Henry Ortiz v. Werner Enterprises, Incorporat , 834 F.3d 760 ( 2016 )

In the
United States Court of Appeals
For the Seventh Circuit
____________________
No.  15-­‐‑2574
HENRY  ORTIZ,
Plaintiff-­‐‑Appellant,
v.
WERNER  ENTERPRISES,  INC.,
Defendant-­‐‑Appellee.
____________________
Appeal  from  the  United  States  District  Court  for  the
Northern  District  of  Illinois,  Eastern  Division.
No.  13  C  8270  —  John  W.  Darrah,  Judge.
____________________
ARGUED  FEBRUARY  8,  2016  —  DECIDED  AUGUST  19,  2016
____________________
Before   POSNER,   EASTERBROOK,   and   HAMILTON,   Circuit
Judges.
EASTERBROOK,   Circuit   Judge.   Henry   Ortiz   worked   as   a
freight   broker   for   Werner   Enterprises,   Inc.,   for   seven   years
until  his  discharge  in  2012.  Werner  says  that  it  fired  Ortiz  for
falsifying  business  records.  Ortiz  says  that  Werner  fired  him
because  of  his  Mexican  ethnicity,  and  he  sued  Werner  under
42  U.S.C.  §1981  and  the  Illinois  Human  Rights  Act,  775  ILCS
5/1-­‐‑101   to   5/10-­‐‑104.   He   also   claimed   that   Werner   subjected
2   No.  15-­‐‑2574
him   to   a   hostile   work   environment   in   violation   of   the   state
law.
Werner  is  a  shipping  company  that  offers  freight  broker-­‐‑
age   as   one   of   its   services;   customers   pay   a   fee   and   Werner
finds   transportation   for   their   loads.   Werner   tracks   its   cus-­‐‑
tomers’   loads   through   a   proprietary   records   system.   Loads
that   have   not   been   matched   with   a   carrier—a   vehicle   hired
to  transport  the  load––appear  in  the  system,  and  for  most  of
Ortiz’s  tenure  at  Werner  any  broker  could  locate  a  carrier  for
any  load.  After  securing  carriers,  brokers  update  the  system
to  confirm  the  transactions,  the  carrier’s  prices,  and  who  gets
credit  for  the  deal.  Werner’s  profit  is  the  difference  between
what   its   customer   pays   and   what   it   pays   the   carrier.   Some
loads  produce  a  loss  when  the  carrier  charges  more  than  the
customer’s   payment.   Werner   compensates   its   brokers   pri-­‐‑
marily   through   a   base   salary,   but   they   can   earn   a   commis-­‐‑
sion   of   up   to   4%   for   months   in   which   they   generate   more
than   a   specified   profit.   Ortiz   usually   received   monthly
commissions  of  around  $1,000.
Permitting  all  brokers  to  book  any  load  led  to  problems,
as   brokers   cherry-­‐‑picked   the   nation’s   most   profitable   loads.
In   2012   Werner   addressed   this   by   assigning   each   broker   a
specific  region.  During  Ortiz’s  final  months  with  Werner  the
regional   broker   bore   responsibility   for   matching   every   load
in  the  region  to  a  carrier  whether  the  load  would  generate  a
profit  or  a  loss.  But  the  system  did  not  automatically  assign
loads—brokers  still  needed  to  search  for  loads  in  their  region
and  update  the  records.
