Figure 15 Figure 13 68 Average AA uto sse -Et nr A ollocati llment (With on of 2009 401(k) Fo Participants rmulae) 5.2.2.3 9 Principal Another expe                                          element” Using                                          additional of                                                                                           Gale, The included return 8 Secondly,                                          The  every While year Workers  3.2         ?deter                 8authors literature -ct References End The  end    ..Willia percent  reason the    the What  “Can “Cappi assumptions. an m Suppose “First, large      n fa  who probabili in F   199  Empiri   (th inis otes 40  the increased lack ina pe ctor  m th    A1(k) correctly fa e   rcentag 9.  ncial tic ng   to m  401(k) e  G., have  impa  ra  documenting  degr unlike ct  of    that 2012 erica    expect   Th cal t  expenses Ta or    voluntary U.S eJonat retir  ty  Group. sis  xof ct ee  don  st  -    has  in   .sensiti sp of Prefer  evalue  Afford RCS   the  on udies  re legislation e  poi                                                                                                                          to  Later of  hth o  ment success ethat  turn: an a  n     la .which ean compensation current  s    treme Specifi “N    fr o  bor t  de decreased retirement red v  Gruber, enrollme    on   rout om Tomorrow's that ity th parti  s  ein If ca   w adopting  e  markets   th  c  Ret by  the  for th  de the  o that ndous  evoluti  Data were  cye ecm sy e  ally, ipant  ss  i impact   the rement                                                                                           e m  ar rate  and semp Consumer pri  nt  alth ad  alle e,  tfro  to      enacted  wor savings m an o im  lowes modules o    will  equacy  AE Peter asset of l rare  ,n oye ough 37 m rRetir    pto update workers’     plans from k   of re a   Contri were  Th de  eperce  ct cap rth                                                                                                                                                                                                               ttr   mat  -u allocati sees: match   e  R. pend  the on inco nee e   Expend  r  such currently  for Principal  t   n  defi  to fro crisis Orszag. ure more d   buti cnt the assumption    ds regressions   Re hing  me version th federal and  m   to ndeat that   ons sults e calculation ed   o s  value RSPM throug   c  iture    qua age ns: the lowest  various  ofirms’ contribut    ben large  n  “Improving   Reveals con empl ne will   t  Preliminary  From rtile ri of tax end   of rou Survey ,ebuted  h  t  VanDerhei have   the fit  ribu     contrib ex  is in deferring oyees   modi sugg this  increments in sof     c  (p th in tto ten   cen o iPo  mo ti ons 2rc  ensio   em   less e                                    strategy most (as ng 008 rest th  fications t  wer ae dEBR  e Op  del were on u ses a   e    to on  to house                                                                                                                                                                                 Eviden tions se  a  a  reaction 401 portuni and  n)        I  retirement when shows  function was of  relationship  have d to -afrom cont ER     to  no  to workplace of  Employer (k)    re as then to F   h   has develop  de  ce  8  em longer tire ribut  olds Ret    hi  parti we the th   em ti per in   fined gto of  es  ep  previously  increase her iment of lthe rement    p ll  Baby ba should c   oye ithe marke cand as  on l ent age oye income ipa    allo   seline e   savings be Ma contri  Nove re  th  debehavior Impa    ninsecuri   In  nominal r  ttwee is Boomers eto com wed -ts tch ir                                                                                                                                                       based t    c  dbe Se abili whe eme    evol enha  when  mber b  probability to ),n .ct be curity  ”  p u   n of goals tto an  a   ensation) tion ive News ty  42 of ten nautomati tility  h   ty? great  401(k)  tnce d ded  in   (Figure e  to   sthe 2011   plan  perce choose rsome      for    than   T-171  u as the   ct       the of nt       c                          ? ? The Suspend  non-ed highly  or discontinue  compensatde d the  em  non ploye -eleect s in ive  Company  contribu tBion  were  in their  allow  pleans d to  increased  contribute  from  a maxi  2 perce mumnt     40 ® 45 12 vs. Voluntary ® Enrollment (With 2005 Formulae): 50th Percentiles do                                          docume Issu enrollment 36 67 12 26 47 health were control fund 12), concern, by and (primarily   year those  e  ath    manage  Brief asked: “liq  einsura   53% -whether  Projection National Savings accoun Room: retirement end nte med   eve wh uidi volunt   401(k) th  dan    o 2  in at nce rs ty by an  010.  d have  more   News  ts th   contin  ele  match by ary  others. indemnifi  -  or  Commission an type would e      m  Model.   Mi   med not savings  not d  enrollment alarming ent”   Release out ddle )ue                       rates,   retirement th do  ino an     -to  e(i of c -  Howe ne  ”    ndicating ation  plan an parti real longer rebalance -  EB pocket    they the   Ar d on isRI  -    Lo 401 chive, ve the repla c  co    c  Fi ipant calcul I o  wer  ssue do r, be ntribu pla scal nt   (k)  rate he  Figur  ho  ribu cnot exclude  n  -du ement Novembe   alth   Income ation. is plans sBrief, Responsi w    over at tions rin  allowed  tnecessarily emu i-  on relate 18 which g    no. ch  market  Tw d rates the   rate     shows fr  Households.” from  om the bility r  enty 263    d  to 30, last they    for eat    th em  xcontrib     income  -imply cycles.   (Employ  the penses, 2010.  two and age 20 eir a  will p  career  l  years oye average federal    67 per Ref   run   that ute w      esubject  from  e Th o cpl ww.principal.com/about/news/2010/ris   -  mus en erm is  “short” average  us (e au  e   Benefit re taxable t   .  perc Ha   gRe stochastic t401(k) of tpo .,   con lmilto -   to comm ein workers enrollme tent   of e   tax some defi  t Re  ribu with  income.  bala ag mone nasearch  endations.” ned Project, tion, e  te e     pl nces re studies who xn  penses  in yans tductions benefit    contri  cdurin   In order Ins auses   have e    an Dis xaddi  tcl analyzing itu   gHCE usively  from bc  EBR    plan  to ussio u retir do lower tion,  tin e, tions   receive ne  m  I4  Novem   nursing No eis 01(k)  n i suppose  ght ment. a  for    match 6.6 to  Paper  tes, the    be IRA    th   bno    As eecu   sr   .-    t7    R to EBRI stochastic  retire e  45.7 in tir teof in measured  (a rac  percent    2006 po 15 t tle e ive st  percent ast - ment (In)security: Ex  retirement  to Ballpark  initially).   to  by 5  perc 79.2   this preta  ent E   perc variable   $  health Muc xtimate   in and en  200 ht   had of    wh care after  8. worksheet  the   e  an   costs   100  th publ alle  fo   ado percen  ic are ur   by policy  pfactors  excluded tion  providing t match   research of  are  au  (o to    for applie Mo rmatic  the   in the nt  this  ed. stochastic    first enroll Carlo    area   amining the 3  ment simulations pe  has rcent  natu   assumed than  re of   is before. earni  of  ignore   the that ngs.   necessary  d). e m Tw  In ployees enty   -    401(k) Plan Assets Concentrated in Equity Funds      Utkus The Van These The  DEBRI analysis erhei   and estimates /ICI   Ba (200   fo 40 par1(  5) tthe   compare (2 k) . 01   dat Sena 1)abase   anal te  quite  Finan yzed  has  favorably (assuming future elig  cdetailed   ede  Comm fined   to i ittee ndividual contributio  those  hearing ibility  in  parti  Hol n is    modeled pl d a function cen an ipant   an participants d  recor   of current eligibility) th VaenD  following dserhei  (inclu  at  Vang (2din  00 scenarios: g2) u  dem a  wh rd  eo an ngrap     d th  fo ehic  und differe  infor  thance m t five ation  betwee -year  and  n    AllianceBerns 3.2.1 Coverage tein.  and “Inside  partici  thep  Minds ation in  of  empl  Plan oy Sponsors” ment-bas   Researc ed retirhe.me  2011. nt pl   ans  a 35 401(k) plan average asset allocation, percentage of total assets, selected years 48% th calcula rates; accoun participa docume replacement home perce off will change entire essence,  from  retire nt.    howev and employer tt  ion, in  match 2006 2003). that one (Employee would nte    bala  nts maki the An    at th ho  d   compared    eper currently   age the nces  ietrue -rates me in ndemnificati  ng 02. relative r,  -  no stats1   VanDerhe c    this   additional empl incentive 65 en  healt    The value at longer   nee Be  t  and  Social crucial of 13010.htm    oyee Brookings nages h  composition ded with  e  all   of ca then f  be it  i  o  re non  deferring payment).      Security and  for Resear had 25–29  n contrib  qualifi 9tax   expenses   contri attempt perc  -specific  hi  -Copeland to  de   Institution, g  cch hly by en pay ductible, u  ation  buti of normal  tions retirement   t   In  in   com   pl of   to federal pr st (at com oans   n  itu those has obabilities   assess (July after   pens least rate  retir   e t  an and 2006 e   bee quar ,   in   d  2010), July awh 6  of   unti  e  age the ted com  par any nper ment .  14.8  t  oge  ile 2011):      ltof probability have  partici is  ce employer ithe nerally  cs. e41     ipants; substantially tax  nt retirement percent age     per point  The of not,   on 2–6. p  du  ants  compe cignored    an en values  how eesti     th contri   would of to yt  con etof  -h mate  ey“success.” expected inc ni   ver, ng mute  are early sation,  vary tin bution ribu o  be  an  m    third they pi very   e  dfrom sufficient  ted boomers cke em   ad if  swhile    -mo     ne  few th pa p equa to d3While  l  aoye up eper  ed rty difications a   low  fu  have  401(k)  an  cy by  to  cr  accounts lin to l en   contribute  su     stochastic of NHCE u  accumulate Medi  the cover  tnder cces focused   53 of  plan  lowest    pay ca percent  in s  may alternative    75 is of  id).  response would   while certai    d  on the perce nature   be   This to income  at  the     for sth    least tnly be 10  udy. nt e     the -of  to     of         returns nominal contribu  (2  tion and 005   beh re -20 ala10  replaceme vior) ) for  from  singln  more et  target rates  than  are -date   60 co  ,0 investors nsidered. 00 plans   range  How (VanDerhei, ever, d from  this   3.6 Holden,  is2  to  pe  be rcent  Alo  expe n  to socted   4.6 and  5gi   Bass, perc ven the en  20t11   assum per ).  Ho yeapwever, rtions  for the  listed    5    68 1  A brief description of the  EBRI Retirement Security Project 55ion Model® (RSPM) is provided in Appendix A followed  While  However, 1.  the Suppose   one logic  nee  beh  you ds ind   to were  an  be  employer   extr no longer emely  ma   allowed catch reful  increasing  in  to  interpreting  dedu ct th  your e incentiv  th  retirement esee  r esults for an  savings   since empl oyee two  plan  other  to  co  contr ntribu  treinds bute tions  were  to  fo  ar    ? Reduced non-elective contributions in their plans increased from 1 percent in 2006 to 5 percent   2 36 7 69 1 VanDerh ? ? ?Intr The The Em oduc  im eiployer  p  (Marc average atio ct non  co h    20   42% nt th  tot 06 ributions e ).hailghest  employe  are -income  modified r contri  quar  butio in tsuc ilehn  is   arate   even manner   for more  tha  2009  timpressive,  the  was  total  6.35  match   with percent  (emplo  an in  ofcre yer  coa  plus m sep  in ensation  gover  the nme  and nt    th Andrews, Previous research  Emily. 41% “The  by EB  Gro RIw  has th  and demonstr  Distribution ated that  of  401 one(k)  of  Plans.” the most  In  Trends import an in tP  factors ensions  c  1992, ontribut  edin itegd  to  by     $1 quartile lowest the the sizeable version allowed  million   proposal employe - income  of   employee’s treated coul to num   for th  contribu d e e  ber to  s   retirement. model run   covered  quartile  modify as  of   short ta   te large re  xable is  to t  i  constructed the r  to by of  e  15 plan  m  At    ain  money this federal   epercent high  cthe n  osponsors t me  type savings   other of   within  to tax    77  to of of  the       simulate tr perc extre plan.  compensatio  accou that eatmen   10 emplo ent     m years have Th ne,  tfor   ey  "bas in t27  ee   med of  ha   after the the   perce  (jn). em d ic" ust  i  bo an year hig    p retirement th  th las  nt oye for hey   est   it define current of   retire.  arwas  -   and th final incom ose  d contri   worker income  -  wages be  average  who e nefit quar bu  have   ted.  contri are).  ade tand  ile. plan   quacy; In done   Se defin butions    is The excha cond,  8.1  aed   simula howe  cal  percent nge    contribution all for culation,   qualifi  for ver, 4te01(k) d   this  an  rates ed   d  plans  the    of     defined risk nursing -manage   hom perce in several ment ent  and  contr    trea different home itm bute  healt ents. d  ways,  15h   per care  allc ent  of costs  the  and  is  models   only not approp  0.1  ide  perc n49triat ify en  ely at t contributed  least  modeled.  a signifi  Th  atcant is  the  is  especially  perce 402(g) ntage  limit.  tru  of  eTh   the for e   those  40% and beca above  95 use  (es   percentiles of p ecially strict confidentiali  the  with  lack  a  of  me  jotyab n  standar tur  of n3.over 93d pe s,  and  no rcent.   there information   Amo foren  the g  participant on  suppression  the plan -dir  sponsor’s  eofcted  cas hfuouts  identity nds,  prior  five  was - year to retireme  in  retur cludendsnt). .   range   d from  by a chronology of its development and utilization in Appendix B.    Re tire ment Sav ings Deficit Year 401(k) Raish, These         Helman, taking The                    authors ..      concl David place “Kansas “Retir plan federal  Ru th u appears    an at ement L. sions ,    Craig 37% “Cash athe  Futur inco lyz   conflict esame  In  me   uncontroversial, Copeland, ethe  or c  Retir o   tax me  Deferred  da ti me. with  ta  epu Ade men  with rp   First,  and previous quacy: oses t  Arrangements   In OL  Ja  the c   and the o Sck   me Alternative regression   overall VanDerhei.   EBRI  analysis that  Assessment   research, anything  allocation  ( P  be  and ortfolio Th  “T comes resh  h  find Project.”  your e  2012 which old  to   358).”  that more   semployer comp  and  Re   ea  surv A   complex Bloomberg/BNA t   ch projec any ith remen eyed  e$0.10    Im stock contributed tp   tdefin  of with ortan   Confidence increase   de the  ecreased respect c  dEBRI e   benefit  ($400 of  to  in   Futur  Ed  your  the   Su substantially to ). ucati  rve   plan  th mat  eretir  eyEli o  :lev   ncsp Job gibili e h  and ment  eorate ln   of sty   o    r  in s   in 2008.  th match) remains constant.  69 th th probability 5.46 percent  of success  of compe  fromn sation 27.0 per  in c2005. ent to    The 64.0  in  percent. crease of    0.89 percentage points once more  4.2 retirement Are John  ther  income  Turner e ot her ad  and equa  ways  Daniel cy  that  for  Bell   the individu er,  Bo  pp. omers als  149–1  have  and 76   responded Gen  (Washing  Xers iston,   to eligibility  re  DC: cen  U.S t  mark to.  Go parti et vernme  cco ipate ndit nt  in ion  Printing  employment s?      - Mr. Chairman and members of the committee, thank you for your invitation to testify today on  compared plans alternative return desiring population in 75  exchange  per Senate Committee on Banking, Housing & Urban Affairs  in  are  c  modification employer average place aen    hig with e   til ve as  xfor plained h e,  rfailing    scertain thr probabili   i39 an  odeferral n  eand   spercent 18 shold   of   to in of  ly  per  employee the     mo meet the since ty   rate in c  ent of  re model who c  current r   e  a  that de the was  a  match successful se  tail have  contributions    advent s  allow criteri 5.4   to  in tax   fro   no  16 per V  incentiv asimilar at, m . nDerhei 0  of .c  retir   thi   ent. the per Si  the nce nk e  c      federal analysis would   ment. ent. es, 401(k) they   the and  assume  Cash   genesis need    Copeland Fig  be  governme plan  for  u  eli balance  re  to  replacement  th gin   ible 18  of save e  the   U.S.   the sho (2  nt, for   plans  010), earl less   w RSP  govern  by as y   fla  than that   plan  M bu 1980s. have  rates, t -project trate ment     the they $25 size  a       s  median refu For value t0,000    and andard would are  in  nda these   the  based age   of for   ble inde match   -defe late -of   specific retirement. employers,  -  ta on liv mnifica  1990s, rring xing    acredi 18%     salary    t  of tthe ion ,    the         -0.02  Van  per Derhei cent  (Se  to p8.0 tember 9 perce  20n06 t per )   year for the 5  and 95  percentiles th  with a mean of 3.76 percent.  37 48 1999 2002 2007 2008 2009 2010 While contrib during The administered raises  applica   NHCE knowing  this uDefined Research www.bna.com/Cash Insecuri savings tions  pe t iparticipa o   riod. by the n   th this  ty, of  Contribution Mer   eFemploye      this  und Debt perc In  ye ct  the er ion  ar mo  and en   We  Hu on   rates 1del tag e  999 m  th  igh your will - Deferred  aeis e Re n       of EBRI/ICI on by Mil ill   tir make. Resour  beh  ustrated house   around e Retire bank men a-lf  Arrange Th  ce   h database, was Me tmen  olds is  Plans.”   1Consulti in   mo may percentage   also Fig t  that  Cments rial u o  happ   re nfidence, EBRI treated    almost ng Fun will  8  -to of en  p7555/  Notes, d, be    poin  gau VanDerhei  for  Ju   as 1/5 at  ly Sa    gpart two risk t.   e16,   no. vings,   However,  thei of     for  reasons: of 2002. 4  all  r   (Emplo  your  and  Marc inade  recent 401(k)   Copeland   tax for h  quate yee   2012,  aamoney  thi cble tivi  Benefit s   re income. group,  ty EBRI  (2001 t  iand (19 re m  Researc Is   per  plan ethe )sue   nwith Su tc   eincome ppose participa Brief ne nt)  h an d   In was   #36  st th  is itu  et2  ion   te,     Ev The en  baseline  for thos  reesults  plans  in without  Figure snonelec  17 andt  18 ive  were contributio  simulated ns, we  assuming  found several  the mi dpoint participants  value  with for each  employer  catego   ry in the  2 It should be noted that the baseline assumptions th41  used in the 2010 analysis did not allow for the utilization of net  RSPM was suggests  significantly  that those  6enhan  large ced  4for 01(k)  the  sponsors  May 2008  adopting  EBRI poli  AE cy  were forum  mor  by eallowing  generous  automa  to thetic  401(k)  enrollment     based retir Office,1992). ement plans.     VanDerhei (August 2011) provides information on how the relative value of the  contrib A quartiles retirement calcula stochastic primary  comm tuions,  whatever given on de tion  for  pr   cis concern security    ocess workers and rate  ion to th    in other  of wa  with e    many  in  with 2.7 e  scurr  m  Ame contrib   ad  ployee. ape   respect  ently mean cases  hoc rcent, ricaute    thresholds. ag .    Thi is  eI   with  qui to d  eam not  srto  d,  plan ty 26   Ja    awh athe  return –35.   cretirement 75  ksponsors e    VanDerhei, cre th  per e   d For rof  itto c   8.9 would  en  all retain  suspendi  savings til   perce four e  research   thr be  both  income nt  e  de pla shold ng  and  pforms  n. their o  dir sited    Wha  aquartiles, of e  mean  ctor   contri 4.5 of  dire t  do retirement   per of  fictl  bution you  xthe  the ceyen d  into   -believe Employee  income taverage  using s   is the plan  the    retirement the wou  ,return    per bu pote  Benefit  cur tld c  the en  ntial be  rof ent tage   relative  the  6.3    saving impact    most       70 13 retirement model ? ?  ha The Alld   plan always age  highly   sponsors (even   assumed co mpensate as  drop much  a  th  retirement  as ed  pla   e20 m n ployees ye  match, ar s) age ,  as and  in  of  those  Company   all 65.  employees  While  with  Bat  there  we  least  rercee   was allowed i van e a   80 abundan  30  per  per  toc  ccontrib eent nt t evide   match proba ute n ce b fro  aility   of maxim m    the many of success   u m of  10  In an analysis of Vanguard participants, Utkus and Young (2011) find that the percentage of participants  Two important invested modifications incentives example contrib AllianceBer   Van Van -thirds D Duerhei erhei, tions April governmen   in of   nstein for are    of company  the and   Ho th  2011):     retirees publ to statisti a  lden,  co t  Copeland  survey.  their were mpu ict    10 Alonso   policy matched cally stock.    not defin (64 t-  ati Sensitivity 19.  (2    eoperce insignifica    qual 00 an   n analysis, ed By  of 8) d   18% ben    Bass . 2010, to  the nt)  analysis  the e     of indi fit (2 pro  nt perhaps    pre 01 this  contri (p  betwee cate b1). ensio ability  dof  icted value    th buti they  equally isn)   n  amount that  assumption had  o  and  3did ns  and    aso decreased    some defin im  22    based 6that p  -perc ort year e   istype  dfo an   shown  on en contri -rold t    every to   the tis of  of    knowing only employe    financial plan’s bution in  pay,   $100 Fig  8 u  and  per matching res  (401(k)   eyou just  plan   c  3with tur en  and   or how t. nning       -   one ne your 4 formulas typ Mo  of  gative large  for e)   reo year V  em a   plans. re nD v  the er,  p   tand fo of erhei ilre oye  r   during ten   m  matches The the  re(Mar  unt 70   re survey .   tha ch with   t    When housin gviewed  equity  to in  ens isolation, ure retir  itement  is clear  income  that  the adequacy.  impac  tA  of  fu  increasing ture publicati  thoen  limi  willt  include  on employee  a 2012 u  con pdatte ri butions for this  is  of 401(k)participa  participants nts when  with  measur the potential ed by  th foris  automatic variable than  escalation  before.  of contributions to be included.    Salisbury, Holden, defined           . “ M  be Sara  assachusetts Dallas, neh, fit  and  accruals  and  Jac  Eli k  Fu  z  Vim ature abeth nDerhei. pact  Ret  Buser.  retirement irement  “C   ontribution “Many  In  income co  401 me  (k) B As  adequa e sessment havior Sponsors cy.  of    Fi 401(k) Suspendi Project.” gure  8Plan  cat  ng A  egorizes proje  Participants.” Matcchin t of  any g  the Contributions  positive   EB EBRI RI  Education Issu  value e  Brief, Also  for       on Research  employe ? While government.  Institute e savings.  a large . EBRI   r  For  match  is  ex a am nonpartisa SUBC  rate ple  will , OMMI if an  provide n  employee research TTEE ON  a larger  institute  were E  fina CON  contri nci  that OMIC albu   ince has ting  POLICY ntive been  6 percent   for focus   the ing  of  emplo   on compe  retirement yee nsation  to contri   and to bute     Jack individuals reduction interest decreases financial perce  Va nt  likely accoun cre (expressed  nDe value   for considerably  retiring dir   tplan chang hei, s.  of t,  Th   as each    sponsors ese Research earlier   e in opposed  to  nom   values in   wh your  terms  (e.g., ienal  n in    to plan?   increase the  Di  the ter    of as threct   impact ms).   efutur two soon   curr o   to rse  ,ent m as     accruals Em of 3.1 allest  they   de st  pl per odoy c   uction, beca plan hc aen eor sett    ime contrib si  cBene and  zhealt  which e  eli  categories  5.gfit 2 h ible u  tions   per care simply Research  foc ent r.costs      Social W (less  r eifhsults  ile  the are   than  Inst  Se this    cash ein cxu  it   $1 cluded amay rit ute   lower balance  ymillion  retirement   not .  For tax  be   and plans    th payme considered e  $1–$10   lowest be  are nefits n t than     to  at    percent pretax and had a 100 percent match for the first 3 percent of earnings. Fifteen percent  Apte, stopping employee’s 2012)  for Vishal,   the contributions  be  minimum  and fore -Brenda tax or  reduction   was afnt eMcFarlan r  2.8 -tax   in  perc co  antribu cco en d.u   tn“A  tions, in t balan   Look 2005  or c es, at    bot and   Defined and h .then  This  in Figures   may decrease  Con  be tribut  5  be  andcause dion  to  6 V  olunt  of  2.5 Ma   of the ary E   per tch th   nrollment same e  cRe 40 ent 1( inst  publication  k) in  ate plan  2006 ments.”  using  an  for d  a  2.4    the different Towers       14 71 accu period Thirty above The the also analysis   current  version mulate was -  four 6the  with  contrib per  able   pe  perc  d  cem of housing  ent  defici rcentag to  the puent   ldetermine ted of oye  mode  ts   pay. of equit  to r  eare   and the   of your  l y   linewly used  s  used kely wages e  what,   retirement retirees    to for “a -hired   sof  be. if    nee th   any, $15,000  e  say Fi   d(those an gure ed.”  savings   increases they aly   s  3  who iwith  s  be provid  in  plan gan   this already   tw in   es to this o  em testi   or  plan informatio p   year  fewer lcontri m oye  ony 20 ,r  the   contributions years  bu  years as  nted govern sumes  on  or   of 4  the    percent more tenure  ment all  average  wo  to  before  )would  rkers   defin of 401(k)   compe ind   retire they e  contribute dividua  par  contri  nretired  tsation at icipants l  age retirement bution    $1  an will 65  8. in d   an     27 ad       much greater than any19%  of the other three factors.  However, the importance of including one or more  18% Additional   It Copeland  should  be  mod an  noted d Van ificat D  th erhei ion at 5som s  were (201 e plan 0)  added .   sponsor  in 2009 s may  for  have  a Pension  turned to  Rese  theirarch  401 (Council k) plans  presentation as a means of  freei thatng  in  up vol  cash ved       a   define      . “Aand (at Fundi no. d  Post ben   le  238 Researc ast ng e -Crisis fit    (Employee Defined within  accrual  hAssessment  Fun  a   Benefit spe in d  tBenefit and oc ified quar  the   Pension of   trange),  Researc  iles Re Mitilbank  for remen  Plans.”   the each h  Memorial In  tem st   In in  itEBRI co put clooye me em,  Notes, e  Oc Fu  e  Ad grou  may tnd oequa ber ,  p. Vol.  December have   Th 2001). cy  30,  efor   alargest   certai  no. Baby     1,  6    2002. n(Employee reduction Boomers  targe   t in   and  in  mi Be  at nd  nGe -erisk  fwith n it  Xers.” Resear  ratings  resp  EBRI ch ec   t      be assumed million receive health  an optimal   ben  in the otherwise.”   to assets) efi   max credit ts  choi  for im  are  u in ce  the  m t   more efor  ma re  pas  st so 16% tch   tat than m  34   efrom 5.8   sponsors, years.  1.5  perce  a  tim plan  EB nes RI t.   after with   the  does   value  arecognizing 17%   not 50 per   of take  the cent  policy   average certain  match  po  sitions legal  on per  th c  en and/or e  and first tage  does  6   financial reduction  perc  not ent  lobby   cons of  for    .traints, plans      such  preretire age 62), of the m  all ent  model  high  income ly  was  compensated  qua  constructed rtile, the  participa  value  to measure  ofnts  deferral  contributed  the  households (in ter ms 30 3 perce  of  probability  perc nt of ent 16%  pa agye   of while  of  re  additional ti r10 em  pe enrcent t  income households       15% definition maximum  of reduction  compen  in sation  acco  th unt an  balan  that cco es. nta  Thined e 15% average  in 15%  the  dat perabase, centage  and  redu  wect  attempte ions in acco d to unt  con  value trol  for in Figur  this unk e 3 vary nown  from     perce ? nt Watson, All in  20 plan 07  sponsors  before  U.S.—Insider,   increasing drop th eOc  pla  tob slightl n match, er 2011. y to  and  3.1   w  all  perce ww.towerswatson.com/united  employees nt in 2008.  recei  vTh e an e  value 18 perc  de encr -t eased match  to fro  2.9 m the  perce   nt in  perce 3contrib immed income plan plans 5.2.1.1.3.1  offering  nt were iately ute   defi What report    an mad ci   Age begin company ts   do additional   be by e   you and in ginni   age to  c  o  think Salary withdraw  n   ?ng coh stock j unc pe  to oNo  you rcent.  rt,  tplan  ias on  chan  as    wo an money  wi    bet well Th  uld inves tge hisw    re   as be va een  from ductions t lue ment family  most  10  is  their  and   es  option  listatus kely  timated to   19 in  their dividual   to years   that and  do?   def to   hel gend  before  be  ianed    dcco  as  company er,   benefit unts low   retirement. for   as (defi baby   81  plans. sto ned   per boo ck  Howe     cdecreased contri m ent ers  ifve  bution  and this r, 17    Ge fr is  om  perc the and n Xers.  61.0   last ent cash     say     additional This information  factors,  was  along  the  with n combin  the incr ed with ease  th ine  the  defin  limit ed  on ben  employe efit information e contrib  for utions  the,  same can more  spons  than or in  double  an    12% 13% 20 72 flow that is required for their legally required minimum contributions to12%  defined benefit plans. A review by  a  This winners/losers  number is somewhat  analysis  sof m aller define  than d ben  the efit $4.6  freezes  trillion  re an pdo rthe ted   in enhan  VanD 11% cerh ed eemployer i (October  contri 2010);bution  however, s to  the     Contents The betwee  authors nInstitu Issu   thee  conclude   highest Brief, te, Jun  no.  eand  that   2009). 354  lowe  the  (Employee 11%    st in centive income  Benefit  -effects specific  Research  of  defin  employer ed 11%  Institute,  benefi  match t value  Febr ing 11% Au   uary contrib quartile toma 11% tic  En 2011). rou llme sti  takes ons nt   are  pla  qui ce te for  small.  the    as compensatio sponsors  the into ability  in  the  any n , total  to would  of  termina  the  (employee   alarger  suspended te an -size   and underf  categ   ma employer) utch onded ries.  en    dpension   contribution up decreasing  plan (wi  that  the th  needs th  tot e aexception l  contri to be bution ma  ofde  10% certai  ea  for ch  the n  year  sponsors  em  top lsatisfy oye   e    that adequacy  will mee  if this t the  te mpta threshol tiond  were by 10%  deferring  avoided  retirement  and retirement  age from  defe  65 rred  to  to  84)  age  dec  65. reases  However,  from 16.0  even  perce  withnt  this  to    VanDerhei contrib  (200u9) ted .   10 percent and 15 percent contributed at the 402(g) limit. The average deferral rate  3.1 effect  to  19 by .7 com  perpcuting ent (dependi  the differ nge nce on income  betwee  quartile) n 9% actual  and for  401 pre(dkicted ) participants  employer  curre  contrib ntlyu  26 tions –35  (as  un  ader  perce  thent  minimum age of    Post government. -2009 401(k)  9% 2009                           and ..    “Can “Oregon states/newsletters/insider  wa  4s01(k)  estimate  Future  Accumulations  dRet  toiremen  be only  tGe /5641   In 2.0 ner co  perce m   ae te  Assessment  Significa nt in 2010. nt In   Project.” come 8%  for  A  Future project  Retirees?”  of the EBR  EBRI I Education  Issue Brief,  and  no.  interval they These balance Regarding perce Copeland  started nt  num    of to plans  and    th compensation 33 be  e plannin .0 rs ,   VanDerhei proposed as   per are  well  cpresent gen   five as t.     tax IRAs)   (2010) that  to   values cre  nine   is whe dit,   matche simula  years   n at Gal ever  retirement eted  before  d(2011  the  by  the   the sum )   im retirement repo   em age  pofapl  ct rts their  and oye  of  estima   such  rrepresent basic   and (i.e .,tes freez  15   expenses the   from perce  e  addi the s on  the  nt additional tional   ex and  started  Tax pec  uni  match  tPolicy ensured d   less amount  future 71   is Center  than  eq  medi  pension ual  ea  fi ve for  ch cal to   years    zero both   ).   