  never
    return to  work as a forklift driver.  To cap matters, the record
    contains  ample  evidence  that  the plaintiff's  disability  was
    permanent  and blanketed  the  universe of  occupations to  which
    plaintiff   a laborer  with a high-school education    might have
              We  need not  cite  book and  verse.   The  court  made
    detailed findings, crediting the  conclusions of four doctors who
    judged  plaintiff  to be  severely  impaired,  both mentally  and
    physically.7  The court also credited an evaluation performed  by
    Sherri  Krasner,  a  speech  and language  pathologist,  and  the
    testimony  of  a   vocational  rehabilitation  counselor,  Arthur
    Kaufman, who offered an opinion that plaintiff was unable to work
    without  constant supervision.   Kaufman stated  that he  did not
    know  of a job suitable  for a person  in plaintiff's condition.8
         7These  experts   included  the  attending   physician  (Dr.
    Martino),    a   neurologist    (Dr.   Whitlock),    a   clinical
    neuropsychologist (Dr. Higgins), and  a psychologist (Dr.  Toye).
    A  fifth  physician,  Dr.  Michele  Gaier-Rush,   also  evaluated
    plaintiff.   CIGNA chose  Dr. Gaier-Rush as  its medical examiner
    but neglected to  provide her with  any of plaintiff's  plentiful
    prior medical records, despite their availability.  She concluded
    that  plaintiff could not perform his usual job but could perform
    a job  "requiring more  mental capacity than  physical capacity."
    She noted, however, that plaintiff had no formal  training beyond
    high school,  and  conceded  that  "[t]his  will  probably  be  a
    permanent  disability as  there  does not  seem  to have  been  a
    significant  improvement in  the past  year."   Consequently, she
    found it doubtful that plaintiff could ever work again.
         8While  Kaufman did say that  plaintiff might be  able to do
    some gainful  employment with  "excessive supervision," and  that
    plaintiff, like  other  patients with  traumatic brain  injuries,
    would  probably benefit  from vocational  rehabilitation, Kaufman
    expressed   doubt  that   plaintiff   would  ever   overcome  his
    impairment.   In short, he lacked  the "capacity to retain  . . .
    On  this record,  the trial court's  total disability  finding is
              Another wave of  appellants' evidentiary attack targets
    the  district  court's  finding  that  plaintiff's disability  is
    permanent.    In  this  respect, appellants  rely  mainly  on the
    physicians'   recommendations   for  rehabilitative   therapy  as
    indicative of the  potential for recovery.   The district  court,
    however, found appellants' inference unreasonable in light of the
    dim  prospects   for  significant   recovery,  the   duration  of
    plaintiff's  inability  to  work,  and the  policy's  failure  to
    require  vocational  rehabilitation  as  a  precondition  to  the
    receipt of  benefits.  These  are fact-dominated issues,  and the
    trial court is in  the best position to calibrate  the decisional
    scales.   See Cumpiano,  902 F.2d  at 152.   Having  examined the
    record with care, we have no reason to suspect that a mistake was
    committed.  See, e.g., Duhaime v. Insurance Co., 
    86 N.H. 307
    , 308
    (1933) (explaining  that, to be permanently  disabled, an insured
    need not be in a condition of "utter hopelessness").
              IV.  CONCLUSION
              We need go  no further.   ERISA does not  apply to  the
    group insurance  program at issue  here.  Moreover,  the district
    court's    factual    findings   survive    clear-error   review.
    Consequently, the court's resolution of the case stands.