Werner  assigned  Ortiz  the  West  region,  apparently  not  a
lucrative  assignment  during  late  spring  when  high  demand
for  transporting  produce  drives  carriers’  rates  up.  (The  par-­‐‑
No.  15-­‐‑2574   3
ties   don’t   explain   why   Werner   failed   to   change   its   fees   to
match   predictable   cycles   in   the   business.)   During   the   first
week   of   June   2012   Michael   Krikava,   the   branch’s   assistant
manager,  booked  or  directed  another  broker  to  book  six  un-­‐‑
profitable   loads   in   Ortiz’s   name.   There   was   some   variation
in   how   Ortiz’s   name   wound   up   on   each   load:   Krikava
booked   several   loads   in   his   own   name   and   swapped   in
Ortiz’s  name  after  the  loads  had  been  picked  up,  but  others
were   booked   early   in   the   morning   when   Ortiz   may   have
been  out  of  the  office.  Krikava  did  not  notify  Ortiz  of  any  of
these   losing   transactions.   By   the   time   Ortiz   realized   that   he
had  been  assigned  responsibility  for  the  loads,  it  was  too  late
for   him   to   search   for   a   cheaper   carrier.   Ortiz   later   changed
the  records  to  reflect  that  he  had  not  booked  the  loads  or  to
show  lower  rates  that  he  thought  the  carriers  had  agreed  to
charge.   He   then   left   for   a   planned   vacation.   On   his   return
Werner   fired   him.   The   parties   agree   on   this   much,   but   here
their  stories  diverge.  We  recount  Ortiz’s  version,  because  we
must   view   the   evidence   in   the   light   most   favorable   to   the
party  opposing  a  motion  for  summary  judgment.
Ortiz   contends   that   brokers   and   managers   always   noti-­‐‑
fied   one   another   when   booking   an   unprofitable   load   in
someone   else’s   name,   and   that   it   was   atypical   for   someone
else  to  book  so  many  unprofitable  loads  without  so  much  as
a   courtesy   email   or   phone   call.   He   adds   that   Krikava   as-­‐‑
signed   him   unusually   high   losses.   Ortiz   repeatedly   ques-­‐‑
tioned   Krikava   about   the   unprofitable   loads   but   he   refused
to  answer  until  saying:  “Why  won’t  you  just  quit  already?”
Ortiz   maintains   that   he   updated   three   records   to   show
that   he   had   spoken   to   the   carriers,   which   had   agreed   to   ac-­‐‑
cept  a  lower  rate  because  they  had  not  picked  up  the  goods
4   No.  15-­‐‑2574
on   time.   Ortiz   adds   that   he   often   tried   to   negotiate   a   lower
rate  after  a  late  delivery  and  that  carriers  would  sometimes
oblige.   Ortiz   says   that   this   represents   standard   practice   in
the  industry;  other  brokers  offered  similar  testimony.
Ortiz   justifies   the   removal   of   his   name   from   three   other
loads   with   evidence   that   the   branch’s   managers   permitted
brokers   to   delete   their   names   from   unprofitable   loads.   This
practice  allowed  Werner  to  satisfy  the  needs  of  all  of  its  cus-­‐‑
tomers   without   saddling   brokers   with   undue   losses,   thus
protecting   each   broker’s   commission.   Removing   a   broker’s
name  from  a  loss  also  moved  the  loss  off  the  branch’s  books,
increasing   the   branch’s   profits   and   the   managers’   bonuses,
which  are  tied  to  branch  performance.  Other  Werner  brokers
testified   that   this   was   called   “spacing”,   as   in   sending   the
record   of   the   unprofitable   load   into   cyberspace.   (The   evi-­‐‑
dence   does   not   reflect   what   division   of   Werner   ultimately
bore   the   loss   or   how   Werner   reconciled   its   books.)   Because
spacing   was   common   practice,   Ortiz   says,   he   thought   it
permissible   for   him   to   remove   his   name   from   the   money-­‐‑
losing  loads.
On   his   first   morning   back   from   vacation   Ortiz   met   with
Kip   Lass,   the   branch   manager.   Lass   told   Ortiz   that   he   had
been   fired   for   falsifying   records.   Ortiz   argued   that   he   had
not   falsified   records   but   had   corrected   them   by   providing
the  rate  that  the  tardy  carriers  had  agreed  to  charge,  and  he
offered  to  call  the  carriers  to  provide  proof.  Lass  showed  no
interest  in  verifying  Ortiz’s  allegations,  noted  “What’s  done
is  done,”  and  discharged  Ortiz  without  further  investigation.