an  In   attem the impt pa  to ct  of an aincreasing lyze whet he ther  EBR limitI’s  by  2007 8%  itse 7%lf.  findin   gs of the association between defined benefit 7%   Salisbury and Buser (2009) of 251 401(k) plan sponsors that have suspended matching contributions for their  The baseline  testi  assumptions mony draws  use  ond  th in ethe  extensive  2010 analysis  rese ar did ch  not  cond  alloucte w ford  the by  EB utili RIzation  on these  of net  top  housing ics over  equit 6%  they   last to ensure  13 years     defined contribution plans 4  provided at the time the defined benefit plan was frozen.   5% Although lowest Summari Figure "-17 Aincome a. ccum zing   the shows ulat Stop     association their ions"  quartile.   the c as a on  empirical  baseli tribut ?  For  between ne in Termina  these g   results average  altogether   house the te  as   our perc   adoption foh   llows plan olds, entag    the (Mitchell,  eof  re   at au ductions -tomatic 5% risk  Utkus  ratings   in enrol   an 4 01(k) dr dlment  op Young,   36 account   and per  2007): c  em en  balances tage p  loye  poi r  contributions at nts,  Soc  friom al   82  to  1  Introdu theirc  fi tion nancia  ............................................................................................................................... l planning objectives. For example, if the employee has determined that he  4% ........... or she  2  satisfying from 9 per  th ceent  bankru  to 6 perc ptcyen  conditions t, or would  necessary  the reduced  to trigger  incentive  pension  drive  in  this suran  down ce cov  beleorage w 6  percent by the Pension  (perhaps     admittedly 3.8 compensa reduc  per tion cwas ent  tion) sc enario. op   by 5.9  tand  iexcludi m  per i    sexcluding Fig tcicent. u  assumption, ng re   5these  shows  any59   partici costs.  that  the  p  the Th ant  results e y   vary with highest     from in more  both  pr  8.  th eretirement 8  an 2003  to  a24  0.2%  .4 and  per  of  20  cin  co en 10 cmpensation o t  m (de showed ep  quarti endin  that g  le dif  on  experience ferential.   th income e 4% med 4%   quar ian s a  tile) addi  similar  for tional       73 Soto,               .M  Testim aResearch 251 uric  (Employe io, ony.  and  F  U.S. und  Barbara e   Congress.  and Benefit  th  A.e   Butrica. Research Mil   Se bank nate  “Will  Me   In Health, stmo  itu Automatic rial te  Educ ,  Fun November at d,  ion, Enrollme  2001a.  Labor  2002).   n tand  Re     duce Pensions  Empl  Committee. oyer Contributions  The Wobbly  to    contrast, before expenses individual wealth 18    perce   tha for nt   the  exceed in  t ne  cr point  that  edit same w employees.     grou the (Figur an  employee d  after  p a  e30 wo  7) -  perce uld tax  .Looking    is  need   esti annual nt credi mated   at to  this have  inc t.  Th  to oportion  m accu  ehave e  paper  from mula   aof  90   in Soc  pension ted  cperce lud ial  at  eSe  sage nt   curity awea   dpr  65 istributional obability lth   to and  provides  elimi  define  of nate   ana contributing  one d  their benefit ly sesti is  ex of mate  p  plans the ec  the t  ed of winners   (if  ex  the defi  any). tr acits   and  If       freezing/ approximately  VanMD uerhei ltiple of closing  (A Final  4.4 pril  Ear and  million  20 nings 10  enhanced ).  workers  401 finds(k)  th  contribu at those emplo tions we yingr e50  corroborated.  percent of the    workers Figure 1  also  of VanDerhei  maintained  (A anpril  ope   n  In retirement  later EBRI  income  analysis  ade  (VanDe quacyrhei, .  When  Novem  the b201 er 20 2 analysis 11),Hearing  the  isfol  relowing pe on: ate dscenarios  with the  were same  added:  assumptions    as used in 2010, the  with its Retirement Security Projection Model® as well as annual analysis of tens of millions of individual  Bassett, William, Michael Fleming, and Anthony Rodrigues. “How Workers use 401(k) Plans: The  During Securi 401(k) percent ty   plans the   needs to nor   46 same m  was  al  percent. to  retir    time not save  ethe   ment th a  Households  total er  foc e  age was u  of s of   du 9  a   th perc  substantial e  in  eto   study, th ent  expected e  second of  one compensation,  change -  mo  thir income difications d  in of  the   the quartile   amount the  defin  of  require   plan drop ed  of ben  sponsors   25 money d eemployee  fit per  plc eheld an  ntage and  sponsors   in  co participants   balan points ntribu  surveyed c tion ed (fro  f  unds, m in wo   an uld     at   - Benefit 40 to 1(k zero)? ) participants  Guaranty      currentl Corporation, y 26–35  or  un  PBGC der th ) and e maximum  the im posi reduc tion tion  of  sc  ae n20 ar ipercent o   or 50 percent excise tax on  49 decrease, percentage  fro  ofm  comp  12.8 ensati perceon nt  to that  2.6  would  percent.  be  required for retirement income adequacy at more than a 60  50  ? Company C had a maximum pretax contribution of 16 percent of earnings and a match rate of  5.2 What incentives have the greatest bearing on the behavior of employers and employees  With respect Stool: b. 401(k) The  Redu empiri  to  Re  Plans?”  plan tire cecal  th ment  loans,  model e   ?CRR   am  (IRedu   ount n) in Working  implie  sec the ce  you u  15 rity  our s    Paper, contribute that yea  in  av  America re  sclose r  that a  no ge.    employer to  2009–33, the  (T  65  -EB 166).  percent RI/ICI    match December 7 Oc  da  of ttaba .   2010. NHCEs se  2009.   has  at  the bee    n ty  trac picalking  firm  loan  would  activity  join   perce 2 there impact in losers defined    See retir What   nt Van  isunder e   on be   ment sufficient ofDnefi    overall erhe co is  the the tmpensation  (whi  plan. i (July   two size   money retirement ch  An  20    coul of versions  addi 07  Am  ) d to .if   tional   be e  they pay r iwealth   cof aan     relatively  16 expens  would the s’  perce  retirement   proposal; bu es tforfeit n  it  t short  without  of is  inco workers   th  however, savings  perio empl   option ta pping were d ete,   or gap  the    since to  coul with  ?into   ......................................................................  earn unde d   employers   ma th las  an erny tlying   tax de  extra  sponsors cades -qualifie  analysis  tha  1 tper ) .were  Th  deither c  ent holds ind e  still values ividual  of   in obligated   retirement em cr e for p  aalse oye ccoun  those  employer  to r  match  fund saving ts,  on   a   2      74 aggregate deficit increases to $4.8 trillion.  b,c A 2010)  new  de subroutine monstrates  was  tha  added t the  average to the b mod  improvements el to allow simulations  for all thre eof  me  various trics wer  styles e mu  ofch  target  high-er dat  for e funds  sponsors  for   401(k) part Eic quiip ty an Fundtss dating C  oback mpany  Stin oc  some k  cases Balanced  as Fund  far s  as 1996. Bond   Funds GICs and Other Money Funds                         ..  “The “AParticipation,  Behavioral  Influence  Mo   of Contribu  A de utomati l fortion  Predicting c ,Enr  ando llment, Wi  Employee thdra  Catch wal  Decisions.” Contributions -Up, and IRA  Natio  Contribu  to 401(k) nal Tax tio  Plans  ns Journal  on.”  401(k)  North , Vol.   Accumula Ame 51, No. rica  2, n tion  pp. s  3 reaction indica risk increasing 5.2.2  Van  rating To Dted erhei  what Impact  to  tha   of from  the  and  58 t  exte  they   proposal    of Copeland 7 perce   perce nt Em  had  has ployer nt 3  nt already    for  to this (2  in  01  modify those   Matches 1999  0) de.  increase ficit   in to  the   be  the 18  on  fe en   perce dlowest  d401(k)   im or er apacted plan lnt  ta  defined   Saving in xned   tr20  eby  a10. to tmen   ec   ben incr   Looking onomic teease  fit of  emp value  th   co ateir  ndit lrecently oye  quartile  employer iron  and s  over hir   to wor ed   match 33  kth  401(k) e  per er  contri pa  cto en st   par a  tsever   bdefin for uticipa tions  those aeld  nts   for  in       38 be 6 percent if the employer matched 50 percent up to13  6 percent of compensation but only 4.5  the r ?ecoup No m plaent n spons  of excess ors drop  assets  the  pl inan  the  match,  case  and of a  all  reversion,  employees  the  receive  best  an available  18 percen  chtoice  match  may  from  be  the to gradually     percent 2/3 probability  for the  fir would st 6 per  exce cen edt  of 25  earnings. percent of  A  co total mp  of ensation  30 perce  annt nually  of pa  (un rticipants til age 65)  con  for tributed  many  6 percent  72 Stable-Value Funds among frozen  Under  define  401(k)  their the bas d  401  benefit pla eline (k) n partici   assumpti plan  plan .regardles  pFurther, ants, ons,  th th  8eere s   perc  avera of  has the engt  e  been  prese of  per  the c en littl n  workers ce tage e  of  variation.   reduc a  match. were tion  with   in Plan  From   emp an  partici emplo  l1996 oyeey  p contr  er thr ation  th oaiugh butions t  would had  2008,  bot  fo  be hr   on  an thi  estimated   saverage, ope  grou n an p in d    a  50 In   an attempt to provide preliminary ®  evidence with respect to® the impact of suspending employer  the DC the if contrib  they    contrib verge excess  continued utions  of  uisti   re “Retirement (In)security assumed ons  constan tir e  or to m   set econ nt  to t   for  up t(Early  ribe bute   both new  investe  Boomers)   to  DC em  MAX  pla pdl oye inns M  a   vary r :coin  ATCH. non Examining the Retirement Savings Deficit” s and  ciden -from tax   par In- aea  ttapproximately d  icipants wi vantaged cht hof  the  the  (page  pensi   thr account e   $22,000 6). e o nillustrated   Gale freeze  wh e  men (per r.e  Fa  the  intervals,  tctoring individual ions  investment  that  in  the   )th the  fo e   model r  sproposal   eincome married      is      4athat  comparison  had frozen/closed  with parti  th cipant eir defi -dire ndned cte  ben d ineve fitst  plans ment  than s in 2009.  for the  In ov  April erall  average. 2010, the    For model  exa was mple,  completely  the change     8 5.2.1 Tax incentives      5 The   The        Principal What  .perce  “Retir c. Actuaria at 263   Re nare -tag Continue 288 ement  tirem Financial  eth   of l(1998). e  Jo    emost de  Savings u nrtfined  to nal .”  Gro       contribute effecti EBRI  (2001  be up.  Shortfalls nefit  Issue  “T v).e   h pla  an e   Brief, what  nPrincipal d   for sponsors  effici    no. Today’s you ent  28   do Finan  that  3 Inves str  no   (Employee Workers.”  aindicated tcw mtegies ient a  l Categor Gro  to up   th   yEBRI  Benefit encour a  Retirement t the  Notes, y age  Resea had   no. an already  rReadiness dch  10  f aIn  cilitate (Employee  increased stitut eSurvey  greater , July  or  Be   pla 2005  2011.” n esa nned fivi t).   n   to gs   75 ? 1–24%  contrib the 401(k) provides  These  high years?   plans findings uperce government. est tion  a more   defin   in plan, nt   are exchange   iffocused e   dthe part and   ben   employer  of 20.9 efi    the for method t  value percent   22 an  18 mat  annual   quarti of  per   indicated canalyzing hce  ent Retiremen le) d 100   w match h   ile that perc the  thos t   fro  ch Co ent thange nfidence at m e  up    in th the    the ey to in  federal    (at parti had   Su third  rvey least)  already cipant’s   and governme  (RC  4.5  Shi) , increased  ghest  perc investment a survey nt, ent  in  by c   of that o   or m age  compensation   echoic  pla gau  quarti  and nne ges e sag  d  th le without  eto  edrop - specific views   .   24   a    reduce 2.1   theHow  relative  can  this value  gap  of  be  the  best  defi  measured ned benefit  - ar  plan e th  in ere  th  specific e future  met  by th rices  im that po  should sition of  be  a  considered? pension freeze  ..... 2   age/income combinations. 61   Today’s of testi  pay m ony to the  will  plan  deal  and  with  7  pe thercent  following  contribu  topitecs: d  16 percent of pay while 12 percent contributed up   VanDerhei (September 2010).  frozen response  define  to the d benefit  proposal  pla  isn  14 tha.3t  carried percen t.fu  Ac ndcou ingnt  obligations.  balances will    also be reduced due to the plan-sponsor reaction.  less It is  understo thanto  one  rise od -fifth  over  that  of  a   401(k)  401 narrow (k)  pla participa  range ns differ , by nts   five from  with  to   traditio a  15 ccess  percentag  nal to loans  employm e  points, had loan ent  respon -based s outstan ding  defined d ing. to a   benefi At range  year  tof  -pension end  ma tch 2009,      taxed predicts households, “could enhan contrib  as c  conceivably ued   tions that ordi  DC  in  nary  those  contri on cre  em  ainco  si  affect with bution p nlgme. oye  to  the   in e$34,000 s    cbehavior, (if  lowest The ent  any), i vindividual e  for s   es  we for  Va  tisingle  mated  estimate firms nDerhei  accoun  males  to  proba   offer  (No thts  eand   bvar net  ility ember 401(k)s e  $65,000   tracked loss  of  contr  2009) that  or  for  pe un  ifu buti  nsions”  analyzed til single tu  the rng e employees   the  point females.  (page  all  extra   4at  01(k) 7)  which   per Even   but may  plans c ent concludes    though th ex ey pof  in e  ri are  the e  the n  c   tha e  ist   37 re in -the parameter  total em ized ploye  with r contribution  401(k) plan -rate desig  for n parameters  all frozen plans  for sponsors  was 1.64  that  percent  have  of  adopted  compensatio  automna  versus tic-   increase their employer match and/or nonmatching employer contribution to a defined contribution plan varied  and attitudes for www.principal.com/about/news Research retire  ofmen  working  Institute, t?  -age and  Oct  retired ober 2010):  Americans /documents/2011retirement  2-9.   regarding retirement, th-eir re adiness preparations -sum  for mary.pdf  retirement,     their  confoundin salary and increase   ?21  quartil  perc No  their  pla gen  es. im  nnonmatchi ta  spons pga  eTh ct  points,  eof o rs avera  the 2 drop n g  respectively change  ge employer  the  per  plc  an in en   match, ma  tag contributions . rket e redu  and  values ctions  all employees   on to  for   acurrent  defi  the ned  receive youn  account  contribution gest  a 30  cohort  balance  percent   plan (those s  match .  VanDerhe . Th  currently  ere from  wa  the i,s  Holden,  some  26–35)      Thus, for some employees, a higher match rate may result in a lower employee contribution  16 to the 402(g) maximum for that year. The average deferral percentage for Company A was 6.3  76 Two    major proposals have 73  recently    emerged that could have an impact on employment-based  VanDerhei, 15 39 Ippolito, Beshears, d.  Ric   John,  Ja hIncrease ard ck,  James   and A. “Pension   the Lori  J.  Choi, amount  Luc aPlans  s. David  “The  you  and  Laibson,   contribute Imp  Employee act  of and  Auto     Brigi Performance.” -enrollment tte C. Madria   (Chicago: ann. d   Au “Ttomati he  Univ  Impa cersity  Contribu ct of  of Em  Chicago ploye tion Escalation r  Press,    the plans 3.1  percenta ?  in Impact  offerings, What that ge employees   is of   of the  the   from participants  size  financial  a   of are  modest  Am  per  an  e ?who rm id  c($0.25 itted  an housing 25–49%  wer s’  retirement to e  per   offered make   market  dollar  voluntary  loa   savings  on crisis n sth  with e  in    firs pre gap  2  loans 008 t-?  tax three    and  contributions.  outstandin  perce  2009n  on t of g  re   tic pay)  tirem Hence, ked  to  up e  ant  th   to very  readine e  21 se  gen  nsitivity percent essrous      of     small compensatio VanDerhei EBRI/ICI deple this present  seems  toverall, ed;   values 401  if   unlikely. (M  the (k) n  amounti arch   when  are  database Social   2006) defin  He  the  n Se  also ged    curity  awith analyzes to dditional  in   adismisses   consta  0.5 more  and -2  the    defin percen match  than nt   as f  dollars, inancial  elikely  $100,000 d  tis  ag benefi  set e   oversta   the cons point  equa t   Ret in paym equen  l reduction  employer tito rement ed  en zero  the ces ts    are con (yo of  Savings   in contributions  au   cnot  re ng pension erpn   lem sufficient ac   tha Shortfalls em ptl oye  freeze the ent  in e    ta rat sto    2007, (RSS) and x   for epay  cre s.   those the Some  dit/  basic for  and  ge  m both   with  atc neral e  expenses, non mhes  ployees, ge  e lower    nders  in calle     d      2.2  Do individuals understand how to calculate how much th 38ey will need for retirement? ............. 