Ortiz  also  tells  us  that  Krikava  and  Lass  subjected  him  to
a  barrage  of  ethnic  slurs.  They  frequently  used  epithets  such
as   “beaner,”   “taco   eater,”   “fucking   beaner,”   “taco,”   “bean
No.  15-­‐‑2574   5
eater,”   “dumb   Mexican,”   “stupid   Puerto   Rican,”   “dumb
Puerto  Rican,”  “fucking  Puerto  Rican,”  “Puerto  Rican,”  and
“dumb  Jew”  to  refer  to  Ortiz.  Ortiz  contends  that  he  encoun-­‐‑
tered   such   slurs   throughout   his   tenure   at   Werner   and   that
they   increased   in   frequency   and   intensity   in   the   months
leading  up  to  his  discharge.
Werner   asserts   that   the   office   did   not   regularly   notify
brokers  of  loads  recorded  under  their  names,  and  it  contends
that  the  losses  assigned  to  Ortiz  were  not  unusually  high.  It
insists   that   its   brokers   do   not   have   the   authority   to   remove
their   names   from   unprofitable   loads   without   a   manager’s
permission.  It  also  denies  that  it  was  customary  to  cut  a  car-­‐‑
rier’s   rate   for   a   late   delivery—even   if   the   carrier   had   con-­‐‑
sented   to   the   lower   rate.   According   to   Werner,   Ortiz   im-­‐‑
properly  edited  these  three  records  as  well.  Finally,  Werner
denies  that  Lass  and  Krikava  ever  made  derogatory  remarks
to   Ortiz.   But   on   all   of   these   issues   we   must   look   at   the   evi-­‐‑
dence  from  Ortiz’s  perspective.
The  district  court  granted  summary  judgment  to  Werner.
2015   U.S.   Dist.   LEXIS   82952   (N.D.   Ill.   June   25,   2015).   It   dis-­‐‑
missed  Ortiz’s  claim  of  a  hostile  work  environment  for  fail-­‐‑
ure   to   exhaust   his   administrative   remedies,   a   ruling   Ortiz
does  not  contest  on  appeal.  The  judge  looked  at  the  evidence
through   the   “direct”   and   “indirect”   methods   that   courts   of-­‐‑
ten   discuss   in   employment-­‐‑discrimination   cases.   See,   e.g.,
Andrews   v.   CBOCS   West,   Inc.,   743   F.3d   230,   234   (7th   Cir.
2014);  Zaderaka  v.  Illinois  Human  Rights  Commission,  131  Ill.  2d
172,   178–79   (1989).   Admissions   of   culpability   and   smoking-­‐‑
gun   evidence   were   assigned   to   the   “direct”   method   (the
judge   found   no   such   evidence),   while   suspicious   circum-­‐‑
stances  that  might  allow  an  inference  of  discrimination  were
6   No.  15-­‐‑2574
assigned   to   the   “indirect”   method.   The   court   did   not   try   to
aggregate   the   possibilities   to   find   an   overall   likelihood   of
discrimination.
The  district  court  treated  each  method  as  having  its  own
elements  and  rules,  even  though  we  have  held  that  they  are
just   means   to   consider   whether   one   fact   (here,   ethnicity)
caused  another  (here,  discharge)  and  therefore  are  not  “ele-­‐‑
ments”   of   any   claim.   See,   e.g.,   Achor   v.   Riverside   Golf   Club,
117   F.3d   339,   341   (7th   Cir.   1997).   Pursuing   two   “methods”
instead  of  a  unified  inquiry,  the  district  court  elaborated  by
saying  that  Ortiz  could  prevail  only  by  coming  up  with  “ev-­‐‑
idence  that  creates  ‘a  convincing  mosaic  of  discrimination’”.
2015  U.S.  Dist.  LEXIS  82952  at  *12.  It  concluded  that  Ortiz  had
failed   to   present   a   “convincing   mosaic”   under   the   direct
method  because  Lass’s  racial  slurs  did  not  have  anything  to
do   with   Ortiz’s   discharge.   The   court   also   ruled   that   there
was  not  enough  of  a  “mosaic”  under  the  indirect  method  be-­‐‑
cause,  by  removing  his  name  from  the  records  and  changing
the  rates,  he  fell  short  of  Werner’s  expectations.
The   district   court’s   effort   to   shoehorn   all   evidence   into
two  “methods,”  and  its  insistence  that  either  method  be  im-­‐‑
plemented   by   looking   for   a   “convincing   mosaic,”   detracted
attention  from  the  sole  question  that  matters:  Whether  a  rea-­‐‑
sonable  juror  could  conclude  that  Ortiz  would  have  kept  his
job   if   he   had   a   different   ethnicity,   and   everything   else   had
remained   the   same.   See,   e.g.,   Achor;   Troupe   v.   May   Depart-­‐‑
ment  Stores  Co.,  20  F.3d  734  (7th  Cir.  1994).  The  use  of  dispar-­‐‑
ate  methods  and  the  search  for  elusive  mosaics  has  compli-­‐‑
cated   and   sidetracked   employment-­‐‑discrimination   litigation
for   many   years.   During   the   last   decade,   every   member   of
this   court   has   disapproved   both   the   multiple   methods   and
No.  15-­‐‑2574   7
the   search   for   mosaics.   See,   e.g.,   Sylvester   v.   SOS   Children’s
Villages   Illinois,   Inc.,   453   F.3d   900   (7th   Cir.   2006);   Coleman   v.
Donahoe,   667   F.3d   835,   863   (7th   Cir.   2012)   (concurring   opin-­‐‑
ion  joined  by  entire  panel);  Good  v.  University  of  Chicago  Med-­‐‑
ical   Center,   673   F.3d   670,   680   (7th   Cir.   2012);   Harper   v.   C.R.
England,   Inc.,   687   F.3d   297,   314   (7th   Cir.   2012);   Hitchcock   v.
Angel   Corps,   Inc.,   718   F.3d   733,   737   (7th   Cir.   2013);   Perez   v.
Thorntons,  Inc.,  731  F.3d  699,  703  (7th  Cir.  2013);  Chaib  v.  Indi-­‐‑
ana,   744   F.3d   974,   981   (7th   Cir.   2014);   Bass   v.   Joliet   Public
School  District  No.  86,  746  F.3d  835,  840  (7th  Cir.  2014);  Hutt  v.
AbbVie  Products  LLC,  757  F.3d  687,  691  (7th  Cir.  2014);  Simp-­‐‑
son   v.   Beaver   Dam   Community   Hospitals,   Inc.,   780   F.3d   784,
789–90   (7th   Cir.   2015);   Castro   v.   DeVry   University,   Inc.,   786
F.3d   559,   564   (7th   Cir.   2015)   (collecting   cases).   The   time   has
come  to  jettison  these  diversions  and  refocus  analysis  on  the
substantive  legal  issue.
We  start  with  the  “convincing  mosaic.”  This  phrase  orig-­‐‑
inated   in   Troupe   and   was   designed   as   a   metaphor   to   illus-­‐‑
trate   why   courts   should   not   try   to   differentiate   between   di-­‐‑
rect  and  indirect  evidence.  Troupe  explained  that  all  evidence
is   inferential   and   cannot   be   sorted   into   boxes.   All   evidence
should   be   considered   together   to   understand   the   pattern   it
reveals.   But   instead   of   persuading   district   judges   and   liti-­‐‑
gants   to   merge   “direct”   and   “indirect”   methods   into   a   uni-­‐‑
fied  approach,  Troupe  was  understood  by  many  as  adding  a
new  “test”  that  had  to  be  satisfied.  How  a  “mosaic”  could  be
a  “test”  rather  than  a  mental  picture  went  unaddressed—as
did  the  question  how  such  a  legal  test  could  be  rooted  in  the
statutes  that  govern  employment-­‐‑discrimination  cases.
We   tried   to   set   things   straight   in   Sylvester,   453   F.3d   at 903–04,   reminding   lawyers   and   judges   that   Troupe   used
8   No.  15-­‐‑2574
“mosaic”   as   a   metaphor   that   was   designed   to   displace   the
unhelpful   direct   and   indirect   methods   rather   than   add   to
them.   The   reminder   did   not   work,   as   the   district   court’s
opinion  in  this  case  shows.  Despite  follow-­‐‑up  reminders  that
“convincing  mosaic”  is  not  a  legal  test  of  any  kind,  see  Mu-­‐‑
hammad   v.   Caterpillar,   Inc.,   767   F.3d   694,   699–700   (7th   Cir.