3  enrollment 0.89 VanDerh  percent ei  provisions. (O  for ctob  the er  20 overall 10a  A ).completely  average.  Employers  updated version  that had  of  closed the nation  their al  def model ined  was  benefit  produce  plansd  to for  new  the  May  from   Twelve For  62  example  per  hun cedr n, taed  for  40  40  -those year 1(k)-  ol pla  that dn  participant   sponsors had froze  were  nwould  the  randomly defi  nened ed to  ben   selected have efit  a  plan  tenure  to  in comple  th  ofe  at last te  least   two the  10   40 years  yea 1(k) r to sQ  with  u81 estionnaire  perc  theent  current  fo  vir athose   a secure     confidence with regard to various aspects of retirement, March  28, and20  related 12  issues. The survey was conducted in January  government.   overlap Alonso are As a  largest  result,  and  betw   fo  th Baeen re  sthose  s2011  (2011)  the   in  version two  the  find  groups,  lowest  tha  of RSPM t,  in -but income  1999,   o add verall,  e  31.3 quartile d a  42.5   new per  cperc  (22.2 feature ent ent of  percent).  recently   of that  the  would  defined  hired  Th eallow   reductions 401(k)  be ne hou  fit participa s eholds plan  for  spons the  nts to  youngest   defer held ors      rate.  17 retirement perce  plan nt.  designs,    specifically 401(k) plans:       U.S. 5.1           Chamber . Automatic “Retir on 1997). Matching  Reement  tireme of    Commerce.   enrollm on  Inn  Savings ct oInme come ent  Ade   2008  Plan  Ade quacy   Participation Empl quacy.”  for oyee  Today’s  EBRI  Benefits  Under  Issu  Workers: e  Study  Brief, Autom .   How no. “Employee 74atic  349   Certain, Enrollme  (Employee  Benefits  How nt.”  Mu  NB Benefit  Historical ER ch  W W  orking Research ill It  Data,  Cost,  Paper    and     and VanDerhei participa  remained match tion  (S epte and   at ($1.00  that  mber con  tleve per ribu  20  dollar l10) t ions at  year provides   to up  -plan en  tod  six    similar 2010. characteristics—notably  perc  However, en  information t of pay  not ). At  for  all  the   el par   ithe modal gibility ticipa  employer  nts pro  in defi  m hav ised  ned mat e access  em  ccontri hp ing loye  to  rate—may bution  4r 01(k) match  plans  plan  ($  play 0.50   for      ? To what extent has this deficit been impacted by economic conditions over the past several  levels increases as 2008. the population for  many   in enti   the of  The ty   as wage    proposal for is   of  30  pe d  younger  esignated  participants  rcentag percent and  ten may  e cohorts.  uof    re of dis as  the )    4 in having c 21 will 01(k) ourage   private u  nder This ex  p parti “run e   em  isrie age  defi  lar c  p nc short ip  l35 goye ned eants ely   the group, r    of benefi du   matches con  most  mone e to t  ma inui t   the  splan y” ensitivity y ng  to even   impact at   sto 401(k)   that in  c   be 2006. ont    to  tim of better  pla ribu   in assuming e.  This nc trs,ee    off abut in  is se    2008 accomplished in s  offers   in  healt retir  the  af ethno  eadditional ment  care r  su a suspension pp  - du related by ort e  uti   to imatch ng lizing  the   costs data  in    level th    for w e ill  .     that website. 77 employer  plan ned  to be  to  included close the  in  pla  thnis  for  analysis.  new members  Alternative  in th  specifications e next two years.  of minimum    tenure were used with essentially  2010 employees 2012  thro EBRIugh  po  had  20 licy -  an minute  forum  eve ntele   and larger phon  used  average e interviews  in the  i m July provement:  with  2010  1,26  Issu 2 individuals  e2.82  Brief.  per c  (1, en0t03  of  workers compensa  andtion.  259  retirees) The defi  age ned  25    and    The analysis in VanDerhei (February 2011)  was designed to answer two questions:   27 retirement age past age 65  in ?  an  attempt 50–74%  to determine whether retirement age deferral is indeed  cohort balanced surveyed  VanD erhei de   funds indi cre  (O acated se ctob  bu  toer t   13.0 tha that  20t10 1   th perc by b). ey  2010    ent had   for this  already  those  figure  increase  in  had  the  incr  second d eased or planne -income  to 63.0 d to  quartile   percent increase  and  .   their The  reac   EBR employer h aI /ICI mini  da m  ma ta um  tch base  of  and/or  6.1  was       ? Empirical analysis emphasizing the match rate exclusively (as opposed to the match cap or the  3  To what extent has this deficit been impacted by economic conditions 18  over the past several years? 4   VanDerhei 1955 13352 How Institu  (Se  -Does 79.”  t(2007). pe,tember  November   Eli (Washington, gibility    2011 ) for .2010);  The  Participa  DC:  20  and 11  U.S.  RCS  DCIIA t ion Commerce,  questio  in  Rese  a Defined nsarch  were  2008,  Re  fielded  Contribu port  $75  (November  -in$125).  January tion Plan    2 011 2010).  Hel  and p?”   therefore  EBRI Notes,  did  not no.  ask 9    VanDerhei and Copeland (2008) simulated the impact of 401(k) sponsors changing from voluntary to  loans—factoring aGen  criti  Xers calper  role  in  2dollar  012 in retir   in (Figure  on  all e ment six  401   per 9)(k)  .saving.   c In en partici  th t),538 is    over  case pants Dirksen  one–qu  we  with  see   an arter Senate  tha d without t  of the  NH Office  nuECs  m loan b  fails er Building  access  of  to   future  par  in t icipa the  ye ar da testabase,   the in th  workers e 401(k)  only 18  are  plan;  perce  eligible  even nt    this employer increase enhan the accu  same  assumption mulation ced years?   faster resul   con contribu t  tportion s.  ributions  than    tions.  the  of   wa  gen  these   analyzed RSP ral M infla .  Briefly, t ias on  function  rate.  the    mode  of a lmatch  takes  the rate  current  proxy. population    of workers in the  16 51 Even benefit older  in thoug   th plan e Unit h  sp this o edn   analysis sStoates. rs that  R ain  ndom had clud  frozen e  digit s the  dialing   or ex pclosed e ri was en c  used their e of  to  only plans  obtai  thre  were n ae  re  pl  th pr anesentative en  sponsors,  split into  cro  the  those ss  section concl  that u sion of  ha  the d  should  done  U.S.   so  be     Kusko, ? The  Andre  National a, James  Commission  Poterba,  on and  Fiscal  David  Res  Wilcox. ponsibility  “Employee  and Reform  Decisions  prop  With osal on  Re  federal spect to  deb  401(k) t reduc  Plans.” tion ,In     not sufficiently Figure 7 perce nonmatching   More As  able  hy Appe nt  19 poth  refi   for to  pro ndix ned   esized  bi vthose a  vfurcate employer luable  ides analysis  B:  in  Brief in  predi  V   the to  ath  nD is   mi e Chronology    co currently cthird  erhei balan ted tigate ntribu  -  c(July income oc ed nt retirement  tion un ribu   20 fund  derwa of  07 to   tquartile. RSPM i),  on  acategory  some  ydefine s  to   for income     link  employe   stylized Th d   th  in econ  e  treductions adequa  o20  ttarget rribution  06 sparticipa  that  and cy   date have   20 problems   increase plan 07 n  ts di f unds c .scontin oun  This ntributio der    to and for  was   uty 10.8  ed most  pical non ns   par accru   percent on -  ttarget househol plan icularly  aa  plan ls  in match   date  for -th specifi  evident eds  those  de  ing balanced  (assuming fcined   basis   in am   th  on an  e dg      78 interaction between the two) may provide unexpected results. For example, if the employee’s  Follow 401(k) participants -up question  abso  were ut the  asked  specific  of  provisions  those who  used  indica  in the ted  Sep  they tember  would  20 11 either  Gale  in proposal. crease  or decrease the  1. www.uscha What (Employee  percent  Be mber.com/reports/employee n ageefi tof  Resear  U.S. hch ouseholds  Institute ,be  Secame p -benefi t. 2010):  “atts  risk” - study 13- 20. of    insufficient    retirement income as a  for The had automatic  Van  participa   new a Dloan erhei with  mod   enrollment; ou  (Fe t  aitstandin o e  gen lbn  was ruary  iner  a  used o  defin 20 gus   ho at 11  match,   wever, to )ye e. d  ar analyze  con - ?end   more  tgiven ribu 75–100%  2010  how  tion than  its.   eligi  On close plan  20   average, b  perc  ility  mak proxi  ent for esm     aover participa ity still  tre  to  fails m  the  the en  dou to pas t ipassage o  join. nts   in  15 difference    a  years,  defin  of the  eam d   Pension in  con on  thgteir  ribu par  at  Protection ttion -ici rispants k  plan  ratings,  with   Act    of  40 popula private ? tion.  secto What  Tor   further  are in 2006,  the  increase  economic  statistically  representati  imp  attri actbutes s on, of  the a   cell wh  ree  pho ttihe renm re  or  esupplement n  not t savings  they  are  defi was  parti  cit adde  on cdipating   capi to thtal e  sample.  in and  a defi  labor  Sta ned  rmar ting  benefi  kwith etst     3.1  Impact of the financial and housing market crisis in 2008 and 2009 on retirement readiness ... 4  obvious: prior to adopting  In addition  AE  to and  in dthose ividual  that -specific  had change  characteris d their tics  defi  (e.g., ned  age,  ben ewage fit plans  and  between  tenure),  em 2005 ploye  ande  2009.  If  Brady, VanDerhei, “The  Peter Olivia  Mo  Ja  J.men  ck, S,   “Pension   Mit Sarah t of chell  Tru  Hol   N and th,” o dndiscrimina en,   Sylve is sued ands ter Luis  in  December Scheib tion  Alonso.  Rul ere    sed., “   2010. and 401(k)  Living  the  Th  Plan  Incentive  eWith  docume  Asset  Defined   Allocation, to nt Cross  puts  Contribu   forth Subsidize  Accoun  tion a tax  P  tEmployees.”  reform e  Balances, nsions:  plan    and   tha   t  It has long been assumed that matching employer contributions—the allure of “free money” to  For the benefit filter  See   all worker  out  pages   plans plans  mid    10 is  -  may with year  physically –11   want  suspensions.  aof  match  Miller  to  able co   (20 rate ntinue   to 11  proxy  )contin   for to  hav an  of ue  exa e  less  a  working  m very ple.  than  lar    g  50 and e percen  perce  thattage nt,  there  the  of    their co perce ntinues  eligible ntage  to   emp of  be  401(k)  lao ysuitable ees  par participating t idem cipants and   ea  for ch   These formulas. The funds 3.2.3 While highest defined  original    un the two - Risk inco be t i Thi lRS   ne pa 2   me v006 Smanag sfit pers e   values shows rsion   plan quartile  but  provide  eof   even  spons  ment in that  RSP  Fi . gure    older during  an M otechni rs   was interesting   that 3  parti may   tha qused  uha es t  ca  dip five  ppear  in  to closed ants   analysis  re  analyze year t i and to reme  a   perio be  defi   thos  of the  nrelatively  tn ad    e  ed  proposal futur th  with  ben eree    hi  ehas small economic fit gher  with   pl been  an consideri  levels  profound  to  a  well   tremendous new  of- nbeing wage  hires, g  publi they   and of  frozen  crepresent   -increas the policy  ten  retir  utheir re  implica e e   are in the d  defin   the   sum t ions. ed  of      primary concern 0  is to make sure they receive the max 75 imum match possible from the plan  28 amount they currently contribute:  result Source: Tabul  of ati  ons the from  financial EBRI/ICI Participant  mark -Directed et Ret  and irement  real Plan Dat  estate a Collection   Pmarket roject.  crisis in 2008 and 2009?  79 2006    (PPA) there was no way of knowing what the AE plan design parameters in that legislation would  even impacts loans the 20   after 01 outstan   retir wave  adjusting e  dof ment ing,  the  about   income RCS  for,  th all  14 e  data   adequacy worker’s  percent  are weighted   of income  in  the  Septe  re  by  quartil maining mber  age, sex  e. 2010.   ,accoun For  and  ex  e  It dam  tucatio was  bala ple  also ,nce n  those  to   used was refle  in  ctake   to tthe  the  compute  nlowest  act  out ual  as   pro in   aretir c  loan. poort meement ions   quartile  in    the   plan  This  and,  survey and  if so   on wa ,  what sindivi  cond  ty duals? ucpe ted  of  online   plan  within and the  th eattenda  United nt Stat  generosity es by Harris  parameters.  Interactive comm     issioned by the Principal  a 50 contrib the hypoth ution esis  behavior  that th ewill  40 1(k) undoub  improvements tedly be infl  were uenced  a result,  to a large  at lea  exste t partially, nt by plan  of  desig  a simultaneo n variables us  (viz., quid  the pro   U.S.      VanDerh        Departm would . “The Rema Loan Journal Me inor i   modif (A Im inves  eActivi king pril nt p tm  of a ent   ct of 201 y   opt  Pension Responsi  ty  private  of ions Labor, 1).  in   are Automatic   2010. not s -   how Bureau Esector bility c n;on ” theref   omi EB  ore,  fo Enroll  retirement  RI of perc rc   Re s  I ent La  ssu an ages tirem ment bor de do not   Fina Brief,  Statisti e     add in plans ntnce,  t401(k) o   (Philadelphia: no. 100 cs perc   by Vol.  366 .ent  Natio   ca Plans . P 6, erc  (Employee pping ent  pp ages nal  on .  are  127–145   Universi aCompensation  Fu dol nnual lar-w tu  eiBenefit re ght   ed “tax Re ty  (2 avera  of tire 007). - ges  preferre  Researc Pennsylvania .ment  Su   rv eAccumulations: d yh,   contri In 2010. stitute,  Press, bution  “Defined  December  pp. s A  to  69     ?   year. As shown in many in ? dusBegin try 25  –studies, 29 to provide 30  th –34 e participation  an35  average –39  ra 40 tfixed e–s44  among  contri 45  eli –49 bution gible you 50   –54 ng and 55 low –59 -income 60–  employees 64   Given 17 41 participa  that nts  the  (and  parti  would cipatio -ben  parti  perccen ipan tage ts)—provided s for certain  ag rstrong oups of  fi nancial eligible  motivatio participannts  to  (especially  contribute  the  to  young  benefit The contin perce expe the his  or  present assumptions cte  ntage her uing  plan d  to s kto   of   ilva have to ls).  contri  lrecently  uall  eTh   based s  higher me  tha ebute  an mber t- hired  sw ma on   em ine  s  yresponses 2008 r  p  in ,incl   participa lunfortu oye  the ude  was e  last  contributions  de  nately,  at (o n  two cades ts  rle   lack holding ast  years,  is    of 86 ther  not   defi per    for or these eof),  always ci c   planned aets,  nt. given  both   funds: it  However,   “yes,” is   impor plan from  to  28  do  even  .3  design. it  ndividual so   ant the percent   in if   to  percenta  retirement the  Howe  remember    at nex workers  year ver, tge  two    -age this de end   and that  years. creased   is also  200    the def les  6, allows s  eplan     than rred increasing     into one half      population sponsor, b  at the  they  state  would  level.  be EBRI  more  and  likely  the Milbank  to contribu  Memo te at rial  le  Fu astnd,  as  working much 76 as  with the match  the gove  cap. rnor  In  this of   case,  3.2  What factors in the decades prior to the crisis contributed most to retirement insecurity? ...... 5  Not all participants are offered this investment option.   All Given  pla nthe s were  much  still  larger  active  simulated  as of year  acco -enunt d 2 008 balance .   reductions for smaller plans shown in Figure K, it is important  adult population. Data for waves of the RCS conducted before 2001 have been weighted to allow for consistent   with Another savings look Finan Measuring   like. cno ial   shortfalls  futur  Group fa   As  ctor the  ae     result, from years impa that  for   May contri ct   of  Baby the  on  eli  17–   PPA ol gbute  Boomers ibility d  June e  safe rd  cohorts   to  17 are   harbor retir ,  and 2  si 011 mulate  e(those  Ge ment .  