2014);  Fleishman  v.  Continental  Casualty  Co.,  698  F.3d  598,  603
(7th   Cir.   2012);   Davis   v.   Time   Warner   Cable   of   Southeastern
Wisconsin,   L.P.,  651   F.3d   664,   672   (7th   Cir.   2011);   it   has   con-­‐‑
tinued  to  be  misused  as  one.
To  make  matters  worse,  this  court  has  itself  occasionally
treated   “convincing   mosaic”   as   a   legal   requirement,   even
while  cautioning  in  other  opinions  that  it  must  not  be  so  un-­‐‑
derstood.  See,  e.g.,  Hatcher  v.  Board  of  Trustees  of  Southern  Illi-­‐‑
nois   University,   No.   15-­‐‑1599   (7th   Cir.   July   14,   2016),   slip   op.
13;  Chaib,  744  F.3d  at  981;  Cloe  v.  Indianapolis,  712  F.3d  1171,
1180  (7th  Cir.  2013);  Smith  v.  Bray,  681  F.3d  888,  901  (7th  Cir.
2012);  Good,  673  F.3d  at  674;  Silverman  v.  Board  of  Education  of
Chicago,  637  F.3d  729,  734  (7th  Cir.  2011);  Phelan  v.  Cook  Coun-­‐‑
ty,  463  F.3d  773,  779  (7th  Cir.  2006);  Koszola  v.  Board  of  Educa-­‐‑
tion   of   Chicago,   385   F.3d   1104,   1109   (7th   Cir.   2004);   Rhodes   v.
Illinois   Department   of   Transportation,   359   F.3d   498,   504   (7th
Cir.  2004);  Cerutti  v.  BASF  Corp.,  349  F.3d  1055,  1061  (7th  Cir.
2003);  Robin  v.  Espo  Engineering  Corp.,  200  F.3d  1081,  1088–89
(7th  Cir.  2000).  Instead  of  simplifying  analysis,  the  “mosaic”
metaphor  has  produced  a  form  of  legal  kudzu.
Today   we   reiterate   that   “convincing   mosaic”   is   not   a   le-­‐‑
gal  test.  We  overrule  the  opinions  in  the  previous  paragraph
to  the  extent  that  they  rely  on  “convincing  mosaic”  as  a  gov-­‐‑
erning  legal  standard.  We  do  not  hold  that  any  of  those  cases
was  wrongly  decided;  our  concern  is  only  with  the  treatment
No.  15-­‐‑2574   9
of   “convincing   mosaic”   as   if   it   were   a   legal   requirement.
From  now  on,  any  decision  of  a  district  court  that  treats  this
phrase   as   a   legal   requirement   in   an   employment-­‐‑
discrimination   case   is   subject   to   summary   reversal,   so   that
the  district  court  can  evaluate  the  evidence  under  the  correct
standard.
That   legal   standard,   to   repeat   what   we   wrote   in   Achor
and  many  later  cases,  is  simply  whether  the  evidence  would
permit  a  reasonable  factfinder  to  conclude  that  the  plaintiff’s
race,   ethnicity,   sex,   religion,   or   other   proscribed   factor
caused   the   discharge   or   other   adverse   employment   action.
Evidence  must  be  considered  as  a  whole,  rather  than  asking
whether  any  particular  piece  of  evidence  proves  the  case  by
itself—or   whether   just   the   “direct”   evidence   does   so,   or   the
“indirect”  evidence.  Evidence  is  evidence.  Relevant  evidence
must  be  considered  and  irrelevant  evidence  disregarded,  but
no   evidence   should   be   treated   differently   from   other   evi-­‐‑
dence  because  it  can  be  labeled  “direct”  or  “indirect.”
Some   opinions   have   observed   that   the   methods   largely
rely  on  the  same  pieces  of  evidence.  See  Harper,  687  F.3d  at
314   (“[T]he   line   between   circumstantial   evidence   under   the
direct  method  and  indirect  evidence  of  discrimination  or  re-­‐‑
taliation   under   the   burden-­‐‑shifting   approach   has   been
blurred  by  the  gradual  integration  of  these  methodologies.”);
Hutt,  757  F.3d  at  691.  Why  have  two  tests  if  they  consider  the
same  information  and  answer  the  same  question?
True,   some   cases   permit   easy   inferences,   such   as   the   fa-­‐‑
bled  employer  who  admits  to  firing  an  employee  because  of
race.  But  even  then  factfinders  must  ask  themselves  what  the
admission   means.   Few   discrimination   cases   are   so   straight-­‐‑
forward—indeed   they   are   often   factually   complex   and   re-­‐‑
10   No.  15-­‐‑2574
quire  sifting  through  ambiguous  pieces  of  evidence.  And  the
direct-­‐‑and-­‐‑indirect   framework   does   nothing   to   simplify   the
analysis.  Instead  it  complicates  matters  by  forcing  parties  to
consider   the   same   evidence   in   multiple   ways   (and   some-­‐‑
times   to   disregard   evidence   that   does   not   seem   to   fit   one
method  rather  than  the  other).
Accordingly,  we  hold  that  district  courts  must  stop  sepa-­‐‑
rating  “direct”  from  “indirect”  evidence  and  proceeding  as  if
they   were   subject   to   different   legal   standards.   Once   again,
this   court   must   accept   its   share   of   the   responsibility—
because   even   as   some   panels   were   disparaging   the   “direct”
and   “indirect”   approaches,   other   panels   were   articulating
them   as   governing   legal   standards.   We   need   to   bring   har-­‐‑
mony   to   circuit   law,   and   the   way   to   do   that   is   to   overrule
Andrews,  743  F.3d  230;  Silverman,  637  F.3d  729;  Hemsworth  v.
Quotesmith.com,  Inc.,  476  F.3d  487  (7th  Cir.  2007);  Rhodes,  359
F.3d  at  498;  Haywood  v.  Lucent  Technologies,  Inc.,  323  F.3d  524(7th   Cir.   2003);   Oest   v.   Illinois   Department   of   Corrections,   240
F.3d   605   (7th   Cir.   2001);   Radue   v.   Kimberly-­‐‑Clark   Corp.,   219
F.3d   612   (7th   Cir.   2000);   Chiaramonte   v.   Fashion   Bed   Group,
Inc.,  129  F.3d  391  (7th  Cir.  1997);  Mechnig  v.  Sears,  Roebuck  &
Co.,  864  F.2d  1359  (7th  Cir.  1988);  and  La  Montagne  v.  Ameri-­‐‑
can  Convenience  Products,  Inc.,  750  F.2d  1405  (7th  Cir.  1984),  to
the  extent  that  these  opinions  insist  on  the  use  of  the  direct-­‐‑
and-­‐‑indirect   framework.   (As   with   the   convincing-­‐‑mosaic
cases,  we  are  not  saying  that  any  of  these  decisions  produced
a  victory  for  the  wrong  litigant;  our  concern  is  with  the  legal
standards  they  employed  rather  than  with  their  outcomes.)
One   point   of   clarification   may   be   helpful.   The   burden-­‐‑
shifting   framework   created   by   McDonnell   Douglas   Corp.   v.
Green,   411   U.S.   792   (1973),   sometimes   is   referred   to   as   an
No.  15-­‐‑2574   11
“indirect”   means   of   proving   employment   discrimination.
Today’s  decision  does  not  concern  McDonnell  Douglas  or  any
other  burden-­‐‑shifting  framework,  no  matter  what  it  is  called
as  a  shorthand.  We  are  instead  concerned  about  the  proposi-­‐‑
tion  that  evidence  must  be  sorted  into  different  piles,  labeled
“direct”   and   “indirect,”   that   are   evaluated   differently.   In-­‐‑
stead,  all  evidence  belongs  in  a  single  pile  and  must  be  eval-­‐‑
uated  as  a  whole.  That  conclusion  is  consistent  with  McDon-­‐‑
nell  Douglas  and  its  successors.