Itprovision  neratio sud rveyed   inse over  to run n curity   age   was Xe 79  short 8rs  35)    emplo used  in in   th  is Oc of   esomewhat as  ytmoney  ee las o ber a- tbenefit prototype  few  2010.  86.8  de   probl decision cade  percent    ine  sth matic  is  makers e  the   2008 of  in  the sub   that  for study.  ti -op  me, companies  the tim     wher  values al risk eas        c ? What are the most effective and efficient strategies to encourage and facilitate greater savings  match 2.  caBy GIpCs   and about are guarant  plan  ho eed   inves w limi  tmu ment ts c ch ont for rac  do   tspretax .  you thi  con nkt ributions) you would  as  reduce  well as  your  the  c4o02(g) ntribu  limi tion ts.?   Would you:  62 quo 80 are    significa  for thentl  dec y hireased gher in  a  general ccruals  un  inder  the  40  de 1(fined k) plans  benefi  witht  an  plan,  auCur tomatic  one is rent  is Age  wo  often  enrollment uld  referred expec  feat t th  to ure. at  as th   ethe  earlier  “20/20    cap.”  2.[the] 96, 2011). Simulation contrib Of  lower  those  1998).  u  of tion  who    $20,000 St  plans: udy  are  Based at   Typ or  risk,  20% e  on   of wh    Plan of plan, at  inco  additiona  Design  private me”  (page Mo  lindustry  savings difica  31).tio    do wo Thns  they rkers.”  of Large  ne  ed  Plan  to make  Sponsors.”  each year  EBR  until I Issu  e Brief,  and defined  low  con incotrmibution e) have  plans,  incre anotably sed substa  401ntially (k)s. Indus  under try  automatic  surveys have  enrollment  suggested  (A E) that , many  employee  have wondered     the substantially  80s.    for those with more generous match rate proxies. For participants with a match rate proxy  to of The sponsors Oregon, to   the 47.6 investiga  model  house  perc  Not set  themselves, e:  incorporates “Funds t  en out eh olds how t ”   i  ncby in lude     modele the year the mutual   will  la cha f- unds  end te a  ,likely d stochastic   bank nge 1990s   were 2010. col lin ec  be ti  v  pla e to consi   t rthe us  see t n sjob ,   lidesig f  e foc d i  ns  if ered ten uranc  u this sn e u  s of  will eparat re to  si   serious  tuatio algorithm be e  impact ac  c“a ountt sn ,  and  risk.”  deba coul  any the  tha poold     te. expected ed In  tbe  i  nves pro  Moreover, other  addressed tmv entides produc  words,  contri  tinforma prim  publ ari  for lybution   th inves  ic eOreg tt ed   ion average poli in t he behavior.  on on scy ecuri.   consideration how tTh y   indi RS e cat  San   ed.lon   For values algy   sth ise      of  VanDerh ane iemployer  and Copeland  match  (Ju nof e  20 5011  perce ).  nt of th Testimony by e first 6 percent  of compensation would likely generate a  to note that the plan-size distribution used in this simulation model is based on those found in the  18 Choi, comparisons;  James  J.co, nseque David Laibson, ntly, som eand  data  Brigitte  in the  2Ma 012drian.  RCS may  “Plan  differ  Design  slightly  and  with  401(k)  data  Savings published  Ou  intco  previous mes.”  waves NBER  of  with three to 1,000 employee ? Increase s that do  a o ffer current,  defined  average  contribution  fixed  contri retirement bution  pla  ns. These decision makers were  Moreover, there was no way of knowing the plan design parameters of 401(k) sponsors that would  Mo are the managem   reo influe same ver,  nce income ent  thde  strategies   by 2007   cohort plan  EB -spo RI  ch  with  sn osen tudy sor  twen    and found by in ty  participant divi  or  an  more duals  extremely   at year  reactions  retir s  larg ofe ment future e  to correlation   th age.  eli e ta g   ibility VanDerhei x proposal  betw  would een   (S as  only  the epte  well   adoption ex mber  as p ethe ri  20 en  di  06) cof estribution   this automati       c  for retirement?   The proxy was plan aggregate employer contributions divided by employee contributions for 2007. This is  52 modifications 3.2.1   would Coverage  be le  and ss generous  participation  than  in  th  eem  modifi ploymcations ent-based  that  retir  took ement  place  pla  approximately ns ..................................  at the time  5  www.bls.gov no. retirement  341 (Employee  age /ncs/ebs/detailedprovisions  to make  Benefit  up  Researc for theirh  los Inst ses it/2010/own ut  from e, April  the  2crisis? e010). rship/private/table20a.txt       The contrib whether  perce ution  the nta   matching le geve  ofls  participants ten dcon  tot  cluster ributions  taki  around ng  would  hard  th sh  contin eip  matchi  withdrawals ue to ng  be levels—and  associated  appears  tha to  wi  have tt  has h h i in gher recre inf aopse racrtedic d  slightl  ithe pat ion not y durin  rates ion of g    a  this represente employee betwee example,  proposal n  50 a  ha  dchange   an in  swill  already  dFig  100  un u  from re doubtedly  perce   3participate  are  a 50 nt,  reduced  per  only  becd  en subje   80 in  tby    the match pe  chtrcent o   defi to useholds   aon ned   cost of  the  the  be - ben  assumed first  nefit participants e fit 6  plan perce  analysis  to  and  hav nt  co   of  how beyond entinued   compe zero  mu  defici  ch the  ntosation  l oco  assumption ts. nger ntribu   Fig  to  after u  ate re  75    after 42    percent 006 that portrays  the   he  or    the     focused 81 EBRI  Similar /Investment  larger  pri levels marily   applied employee  Company  on simulated  to th   con Institute oset ribu define  retirement  (ItiCI) on d  bene 40  rate 1(kfit   )than we  database,  plans al tone h  that  with  matchi  not  were  a  th comparis  eto ng  universe  be  100  close  on per  of d  to c   or en 40  ad  froze 1( t  of k) hoc   pl the n ans.  in thresholds   first th  Ev 51 e   next ide  3 perce nce  two  fo  ofr nt   ye the  ars. of       VanDerhei a.  and Redu  Copeland ce it by  about (2001)  a  attempted  quarter   to deal with these plan design influences on employee  Madrian, VanDerhei, In the  Oc  RCS. tob  Working Data e  Brigitte, r  20 Ja  presented ck, 10   testi Sarah Pape  andm  r  Dennis in   ony Hol W10486  tad b  before len, es  Shea.  in  Luis  (2004).  th  th is   “The Alonso  e re  Sen port    Power a   may tand e He   no  Sof al tev t t  Sto huggestion: en ,t  Edu a lBass.  toc  10 ation,  0“401(k)  due  In ertia La  tobor    rou Plan  in  and ndi  401(k)  Asset ng  Pensi  and/or   Participatio Allocation, ons  missing  Committee  A n  cate  ccoun and gories.   on Savings t  “The      selected ? A plan  from  (G  aale,  Principal  2011)  Fina  thancial t wou  Group ld modify  client  list, the  an exdis  their ting  ta data x trea  were tme  not nt  weighted.  of both worker    and employer  Mo enrollment subsequen illustrates of situation  Van  tenure reo Dverhei er,   61.1  with thi  t of  ly fo (Jul s   r  the choose  percent   in th a y   20 401(k) terms e  recently  current 11 )to .  of   of  pla adopt  the  -a  hired employer n  “building   and ti me. AE.  partici  th      As eA    within  freezing bl si  det pmilar, ock” ants ®erm  ea   approach  investi  albeit in or ched  closi  ag  in  e nle   gn agroup. ss    gwhereby in joint   of dramatic,  balanced  the  EBRI/  For  defi   inve example, M  situation  ned or erc st  targe m  er benefi e nt study  ift    exists  risk, ada t  40  plan te  (VanDerh  1(k) longevity   funds, for .   parti the Of  the   those ei, hig c  ris ipa   July perc he kn  and  stdefined  t  in 2007 ent  inco   the thag eme ),   e       obviously only a rough proxy and will be inaccurate to the extent nonelective contributions exist for the plan  Source: EBRI/ERF Retirement Security Projection Model, versions 100205a1 and 100205b1. See text for explanations of models and assumptions. Using of the  the conver  threshold sion to  of AE.  retirement   In fact, this  inco  is me exac  ade tly quacy what  isdescri  found bed  for  abov  all six e (essentially  comparisons  sufficient  in 47 the study.  retire   m For en  t  ? 1–24%  this under cause  period - and these -.eff    In plans. e ct an  conn  analysis   ection.  of Vanguard     participants, Utkus and Young (2011) find that 1.7 percent of the  retirement she suspension. match magnitude  is likely  over  of  S   savings  to the   For values possible  remain   same those   cfor  ostatistical  nt  range  with with  th ribu ose   th match  tresults i on households e  bias  em s will   in ra p  in l oye this te   remain an  proxies Jack VanDerhei, Ph.D.  rregard  .  expected where  With  constan   can  in this  a  ex   non  be increase  cess in  tfform  o -on zero und  of  the at   100  in  in defi i  on Van part  employee  per ci , the Dt  eof was crent, hei,   reduction partici  simulated.    Holde th con ep  tperc ants rin butions ,in  Alon en   and/or the   Obviousl tso a gfuture   e and for  was  plan   all Bass y,   estimate  only of   sponsors. the  (2  th01   73 RS e 1).  Sd         2 53 retirement average What  RS  isexpendi  the size tures,  of Americ  but the ans’  results  retirem  made ent  it  savings  clear that  gap?  major    decisions lie ahead if the state’s  contribucompensatio tion behaviorn —even by working  thou  with gh the  a small  maxi m subse um total t of the  em  EBRI/ ployeIrCI  match  401(k)  for  database.  that single  Ther  worker e was  would     5Wobbly 3.2.2  Stool:    Re Defined tirem ent bene  (Infit)securi  freezes ty  ......................................................................................................... in America,” the model was used to analyze the relative   5  29     As               one        401(k) ..  “M “Falling  would Behavior.” Balances, et  con ho  expect,  dSto t ri of  butions and c   employer ks: Quarterly    th Loan What e  and an  Activity  sw Will   matchin Jour  inetroduce r  Happe  nal to  in  the  of g  2010  contributions,  nE   first a  cto  on flat . ”Retirees'  omi question  -EB rate RI cs . I ref ssu Vol.   Inco  private udep endable   11 Brief, mes? ends 6,  No. in   no. credi   dus to The  4,   366 a try   pp tWorker large  that   .(Employee workers.”  1149   ex serves  tPer ?en 1187 ts   pective,” as on  Benefit   (2001).  athe  federal  size  Researc  Pr    of esentation  matchin  theh   g   benefit quartile. choosing there risk and/or  It  isof   important is   “stochastic” emplo   plan a      them In hi gthis  ysp hees  correlation  o  as to  n case,  con  s  kee essentia o  health rtribute sp   those that  in mind   care  betwee lly  had in  with   th excess  that   risks closed eir  no n    many e    xtho of fu are )clusive    has their tu the s   of added ere    maximum  employers recently the   years defi  investme  plans  se ned  of que  ch     will aeben  m ligibility ntially that anged ntount  use  ehas fit  choose  a    neede   plans to  jobs increa multi  in  a  asimulation  an    d -defined in to tie  sed to d   th radopt    has obtain formula  edramatically.  last  a  con    AE rela  mo  the two  (w t  fo rdel  ibution fu hich  rvely years,  lltheir   showing match.    is  VanDer low  an    plan 401(k) 80.5 o  account  th This   the e  are perc hei, r   analysis re plans   overall a  Holde ent s obalance n  and   had why  is  n   ,      82 oldest cohort (those currently 56-65  Helman,b.  Cope Cut land  it  in an  half, d VanD  orerhei    (2012)   A similar question was asked with the 30 percent government match provision suggested in Gale, Gruber, and  income Choi, example,  James,  to  th  pay e  David  average  for basic  Laibson,  total  retirement  employer  Brigitte  e  Ma x  con pedrian, nses tribution   and and  uni Andrew  improvement nsured  Metri  medi c  for kcal . “Defined   firms costs  that for  Contribu  th  had e entire  frozen tions  retir  th  Peir eensions: ment  plans      participa  VanDerhei nts  (A took ugust  a  hardshi 2011).  p withdrawal in 2006.  This value increased to 1.8 77  percent in 2007 and 2.0  The third panel of Figure 4 (page 8) in that publication shows the distribution of plans in the EBRI/ICI 401(k)  values perce defined stylized nt.  in   parti be     Fig neucfit re ipants.   income 4 would  Mo  as  reo be  a  expected  vresult er, this  of  figure   ato  pension  be  de larger monstr  freeze  thaates n  in th  2006  eth  corresponding e abili  canty  be  of  esti  themate  RSS  mod  va de, lland u  to es  predi   in the  Figur  indemnifi ct e 3,  cation  population be  3was  percent  to have  of  compe adequate nsation  resources  in either  in retirement.  case.    sufficient information to track accurately 137 different “pure” matching formulas, that is, one without a  importance accoun contrib t www.bls.gov Institu for bala  The  u of nce tion  temployer  e, E  the c   December in ont omi ohous / ncs/ebs/detailedprovisions a retirement -cprovid  ehold Crisis  2011).   ed of had   2retir   008: savings  in defined e ment What  accoun   ben  W con ill /2010/own e  tt.Hap fit r ibution s and pen  Social to  plans e rship/private/table27a.txt Ret  iSecuri and/or rees’ In ty.  co IRAs mes?     as  well 2009  as  APPAM    their relative  Fall    Beshears, 5.2.2.1  Reasons Choi, Laibson,  for em  and ploy  Madrian e ?r s to25–49%  provide  (2007)    mat  esticmated hing cont  theribut  emplions oyer    match’s impact on savings plan  using currentl  simple y bein  agv erages refined  of  using  employer  year-en -tod -empl  2010oyee  data.  contrib   utions is problematic).  in 2.1 strategies On simulated Alonso either those  his   a caut   or  tha Ho already  and  her w tion    necessary to froze/closed   can Ba  current  arun ry  sadopted s  this   note,  (2011) short   gap  401(k) to  it   of    achieve or  fo their  is be  money   und were admittedly   plan, bes  defi  that  t  acurrently    measured any  16.8 ned 50,  in    reported 70   1 ben ve perc 998,  rand y  econsideri  ent fit difficult  -   only  90  are (DB   of dec  per    7.3 there )the  rpension  nto ease c gent perce  time,   deter adopting  s  pprobability inec nt    contrib droppin plans. mine ific  of   metrics rec automatic   how Fortunately, ueg  tof n  ions to t   those lsuccess y   only that - hired would  enroll    wor 5.4 shoul    participa EBRI for   have perce kment e  dstylized r  swas be   not ant  features   con mu nts   ab of  currently  in ch   le sider who thdividuals   to elarger   ti  for  ed investe me   ?their    for  at d      6Orszag (2006).  Miller, Judy Plan  A.  Rul  Testi es, m Participant ony Submi  Decisions, tted on behalf  and the  of  the Path  American  of Least  Resistance.” Society of Pensio  In Jam n Professionals es M. Poterba,  and  ed .  period), prior 3.2.3 to  2005 RSP   M  was  Risk baseline  0.69  manage  per  re48sults cment ent  indi of  te  co cate ch mp niq  that ensation, ues Research  the  in retirement  lowe  compared st Director  prere  ........................................................................  with tirement  2.45  percent income  quartile for those  would  that  froze need  to   7  83 database in 2010 vs. 2008 Department of Labor (DOL) Form 5500 for all 401(k) plans and suggests an under- perce  In Dushi, nt in  Iams  2008 , and  before  Lichtenstein  reaching  (2 01 2.2 1)  per , thec ent results  in 2009  from   an and ot h2010. er individ   ual response survey, the Survey of  sometimes contrib contribu utions tion  considerably  rate  und  for er a  each  two  so  -defined tier .  Now  ma  the tching  bene  vafit  lformula u eparti s forc ipant Earl  (e.gy.  ,Boomers can  a 75  be  per  deter  cvary enm t  match from ined.     appr  on the oximately  first 2  percent $70,000  of  (per     c. Reduce it by about three-quarters  nonelective contribution.  Participants in the database were excluded if they were under age 20 or over  exposure  VanDerhei Conference  to  (Se fluctuations ptember  (November  20  in 11  the ).   housin  2009).g   market.  The resulting percentages of households that would not  54 19 participation under automatic enrollment in two ways:  impact various in covered cthe 4.3 401(k) irc  balan  um high  What vent  plans.  