With   the   rat’s   nest   of   surplus   “tests”   removed   from   the
law  of  the  circuit,  we  can  turn  back  to  Ortiz’s  claim  and  his
supporting   evidence.   Stripped   of   the   layers   of   tests,   our
analysis  is  straightforward.
Ortiz   relies   on   his   own   testimony   plus   depositions   and
declarations   from   other   brokers,   managers,   and   former   em-­‐‑
ployees.   Some   Werner   employees   testified   that   they   always
provided   notice   when   booking   a   loss   in   another   broker’s
name.  Some  went  further  and  said  they  expected  someone  to
ask  their  permission  before  recording  an  unprofitable  trans-­‐‑
action  in  their  names.  Several  brokers  testified  that  they  (and
other  brokers)  sometimes  removed  their  names  from  unprof-­‐‑
itable  loads.  Werner  could  not  identify  an  express  policy  that
forbids  the  practice.  A  former  Werner  broker  with  20  years’
experience   submitted   a   declaration   that   while   at   Werner   he
had  sometimes  asked  tardy  carriers  to  lower  their  rates  and
that   such   action   reflected   industry   practice.   Finally,   several
current  and  former  employees  recounted  that  Lass  and  Kri-­‐‑
kava  had  directed  ethnic  slurs  at  Ortiz.
A   reasonable   juror   could   infer   that   Lass   and   Krikava
didn’t  much  like  Hispanics  (or  Jews,  though  Ortiz  is  not  Jew-­‐‑
ish)   and   tried   to   pin   heavy   losses   on   Ortiz   to   force   him   out
12   No.  15-­‐‑2574
the   door.   A   juror   also   might   infer   that,   because   of   Ortiz’s
ethnicity,  Werner’s  managers  fired  him  for  using  techniques
that   were   tolerated   when   practiced   by   other   brokers.   In   the
end   a   jury   might   not   credit   Ortiz’s   evidence   and   could   ac-­‐‑
cept  Werner’s  explanations.  But  given  the  conflict  on  materi-­‐‑
al  issues,  a  trial  is  necessary.
Ortiz   also   argues   that   the   court   erred   by   excluding   as
hearsay   his   affidavit   that   described   what   Lass   said   at   the
meeting  where  he  was  fired.  This  likely  was  not  hearsay.  See
Fed.  R.  Evid.  801(d)(2)(D).  But  it  matters  little.  Since  we  are
remanding   for   trial   Ortiz   is   free   to   call   the   attendees   of   the
meeting  to  testify,  and  they  can  share  their  recollections  sub-­‐‑
ject  to  the  rules  of  evidence.
Werner   argues   that   it   would   have   discharged   Ortiz,   re-­‐‑
gardless   of   the   falsified   records,   based   on   information   that
surfaced   after   the   discharge.   In   short,   Werner   discovered
that  Ortiz  was  using  his  work  email  to  send  and  receive  sex-­‐‑
ually  explicit  messages  and  photos.  Werner  asserts  that  this
violates   its   policy   on   Internet   use.   This   information   might
have  justified  a  discharge,  but  the  parties  agree  that  Werner
cannot  use  this  alternate  ground  to  support  action  that  Wer-­‐‑
ner   took   before   the   email   problem   came   to   light.   So   Ortiz’s
inappropriate  emails  may  reduce  his  damages,  but  they  can-­‐‑
not   absolve   Werner.   See   McKennon   v.   Nashville   Banner   Pub-­‐‑
lishing   Co.,   513   U.S.   352   (1995).   The   parties   give   conflicting
accounts   of   whether   Werner   enforced   its   policy.   But   since
they   agree   that   the   after-­‐‑acquired   evidence   cannot   retroac-­‐‑
tively  justify  Ortiz’s  discharge,  we  leave  it  to  the  jury  to  de-­‐‑
termine  whether  it  affects  Ortiz’s  damages,  if  any.
Because   this   opinion   overrules   two   lines   of   cases   in   this
circuit,  it  was  circulated  before  release  to  all  judges  in  active
No.  15-­‐‑2574   13
service.  See  Circuit  Rule  40(e).  None  of  the  judges  favored  a
hearing  en  banc.
The   judgment   is   reversed,   and   the   case   is   remanded   for
trial.