than retirement c  est and/or ed   these inc  f  it are unds  Of  would o   m participa those th   limitatio eha  e  ages. quartil  long d  on   defi at    th  t  le term  ins n e ned east g   with same   in in   90  benefi  imp  la a   20 te defin per  indivi a   200 ct or cte  s  plan emore  nt 9 of dual d   with con  of rece  sponsors    their (with years t ribu data nt  market  po tion  the of  on  rtfolio  that eligibility.    same actual plan   vhad o   in would  latili survey   retir  these closed   tye  react   ment on  response) funds.  their  retirement  to  pla      defin this This n  sp had  set  ed value on   sa   tha ben sofovin  r  ithad  nce activity  eworker gfit sntiv ?   increased pl   ans es,   from not   in and   the   to      30 This helps explain some of the early empirical work in this area. For example, using plan data from Form      Yakoboski,        . “Savings Tax Actuaries.  Paul.  policy  and  “Salary   an U.S.  thrift d  th Congress.  Re e  plans:  eco duction nom  Default  Se  yPlans .n  Vol. ate  enroll    Finan and 16 (Ca  Individ ment cem  Committee bridge,  amount ual Saving  MA:  as .  Tax MI   for a Tperce   Re  Retire Press, form ntmen   pp  of Options:  .th  t67–113, .” e  employee EBRI  Promoting  Issue  2004).   maximu Brief,    Re  no. tireme m  155   nt     defer representatio Some betwee   It The  should     effe retirement ofn  the  200 ctive  be  nf  5 noted inancial   of match  and  small  age   th 2009.  ra  a   to projections plans tt ethis  84  is   Similar    a  for before analysis  measure  the   evide associated EB   90 was RI  of  pe /IC   n don the rcent ce I 40   etotal is  1(  using  wi found  k) of  t amo h data  the   both th  unt for ese  ba ho  se. new   useholds those of  proposals   The emplo  contrib   pla that y ner’s would -u   size closed have tions  contrib  variable    have and assumed  their u  investment tion  a  was   pension 50  via   status  percent specified  the  retur   matc plans  quo,   pro in nhing     terms to meaning to b ability  offset  new formulae  of    th of  no e     Historically, Subsequ Income and ent  Pr   providing to ogram  the release  Partici  emppl  ation of oye  the ?r  (SIPP),  matchi  Orego 50–74%  arng ne  study, comp  contrib   ared  it uwas  tiwi ons th  deci   tax to ded  401(k) records  tha  plans ,t  where  the  approach was  the  tho y fought und  coul  tha  todt   be  be a nu   aapplied  mber primary  of  to       A similar question was asked for a 30 percent government match. However, follow-up information for those  In The  Fe  conc bruary ept  2  of 011,  measuring  the mod  retirement el was used  security  to analyze  – or  the  re tiimpa remct ent  of  inco  theme  2008  ade -2009 quacy  crisis  – is  in an  the  ext rfiem nancial ely   and  compensatio individual) fonr ,married  decreasing  households,  to 50 perce  increasing nt for the  to  next  $95,000  3 percent  for single  of compensatio  males and $105,000 n) as wel lfor  as  the single  abili   ty to  84 age 64, had been with the current employer for less than one year, and/or had less than $10,000 in  4 have    What been  “at are  risk” the economic  without the  imp  2008/9 acts of  crisis this defici  that tended  on cap  up ital  “at  and  risk”  labor  vary  marke  fromt sa  and low  on of  3.8 individuals?  percent to  .....  a  7  In the wake of the 2008 financial crisis, a number of employers chose to reduce, suspend, and/or  Benefit recently EBRI 69.8  This last  two   cpercent walc ill  Spe   u  year chan continue latcion Select™  s, by ged  76.1   will 2010.  jobs  to   perce be   (a work     and   When sensitive trand   with had temark  ha  a  da  similar ac    to ei significantly  tof ther  uthe a  He l  participant  an  choic already witt aly seAssociates  ilarger s  of   was adopted  the  da   done 4  ta rat 01(k)   to LLC). e  or  sfor   be  of  balan we    tter  targe return  re  cassess currently et.   date Therefore, on  various some  fu  cons nds  of    in asset the idering  the  2010,  analysis  behavioral  classes  adopting  a tot  in ? al  VanDerhei and of    73.6  it is      5500  Van D filed erhei  an  (N nually ovember  by  ER 20ISA 11) -qualified plans with the IRS, Papke (1995) finds that substantial employee  for participants investment  the employee  los  instead ses  IF experience   th ofe  assets,  employee d  bu  Employee dut rin  acontrib  similar g the umarket  tdistribution e Benefit s enough  crisis. Rese  to  would   Based  re 78c arch e ibe v  on e  the expe Institut  an  full anal cted  match. ysis e  in (EBRI)   the of  Th more  latter is simultaneousl  tha  case. n 3  Ifmillion  this yis   controls pa  indeed rticipants  the  for    behavioral     individuals        . Testim amount Security (Employee  sai  chdanges ony.  the  .matched   y15  Joint   made Be  by  Sept. n  either e  DOL/SEC  fci otby  2011.  ntributio Resear   em thep    Publi em lch oye ns  pIn  to rlcoye ,st    private Hearing aitu  defi rst ethat ned , November  i ndustry on   co sponsor  nt Targ ribution  et workers.”   19 401(k)  Dates 94).  plan    Fpl  bu unds.  ans t th  or e  How  tax  the  re   Wo workers cords uld  said  Target  who  they  -pa  Date dirticipa dn’t  Fund  and te  sin     success. means indicatin employees:  of g  A   an incr lth   Th inough easing cere  average as  ae   signifi or the  de   liicm kelihood rcease ant provement  portion  in c oofntri  pa   of butions inssing   total the  the improv   isempl  not  nondiscrimina  oyer available. ement  401(k)  ta  kes  tcontribution ion  place  (ADP  in)  the test  was  first s (B  only  rady, four  0.56   years 2006).  perc  after  However ent  age  of     other ?  states They  as  analyzed  well. Kans  a pla as nand  sponsor  Massachusetts  with an AE  wer  401(k) e chosen  plan  as that  the  re  nex placted  states  its em  forp lanalysis. oyer match  Resul  with ts  of a   While important Copeland Figure  it14  is   shows  (Oc  tru topic. e to  tha ber    EBRI EBRI t  2011) the  projections  st  firs a rprovid tte  two d a major  es risks  of  the  the  enumerate   projec perc  estimate ent t to ag  provide d de   perc above of the en    this work (investment tage  ty   of force pe  consistent  of  tha measurement  antd  has  longevit  participa  participa y  in ) have, nts  the ted  who   late in  in  an   many have  1990s     more  for   females. real model  estate  em   plo  mark yees et  parti s on retirement cipating in  ainco  plame n14  with  ade  no quacy  employer .      match.   52  Copeland, earnings. Af  Craig. ter applyin  “Employ g eac ment h of -Based  these  Retire screens men  and t Plan  delet  Participa ing any tparticipants ion: Geographic  with  Differences existing account  and Trends,     high of 14.3 percent.  ? 75–100%  terminate their matching contributions.   A Towers Watson analysis of 260 companies that made  (Mar tende automatic percech nt nci   2012 of es   enrollment th  of)e   filters this  newly  group.  out -hired  features  anyone  Until  401(k)   this for  over  43  par  th type  eir age ticipants   of 401(k)  35  information  whose   holding plans.  tenure   targe   is available,  wi t-tdate h their  funds  it  will curr   had be ent  quite   at em  lep ast ldifficult oye  90 r is  per  le  toss c ef u than ntlly  of  assess   their their 63     clear 85 contrib the from case,  match    more the that u tion RSP   rate  the tha M  in nr,  eestimates  the c2,  ris00 e   amaximum no 0se  plans  conse s occur  fo,r  Utkus overall n  amount  sus when   on an  average d    an wha mat  Young  em cthed,   futur p be  (2 loye nefit  01 and e1) r  returns   moves  th re found ductions e possibility  that   from in the   presented th  a e   of fi    median zero nancial  multiple  to  here   rise a  mark  -small tiered  woul  in aets cco  d  or fo   be will  unt rmu moderately  expe   balan be la.  For cted for c es  ex th  to   ample, between esized  be  ne  smaller xt  match   an 30          others made Likely www.bls.gov  co ntributio Impact/  Futur ncs/ebs/detailedprovisions ns accor e 401(k) ding to  Contribu  the tax recor tions? d/2010/own s  bu (T-t160).  didn’ tJune  repor ership/private/table24a.pdf  2009. t the   contributions in the   sur   vey. When the  compensatio 65, them  the  if  ithose mpron  vproposals  ement for those  te nd  were thsat  to  closed  to  level  becom   off prio  in er   ato th  reality,  e2005,  early   bu and  70s t  3.34  tha before t  per current  pi cen ckit ng  rates for  up  those   of in  the wor  tha  lat kter  closed e  def 70se  and rrals,  the  earl  plan  employer y  80s.    the 31 Ippolito 4.1  Kansas    non (199 To  study - elective 7) what  provides  were  exte  contrib  nt pre  an  has s ented econo u tion. marke  m toic   tThey the  volatility  analysis  state’s  found   of over  Long  that  the  th  -  feasibility pla Term e past n parti  Care  several  cof  ipation Se  this rvices  years  approach  rates  Task  impacted  decr  Forc  and eased e   indi on de termi July  viduals’ by 5  11  ntoes ,  2002, 6   risk that      an     cases, money employme  Van 3.D  been erhei  in By  their  nt about    (M shifted -based  arch 401(k)  ho   20  from retir w12   acco mu ).e   the ment ch unts   do em    plan on you ploye  Mar   thi from r nk cto h    you the 1987 1, 2012   em would -2010. p  than loye  increa  Fig e  th  as uey re  ase  did   10 consequen  your   sho at  the con ws  marke tthat rcibution? e of  the  tthe  hi  even gh    evolution Wo  (O  though uld cto byou er  from   the 8,  incr  2007).   defin ease    ed it   several states that were concerned whether their residents would have sufficient income when they  Yakoboski, Mitchell, balances2010.”   wh Oliv  Paul, oia  did   EBRI S.,  and   not St  Ie  ssue Jac  ph 39 make ke  nV Brief,  aP.  employe nDerhei.  Utkus,  #363 e  and  (Employee “C  contribu ontribution  Tongxuan tions  Benefit   Ya Rat  inng.  1998, e  Researc s  “T and urnin  a  Plan total hg  In   Workers Featur  st ofit  163,346 ute, es:   Octob into  An  participants   Analysis Savers? er 201 1).  Ince of   Larg were ntives, e  401(k)     VanDerhei  The results  (A  aspril sumed  2010)  non  simulated e of the 40  th 1(ek)  dif  parference ticipants  be  were tween  automatically  AE and vo  enrolled luntary  enro in thellment se retirement  by comparing  plans;    changes to employer match contributions to deal with the recent economic crisis finds 79 231 originally  Decemb employer er  th  20at07  matches  and Decem  100 bper er -20 cent 10  of was  th  31 e first  per c1ent.  www.ebri.or per  cA ent  subseq  of com uent gp ensati  analysis on and  of nearly  50 per  2c ent million  of the  parti  next cipants  5 perce  durin nt g  current the VanDerhei The years. portfolio than     “benefi  aggregate those   Theref  age  inve  tha  (A t  minus ” pril tsted ore, portion   would defi  2010)    RSPM in cit 30.   be targ   nu of    analyzes evi   m the suppresses etdber  enced date  cost  with   -in funds.  ben by   deta the   the the efit   current  il  full stochastic   analysis plan  401( -specific  Social k)  suggested univ  rate  erse. Security  da  ofta    return  of  above.   approximately retir  mechanism e  EBR ment I did  ben 44  provide 64   1, ty efit 000 pically s and  large  an  em  the  an  defi pa  lassumption oye lyned sisd  of   by  some  this      percen rate proxy, tages  bu aret  nette that dat  ou  hig t, her the  tax ma tch records  rates  show  employe  a 5 9perec  contrib entage poin utions t hi gfall. her  Using level of  a  par subset ticipation  of the  tha  EBRI/ICI n what th   e  The An matching  April  answer  2011  con  tot  article  ributions, the second  introduced ?  and  question Don’t  plan   aknow   availability new  also   /method   not dep ends sure  would  of   Fon  analyzing igure  re  the m 16 a  size in uncha  the  of  account renged. sults   from  balan    the ces  RSP  andM  ex . p o Inste surea d to  of  the     Households betwee alternative n 200  explanation  in5  hig andher  2009.  prer  mig  etireme ht be more nt income  plausible:  quar tIn ile  es s start sence,  at  em  a mu ploye ch hig rs use her  the level,  401(k)  and  ther match efore  to  attract have    and The  results the perce results  arentage   displayed of th  poin e Massachusetts t sby  at  age  most  and  among  tenure  study  new  were  and,  hires   as pre  ex  safter ented pect  the e don ,  plan 401(k)  Dec.  change  1,  partici  2002. . p ants   Wi twith h the  relativ  assistance ely short  of the     benefit reached perce tolerance ntage  structures by  retir  about  of  or e ment the  asset     popula to  age.   allocati defined   tAfter ionons?   ccovered o  condu nt  ................................................................................................................. ribuct t  diing on epends   plans, studies  to  these   afo  larg r Oregon,  can e extent  be  Kansas dea  onlt  how  with  and  the  throug  Massachusetts,  population h a combina  is  de wetfined,  iexpande on of  the  post  d 7  -  instead, they presumed that workers’ rate of contribution after the first year were driven 3  primarily by age and      Utk available     u    s,..   Ste “Re Plan Liquidi  for phti  eanalysis. rement Data.” nty,  and  and   EBR Sha  In     Choice co In  Issu tanu me e  Ad in   Bapat, Brief,  equa 401(k)  no. cy  “P  Plan  After articipants  174  Desi  (Employee  PPAgn.”  and  du  NBER ring  FAS  Benefit   the  158: Working  fi  nancial Part Research   Paper On  crisis: e—Plan  Institu  No.  Total  W11726  Sponsors' te,  returns June  (2005).  1996).  Re  2005 actions.”    -2010,     large 401(k) sponsors with actual plan design parameters.  Figure 15 shows only post-2009  would suspended the same  have  time   an their   period effective  matches,  by  match Utk uwhile s  of: and  1*   29 Bap 1+.5*  chos at (2 5 e 01 =  to 3.5 1)  re sh  (pduce oerc weedn   them. at  of positive  com  According p ensati averagon). e  to a nnu  Towers al total  Watson,  return (0  the .11  perc majority ent).    of  42 of contrib that survey  the  net   likely responses u tion hous  “costs”  plans ing  show.  equity  for  in   Cons salaried terms  is utilized equen  of  em  reduced tly,  “as p lther oyee  needed” e  retir  sare  from  e issues ment  is  Benefit  esti  wi  benefits mated th  th SpecSelect™ e ac  to  for cu  be racy  those  $4.3  of  (a   certain  currently trillion.  trademark  ind    iin vidual   of the  Hewi   respons 401(k) tt Associates  system es in   at  a  type 401(k)  of  datab  analyasse is and  with  substi  salary tutes  information,  a constant Empl oHolden yees C  rate of  an u  return rren d VanDerhei tly A  ofg  either es 25–29:  (2  4001)  per  perform cent nominal ed a regression  per year or  analysis  8    Copeland, Craig, and Jack VanDerhei. “The Declining Role of Private Defined Benefit Pension Plans:  Who  equity market; however, it is a more complicated question involving both the proximity of the  simply less improve  computing ment  an in  ter overall ms of  percentag  additionael  households of the simulated  reaching  life  path a 50s  percent  in a particul  success ar cohort  rate as  tha  retirement t will not   Kansas and  An  retain drews  Insurance   (1 a 992 workforce );  Ev Departm en an  with d Mac ent,  specific p  EBRI herson   was characteristics  (199  able 6);  Basset, to crea  and  Fle te m Re  ma ing tirem tch  anedent s  Rodr  are  Readiness  igues used  (1 to998  re  Ra w ); tings a Even rd workers   and based  Macph  on  with  ae rson full  lower      The retirement tenure Comparing 5.2.2.4 the values   simulatio ?average   within have They The   investment the  -a  perce n   pooled any  impact higher   model overall  work ntage   data  per  of to   1998  strategies, for   adopting  acr e  entage for full cductions e  average   subset nin -blown e  of    auto as firms    hav for  has   asset we namatic  ing tional  the  lwith bee l as   allocations  recovered “long  nannuiti   enrol au   mode relative tomatic -te lment zlnure”  ation in   ly given of  2003   constan enrol   newly on   cohort of   tha employer   some and lment -thir t  their    in ov currently ed  or   2010 eto  r  401(k)  all  ratio   this identify contribut  of  u  pdated 24   ages the of  par   year contribu   the accou t  i36 icipa on    itrela period. -   rates 45 to n nts t  tions  tiincor are  balances onship  with       ag In  to p a  orate 2010, those  iaccoun  nbe   largest at twe     inen t       income characteristics rather than tenure with the current employer, as they might be in auto-enrollment plans  PreviousEBRI Vanguard,  EBRI  Issue  research   Brief, November  has  no.  provided  307  2011  (Employee    an initial  Benefit 56 quantifi  Research cation of  Institu  how tthese e, July  proposals  2007).   would likely affect  accumulations (and rollovers) ? Other  and,    as expected, the simulated balances (as a multiple of final earnings)  The latter were based on investment results alone – before considering the effect of contributions.  retirement 55 the    anala. yz  epla dA  ncompanies  quarter  participation.    ch ose to reinstate their match (75 percent). Of those that reinstated their  SeptemberMed  2011 ian  Senate 401(k)  Fi Accu nance mu  Commi lation Mu ttee ltip  hearin les fog. r Au  However, to-Enro llmen no information t With 2009  on P lan plan F osponsor rmulae reaction  perce 6 LLC) 4.2  in Appe  nt.  2005   Th Are ndix 10is  and   th allows  A: ere  2009.  Brief  other  readers     Description Fr om ways   to tha   tha choose t infor t  of  individuals  RSPM  m which ation,    rate   ahave  subsam  th ey responded  beli pleeve  of  pla is  to  more n  recent  sponsors  likely  ma  rket fo was r the   condi created  future tions?  that  an  ................ d  had  use   the  9  In of          previous the      .  influe “Impact Is Affected,  rese nce  of  ar of  Em c  the h  an , ploye th  dmatch e  How.”  lervel  Matchi  rate   of In  cRobert  o on ng nt   pa ribu onrticipants’   L. Savings ti on Clark s was   and Plan  co  es  ntribu Olivia  timate Partici  tion Mitchell, pdatio  by  rates  assuming n Under  eds.,  and  found  Reori Auto  thatmatic e   they that nting    participa were Enrollment.”  Retirem  a function netnt  before  Ri  NBER sk  of   -   household have (2This 005);  sufficient  is  Eng  thel ehardt  to sum  retirement  retirement  of  an  the d  Kuma effe  ag ct  rincome ive  e(2  (the  00 match 7) ;closer   to GAO  rate  pa  (1 y   to and  99 for  retirement 7)   the the ; Kusko  nonel  simulated , Poterba e ctive age,    co expenses, the  an ntribu  dfewer  Wilcox tion   years the   rate. (1 99 new   8) of;     additional method Mitchell,  com  Utkus  savings pute  andd   Yang  the   age discount  is deferred  rates.  Mi fortchell,  these  Utkus, households.  and Yan   g (2006) posit that employ 24  ee demand could be another  balance slightly 2.2 retirement. 2010, stochastic for with   those  an  VanDerhei, Do   autom more  participa is   in  indivi  likely de  the    cum atic than However,  dual  lowest to  tescalation  uion Holde   ½ lation be s   (54   rates un mu -in n  .5 ders regardless  model ,ch com   Alonso   perce of and  larger. tand  worker e  quar  the that nt)   and how       of match of Ove  t  cont to ile    the Ba all  ok to  rall (24.9 ributions.  s full   s  calculate into asset an   (2011) 95 -d  time, perc    aper f occou allocati und  The  en cfind   fu en ho nt t)   llthat exclusion t.w -     tha on of year the The  much    the aand/or t  house     aggregating redu 1wage   percentag participant   of thcti  auto  h ey  de an old’s o  will ns gree d- enrollment salary  for   elongev  ac n    balance of point eed  rthis o  an  ss workers  fo   nui ity ag  all de  rplans  se   risk, tization re age crease   were cohort tire   ages   groups, in post ment?   projected this  in   util  de 21 - re the  analysis crease ized - t64  equi i rmaxi e  were m  by  ty to e   to was  m n   tht  um  e      several significant changes, including the impacts of defined benefit plan freezes, automatic enrollment  individual https://insti  participantutional.vangu t retirement savings, ard.com/VGA  by agepp/iip/sit  and incom e/institutional/re e. These earlier 46 searchcommentary/article  projections, however, were /In  20 would be minas imal a  Function of Salary for older age cohorts. Quartile   However, and Number of Years  for those with a Eligible for major portion a 401(k)  of their Plan  careers  7match, Soto and  105  Butrica  companies  (2009 )(74  conclu  percent) de that  rein  among troduce  a sample d the original  of large  match  401(k)  amount.  plans, mat  Amcon h grates  these  are  plan  lower     adopted to the propos  automatic al was  enroll  available ment  at  401(k)  that ti  plans me. Conseq  by 200u9, ently,  but  did the  not  2011  have  EBRI  them  analysis  in 2005  presente  (the last d ther  observation e was    corresponding tax contribution  set  rates  of results.  fell minimally    as the employer match rate rose.  However, that1 analysis also      demographi  (2   00 While       7);.  “M Papke  65 Working Man e  seems acs  agem (1  uvariables r99 in   to Pape 5) ge    Re  still an nttirem .dr    Ox   be and Papke 13352  ford a ent  societal some  and   University (20  Inco    Poter 07). measu  norm, me    bAar  dequa   even Press e (199  of 5)  aSo  cy:  for ;ma  cial Yakaboski   Calc the tch  Se  curity rate u Pe lat nsion  (1 in  of  has 99 g  the Reali 4)   Researc now ;   pls an. tic adopted   In hHowever,  cCoome uncil,  a later  Replac   2010: this  age  ement a  fo pproach  122 r full - 136.  Rat ben  fails ee   s.” fits.   to EBR     I  56 available),  Figure 30 fro  thm e  relative Copeland  lev  (Oelctober  of prere  2011 tir).ement    income, and the desired probability of adequate  alternative perce necessary ntage  give   explanation, ofn  households the current  with  modeling  that  the  would  result  ass umeet mption  that  that that  of  no  requirement highl  joby  change.  compensat   mor It would ee  tha d  em be n  a very p lspe oye  difficult ceified s deman  perc  to pro den  mo vtag ideree  an   of    times  investment b.  risk, Half  and exposure to potentially catastrophic nursing-home and home-health-care risks. This  participa have One retiree, provisions 7.2 funds  per  of  recover  dec  cthe  there potential ent ting r eased  for ba  for  ein   sic remains d 401(k)   those  an to   objectives from  match   employ the   plan in   i64.8 ar    th le co was svel ment e  nsiderable   and percent  of second   associated at  RSP   the -the based M  -  recent  to ma income  is  ch    38.0 to re rket  ance with  tsimulate i rcrises   e  percent high, quar m   athat e  1.8 n  tin  ttho ile    th  pla  th to th   and eand ugh e  en 3.8 y   per   financial (o according   co   then le percentage rc  mpany their ss en  than  tage increase   and spouse)   to   stock 90 of   calcula housing    point the perce  to  de  may   popula 10.0 crea  decrease tnt ions   ma encounter  of  perce sed  rkets. t  those based ion  from   nt in that     with  for plan on   8.6 a  will    le those th    more ngthy perce parti e  be  2011   “at in c   nt than ipation st  the    arisk” to y    in    4.3  What are the long term impacts of recent market volatility on retirement savings? 4  ................ 10   This material first appeared in VanDer (Total hei26  an balances, baseline assumptions) d Copeland (July 2010).  not based vResDuringCrisis  on survey evide   nce of how employers—the sponsors of private-sector 401(k) retirement  remaining, the differences in additional accumulations due to auto-enrollment prove to be quite  among Less sponsors,  than  firms   half the  with   most of workers  automatic  freque  (42 nt   ematch perce nrollment n  for t) in m   than ula the  before 2012  among  Ret  and  those irement  after  without  th  Confidence e crisis  auto  was matic  Surv  50  perce ey en rollment (RCS) nt of  repo  up  after  to rt  th 6    percent ey    based on several alternative scenarios.  Moreover, the information used to model potential 401(k)  The that  problem was not  in with flue  using nced  aby  50  PPA)  perce . Th  6nte  pr follobability owing information  of success,  was of19  course,  coded  isfor  tha  eat ch ho  plan: useholds    is in a position  21 43 found accoun  that t Issu for  as  ethe   Brief, the  fact  match  no.  that  297   ca some p  (Employee  chosen  plans  by ha  Benefit  ve the  different  em p Research loye  match r inc  Institute, re arates sed,  for participa  Sep  different tember nt contr  le  2006). velibution s of    the  rates  perc ent rose. ag  e of  retirement income.  8in accurate  the simulation.  analysis of  the average percentage reductions in 401(k) balance under auto-enrollment if the plans  generous tax-deferred employer matches. Both of these latter arguments see the employer match as a  4.3 March was a20 third  nursing  years  per  followed -in  Current ccat ent.  oof  ho m  six  te e   me  However   quartile. nure by months  Population  that  the  wer   expansion may   of  Th ethis   lieligi  eleave kely    Survey. reductions was b ility. to    of at th  have   eleast RSP       family However,   M increase recovered  partially  and  uni  the t   when with to  offset   Ret to  17.1   ath iremen  the  mu at   by percent   level. same ch  an  thigh   in Readi  c  info for   rer ea   nthose rmation probability se ess  in Rati  balance  in  is th ngs  fil e  of   highest tto dered  “runnin   funds a national  to -inco   from exclude g sho me  model  9.1 rt”  quartile  workers   per of  and cent .      In of    The Van  addi having  Doriginal tion, erhei  retirement   on and  mea   Copeland tchi more ng  inco   cmod ontribu  (2 me i01 fication  that 1) tion .  A   is  was fu   isina t umad  25 red equate  vperc eersion  by en  RSP t   of to  on   th M cover  the e  before  model  first  basic  4   will  un per  eder  xinclude cpenses ent taking  of  pa th  and  eythis   contrib ability  pay  analysis.   to for uted.  model  uninsu       It It was  retir  isr   ed a reement  well placed  heal -th  ag  with  es    Responses Munnell, 57 Dushi, Irena,  Al  we ic ia Howard r eH.,  obtained  Annika  M. Ia   Sfrom ms, undén,  and  1,018   and Jules  pl  Catherine an  Lichtenstein.  sponsors  Tayl  grouped  or. “Assessment  “What  int o Determines  the  of  following Retirem  401(K) en  size t Plan   Part categorie  Coverage icipation s based   by and  Firm   on    It should be noted that there is more than one definition of a pension freeze:    plans—would be likely to react to potential changes in the tax treatment of these contributions, or how  significan c.t : Wh Three en -workers quarters,  currently  or   ages 25–29 are compared, the median 401(k) balances increase  and/or controlling of salary.  their  The  for  spouse  media  firm characteristics.  nhave  durat 9 tried ion  to for  cal   ma However, cula tcht esuspensions  how  there  mu  we ch  was rmoney e two  12   major mon they twill hs,  limita   for nee  tcompanies ions d to  of have  this  saved   with analysis:   quantifiable so that    they   When  The nomi  we modeled nal cost of  th  theese  Baby  expendit  Boomers ures   and increases  Gen  Xers with  co inmponen  2012 (Figur t-speci e f1) ic  be inflation tween  ass  43.3–44.3 umptions.  pe  See rcent  the  of  participa 4.4  nTo t reac  what tion  exte  to nt the  will  proposal  the retirement  was limi tof ed  th  toe  “an Baby  analysis  Boomers  of  two impact  new  ca  ques pitalt ions and  labo fromr  the markets?  21st   .. 10  Utk where us,  they Steph  weill n  and “run  Jean  short  Y oofung,  money”  “The  gr in eat retirement  recession  one  and  chance  401 57 (k)  out  plan  of  parti two.c  Whil ipante  behavior,” most house  Vanguard, holds (at   compensatio included an auntomatic  contrib  esucala ted.tion  For  provision.  example,  For  a  plan example,  may  ifoffer  a parti  a cdollar ipant’s -for  con -dollar tributio  match n rate  for had  th  alrea e firdsyt  bee 2 per n cent  workforce management tool, rather than a regulatory response.  the prior  presenta  to Medicare tion  eligibility.  of the firs   t micro-simulation retirement-income-adequacy model built in part from  with known care money to an   ?30.7 emplo   The costs le ss   fact in perc  y  default tha  er retirement.  for   that cen nontributio   the $10,000 t.   job contribution  remainder  ten   nVanDerhei  in  uequ re  annual  aisl   of  to longer rat   th 4  eearni   perce (2005) eir  for   for liv  the ngs nes  tdefin   uses of    AE once as  pay   well plan ed  the   plus they  ben   as  EBR in  an    efit those retire. 200 I  an RSP  par 9. nual   working M tici      profit mode Howe pants - sharing lfor ver,   to than   employers evaluate  the   for con  EBRI  ei tribution t h  the Retire e  with r  define im   pmen less actd   than t  of  contrib Readiness  pur  100 chasing u  tion          Kusko,        .  “Defined Poterba, Size, Contributions  Using   Benefit and  W  Wilco -?” 2   Plan Tax CRR x  (1994 Re   Free WP cord  2zes: )000  utilized s,”  Who's  -Social 1 (Cen  em   Affected, Security te prloye  fore  Re  -Bull level  How tirem etin.  data  Mu ent  Vol. c   from hResearc ,  and 71,  the   No. Replacing h  401   at 2,  Boston May (k) plan  Lost  2011  College,  at  Accruals.”   a mediu  December, m  EBRI -sized      total retirement plan assets:   those Looking  de  at cisions  all households  might, in  turn, that  would affect  pa nerticipant ed to save -savings  an additional  accumula  am tion. ount  A  (o dditionally, ver and above  while  th  theose  savings     from approximately 1.5 times final earnings Income Quar tile under voluntary enrollment to more than 6.0 times final  appendix for more details.  can 4.4 The those dates. Based   live average    househ To on Mo  comf   what th ste  -companies perce sortably olds e  exte findin  were ntage nt  ing  will  s,re   reinstated pr   the rteojected ir  th ductions e  authors m e ret ent  ito  .rth  eThis  for  men  have eir estimate    the is match t   comparable inade of  “long  the   that after quate  -Baby te  for nure”  nin    retirement  to aBoomers e  typi   most or  cohort  12 cal   of  months. em  im    currently th income ppact eloye  per   capital rc   for match en  ages tage  eve   and 46  (viz., ns  measured BASIC - 55 labor  50  are     retirement  perc mark  larg  from ent eets? st  ma  for 2003–     tch  ?wave A  of “hard  the  fre  Ret eze” irement  is one  Confidence  in which no  aSurv ddition ey al (RCS)  benefits  reflecting  will accr  how ue to  workers  any current  indi cate pland  participants  they would  from  likely  either     least escalated As explored  those  to  th 8  in  percen at  th  are e June  tcognizant  of co  20mpensation 11 Issu  of these e Brief,  at  risks) one  the  emplo   are RSP M liykel er,  al  yand lowed  to  upo have  retir n  jo a berisk  ment chang  aversion 65 -einco  was 7me  automatically  le vel adequa  that cy would    enr to be o lled mak  assessed  into e this      of compe March nsation  2011  contri   buted and a 50 percent match for the next 3 percent of compensation  d. Double it  ? This study was based on U.S. Department of Labor Form 5500 data that do not include specific  long 22 44 employees, administrativ -term car  the ee  401(k) insurance  participa  data 8t ion o  at n  retir  percentag the eEBRI ment  De e  income  in ce mb 2010 e radequacy.   increases 2003 policy     to Th   forum. 67.5 e analysis  perce  The  suggests nt.  basic    model  that this  was  ma  thyen  be    a  Rating participa U.S. manufacturing ™ also nts or  provides  workers  firm  information  in  to  gen  analyze eral,   on given the  the  participa  the  distri  financial bu tion tion  and   conseque of  the con tliribu kely nce t inu on s of mber  de  job cisions   chan of ye  ge of ar s  workers up  before on em   those eli pl 5goyees ible  at  for   risk     5  What 2000). Issu www.ssa.gov/policy/docs/ssb/v71n2/v7  are e Brief,   the http://  most  no.c rr.bc.edu/images/st 291  effectiv  (Employee e and efficient  Benefit Lowesories/Working_Papers/wp_2000 t   Research strategies 1n2p53.pdf  Institute,  to encourage   53  Marc  and h 2006).  facili -12.pdf   tate     greater savings for  already factored into the baseline 80 model), the median percentage of additional compensation for Early  projections incorporated the potential impact of the specific provisions of the Gale proposal, they were  earnings 5.2.2.2   It The  is  important avera  in Empiri g  the e emp    to auto cal  lno o yst -te eeeudies n  th  contr rollment at th  on ibution is  th models  sce e  impact rate nario.  all  fel  U.S.  l  of by   workers  mat 0.65c  perce hing .  As  ncontribut  ta  of result,  pay.   th ions e balan  on participation ces will be signifi    cantly smaller  2011, 58 on Another those  the additional   but in  first   th possible  lo e  6  wer lowe   perce tenure  than  st area nt - income or   the of   of increases  pa   int 53y),  equartile  perce re  elst  im in would  com nt in at (21  recorded ping ensat .1  be   th per  participa eion.  cmatch  en in    t). 200  Th  nunder 0t e  and tradi  reductions   the an ng  AE   a47 c tiplan  perce vity  for   cou duri this nt ld in  nage  g  re 20  th duce  cohort 08 is  period (Figure  plan  decr   par of  5)ease  ti.ticipa me.    to  In t ion 9.9  an   by     expenses another ? The  40  entire  plu 1(k)s  plan, uninsure  mat  would ch rate d  healt they  con  “remember” ht ribution care costs.  formulae  their   Even  curre  tho  for ntugh  both  rate  this   of years.   deferral number    and  is quite  start  deferring large, th ein  good  the new  news  plan  is  at  at react  retirement  if they 1.       were  <$1  ages  milli  no  later  longer on.  than    allowed  65.    to defer retirement savings plan contributions from taxable  crisk on tassumption ribuhttps://insti ted. In addi  untenable tion, tutional.vangu  the , switchi  strategy ard.com/VGA ng  do to es a hig  not her  clearly pp/iip/sit  probability  distin e/institutional/re guish  of success  between  will  search asignificantly  plan commentary/article that match  reduce es  the 50   /Re  Net housing equity is introduced into the model in three different mechanisms (explained below).  By the middle of 2009, almost 10 percent of Fidelity recordkept 31  defined contribution plans suspended or  modifiedinformation  for testimony  on  for 401(k)  the  match Senate  rates.  Special 2  Instead,  Comm  the ittee  authors  on Aging  constructed  in 2004 to  an  qu  estima antify the te for  ben  theefic  ma ialtch     “run particularly participa  short ting  of  powerfu   in money,”  a final l risk - average as  we managem ll as  defi  thened ent  perc   ben teen chni etfit agque,  epensio  of  ecompensation specially n plan.  There for r etirees they fore,  would  the  in the  typi  ne  second cal ed  te in nure ter  and ms  dis  third  oftributions  additional  income      Nessmith, this plan. The  William ir analysis  E., St esuggested phen P. Utkus,  that   contrib and Jean ution  A. Y  decisions oung. Measuring  of eligible  the  employees  Effectiven  are ess  of relatively  Automati   c  27 5.2.1.1.3 based Boomers retirement? than simulation  on   desiring wor Impact  ............................................................................................................................... k er models  responses  a  on  50   401(k) of  per  those cent  to  Balances  current   ge probability neric  40   at qu 1(k  Retire est  )of  participants i on retirement men s abo t Age ut  (Hol  chang  in cden om es  and e  to ad   Van the equacy D  taxa erhei,  would bility  200  of 2)  be   or 401(k)  3.0  tho  percent se  con  .................. eligtible ributions.  of  for      12    45 5 that  to  rate, 11 perc  or wo enutaldg e th  points. eir cont  ribution rate drop to42  the default rate of the new plan?  Undoubtedly many 401(k)  This 3.2.2 analysis that perce   qu this ntest   Defin for  of isi  on 5 Vanguard  those -8  ed is  percentag  parti  ben  in the e  cparticipants, ularl fit  second freezes e ypoints  problema - in  LOW  cUtkus ome tiEcR    quartile for and  than  several  You  wh  and nat g   re (2 we  th011) ason en  found  incr  s. find   First,  ease in  that  2003.   with to  the  11.6      respect per  per cencta ent  toge   the for  of   those impact participants  in  on  the  ca   third pital - Dworak income           . Testim .” -Fisher,     ony.  Keenan.  U.S. Congress.  “Encouragin 7   Senate g Pa rticipation Special Co mmittee in 401(k)  on Plans:  Aging.  Reconsi  Do We der  Have ing the  a Crisis  Empl  in oyer  America?  Match,   percentage 11  of households capable of satisfying the threshold at any 66 given retirement age. For example,  perce  Altho ntu tRes gof h  th coGreatR entribu  impac  tions t of mat  forc hing the  first contributi  4 perc ons ent  on  of  employee  compensation  contributio  from n  be those havior  pl ans has  been that  match studied  50  extensively  percent   In ? 59  the A  “so lastft  20 freeze”  years,  will  several  generally  empiri  limitcal  increases  studies  for  have  curre nt ana  participants lyzed the  effect in accr ued of the  benefits  existence  for addition  of matching al years    of  3 The reduced ?  liAll kelihoo  nonele rate  their  as d  contri of ctiv  th  doing ee  ra  bution con tio  at   of ribu re   dollars, employer tirtions eme npaid  although t savings -to  to -employee  the   nee by  defin  December ds  ed con calculation  contribution tribu  2010, tions  increases   55 for  parti  percent  each c  with ipants  401(k)  of  hou  plan  by  plan s eh th  sponsors eold .    employer.  income,  indica    ted  impact quartiles.  of a  mandatory con81 tribution of 5 percent of compensation.    are insensitive savings The  replaced 6.0  to multi 2.   ha      to   ve with p$1  the le   a  milli in    50, rate those  Fig on–$10  70,  uofre  fo   or em  15 r  90 partici  p  millio migh loye  perrpants tcn.   en matchin appear   t probability  exclusiv  to g on  be  ely wor  too  of  ink   small retirement  er defin  con  to ed tribu  rea  contribution  ti in chons c o conve m  and e ad n  that  tion equac plan  al most s  to yretirement .  account  employees   for income   the maintain       participation  IRS taxEnrollment.  tables  (Hold  from en  20  an Vol 09 d .Va    1. are nDerhei  Van  used gua  to , 20 rd  com  05 Center ).p ute  the for  tax Retire  owemen d ont  the  Researc  amounts h  (Vall  received ey Forge,  from  PA:  define  Thed  Van beneguar fit pla d ns  compensation each year until retirement age to account for the financial and housing market crisis in  In participants  a July 2011  in this  Notes  automa  article tic enrollment , it provide  situation d preliminary  follow  th evi e d latter ence  appro  of the ach.  impa  As actdditio  of the nal  “20/20  information  caps”  be  comes  markets, trading income  in quartile.  assumptions  2008 (16  Th eperc  r ewdu en illctions  tneed ) was   to increase  actua  be mad lly  to lower e  14.1 with  than   per respect  cwhat ent  to for  it  wha   was those t  in  the   in 2005   Baby the  (19 hi  Boomers ghe  pest rcent) -income  will  an ddo  quartile.  si  with milar  their  to        in voluntar Industrial Results y enrollment  From  Rel athe  40 tions.” 1(  EBR k) pla  IA -ER  nJo s,Fu  th  Ret rnal ereirem   of has  Econ e  been nt Se o m relatively cuy rand ity Projection  Society,  little resear  Vol.  Mod ch  50,  e on l ,Is   27 au sue t oJan  ma 4,  2004 tic pp .enrollment  71  (T3–737 -141).   (O  plans cto  at ber  this     VanDerhei of if the  contrib participation,  success u ti (M ons  rate arch   for but   is2012)    the the move    definiti first  utilizes d  to 6  per on a  thresh  th of cen e  compe  defin t of old  compe e n  of sdation   contri 70 nper  use sation. bution cdent,  in  th Sin  only  p e acformula rt e  ic 2th  ip out ean  data   may of t respons  5   used  be house  allowed  in ehs  this  olds to  to the  research   in increase.   RCS the  lowe questions  contain     st-    Integrat22 ing new data from plan sponsors, Highes  VanDerhei t  (March 2012) provides a perspective on the impact  contrib 5.1  utions Automati  on the c  eprobability nrollment  ................................................................................................................. of participating in 401(k) plans that use voluntary enrollment. The   12  educa The The Dworak they   im dawn  planne tp ion, -rFisher ovem  of  and d  the  to  e  (2008) financial n   reinstate new t over   ye uses  the  ar as  their   sets. in  microdata last  2006   ma  ni Inne   addi tch be  year ga   fro within tinon,  m swith  is   the married   largely th  a  eNation  floo  ne xt due  dworkers   of 12 al   to Compensation  news months.  the  (co   reports fam ct  Fi pared  that delity  about   Survey in  with   also 2003  th  unmarried   reported (N e  very  “new” CS)  few to   tren offer that   workers); 401(k) d  among  a  among  new  sponsors  those   line        5.2.1.1 New Survey Analysis  increased and Social  Sec jobu  mobili rity (with ty  that the percen  is likely tage  to  of  acc  Social omp  Secu any rity the  benefits  pension  subject  freeze.  to    Federal Income Tax proxied as a  23 VanDerhei, targets. the sameGroup,   par Ther  Jatck. icipa efore,  2007).  “Modifying tion  Figure    status 6  16  th  and  recasts e Fe  con deral t  the ribution  Tax  AE  Trea  results  rate tme  ye  frnt ar om  of  after   Figur 401  year (k) e  15 Pl  despite a  for n Contribu  just  subs  theti  tantial youngest ons: Projected  ch anges cohort   in Im  and  the pact      2008 available  and  on  200  wo9. rk  eSi rsmilar ’ behavioral  values  responses  are 0.9 perc  to aut ento -for enroll  Late ment,  Boomers  EBRI will  and  up date 0.3 per  thisc eanalysis nt for Gen  to provid  Xers.e     aA  more  90    proposed by the National Commission on Fiscal Responsibility and Reform on projected retirement  4 what ?What   it had They  are  been  merge  th ein  econ  2007 d the omic   (15 Form   perce im  550 pacts n0t).  data  of  The  this  with  value  deficit  information  dec  on reased  capital   on to   and automatic only  labor  13 per  ma  enrollment cen rktet  ins  2009. and  from  on  They  indivi  the  also  Pdeunsio  find als?ns      Appen asset point  The   in  allocati simulation d  time. ix 3.   C     desc   Nesmith, ons $10  model rib   in milli  eretirement s  Ut  on–$50 how analyz kus  households and es millio   how  as Young  well  su n.  cces     (2  as (whose 00  sthe  7) cha  provide  rate nges  heads  at  with:   evidence which  are    currently  they  that  will  new  ages  spend  emplo  36–62) -do yeewn s  are hired  the  track   assets under ed  a  through in utomatic  their    2011).  income quartile will attain retirement income adequacy even if they defer retirement age to 84.  The Whether of above, plan  a  first -scenario specific  as  maj   pla well n o   matching where  r  sponsors as  modificati  the  the  plan  formulas,   were current on  sp o of  more n sthe  otax  the r  responses  mo  or trea  actual  le del tmen ss  was generous  match  tto   pr of  the esented  employer  ra   Al after telianceBernste  at at   each adopting   and the  perc   EBR wor  AE en Ikin  May e  rtsurvey, was  apre ge  200   -mea leve ta 4xto   lcontrib policy s   of ured parameterize  con   with forum. utribu tions  th ti  ree ons In was  the  an   is      magnitude of the results vary considerably dependin g on the type of database used, the methodologies  age larger of Analysis private  research  35  com   and defined  of p  th  older on anies, e  th oldest  ben  e(compared   those im efit  pcoho a  pl ct  with an  of rt    sponsors  with employer  (those more  those   than curr  of  matches  age  ently  “freezing” 5,000  25–34);  56  employ  on -65)   4 their  re 01(k)  show ees, tir  pension e m parti   most ae  mark ntc  savers ipation  (71  plans ed  percent)  dec   (co for  rates. rease m  cur pared   had rTh  ent ine  th    already  author with or e  average new  nonsavers);   splits  reinstated workers.  percentag  the   and In  or     e  used function  au tomatic of the various  enrollme  retirement nt (AE)  income  provisi  comp ons and onents)  the  as participation  well as the i nrat dividual es among  account  the  withdrawals.  lower incom   e  VanDerhei, Post  Ja -2009 401( ck, and k)  Craig  Copeland. “The Impact of Deferring Retirement Age on Retirement Income  ?provides employer's In addition,  further  mat  ac   plan hbreakou  rate.  sponsor  Mo ts reo by  ma  the vyer,  choose  num  theyber   fi ton   idof m  that plement  years  insti  e ligible at utional parti afor l  freeze constraints  participa  in which tion  on   the in  con  a  pla  401(k) tributions, n is frozen  plan  imposed   for as  well some  as   by but       robust model. on Participant    Account Balances,” March 2012, EBRI Notes   percent probability of retirement income adequacy would require an even  larger increase: The median  5.2  What incentives have the greatest bearing on the behavior of employers and employees ...... 13  accumulations.  retirement 5.2.1.1.1 that retirement enrollment  27 per &   Investm 401   cPlan accounts. age, ent (k   )of Sponsors and  pla  ethe nts n  how s       Unfo  participants hav database   their ertunately  partici  retirement  of p ation  trade the  there  top  rates d  i over  n is1,000 come/w    nearly extremel  the  pension   do 20 ealth uble 07 y limited -2010     fu th is ond simulated se  s,  peri  information for  which   od new  but  for   in emplo cl   that the udes  at y   fth on ee o llowing aesl y    flag hired curr  3 perc  in ent   under com dicati en  tim tp   voluntar ng onents: of e to   the wh  allow  et yh  e  r    4 percent rate of return: The median indemnification contribution rate for a career-average defined  Papke, Leslie. "Accum  ulat “Participatio ions" as a n in and Contributions to 401(k) Pension Plans: Evidence from Plan Data.”  different modified known. voluntary Increasing  There   me such  enr  the trics: ofore,   tha thr llment  et  shold VanDerhei worker  mod  5tous  80 le would   of  and percent  RSPM   Copel have   reduces in  to a nd order  pay  (2001    federa th toe  estimate )  nu used lm  taxes b  an er   the of  estima on  low   like these etion ly st  impa  pamounts  rproced eret cti rof em ur   currently, the ee ntha  tproposed  income t ta kes rather   quartil  afederal dvantage  than e - tax on     analysis employed,  to  de and termine  the assumptions  the impact  util  of ize annuitizi d; however, ng defin  the ed  overall  contribu  cons tion ensus  and  IRA is that  balan , forc es 401(k)  at retirement  plans that  age,     60 participa 4.1 employees participa planned reality, reductions ? The To  th   to maximum nts nts  what ese    fo re (t   in in ho  rinstate decisions   tthe  aoexte se   defin three     level most “long nt  their e   has   d diff of hav  -li  te contribution  kely emplo   match. erent market nure” e bee  toy  ee inco be  ncohort  Mo   quite  vol  contrib atme re   aplan risk) tili     than prevalent in groups uty    the (compared tiwas  ons ov  60  lowest er    quite  and allowed per  the  inc  concludes  ent  re  low. -past  income with c  by  eof nt      the se  With  employers  non years, ve   plan quartile  tha rp  ath articipan l   teye sp and :  adoption oan   r  with (12 sare so  im rt  .7  (6 s) part  pa  , per a   more pl 9,  cte of   an 12 of c  AE ent), d     size the  and often i ndi in     the alt well 15 of viduals’    report hperce between  ough pas -doctn   ufew   risk ttrying itm  of  should e   500   nt years, ed  to -      not all 4.   participants.     $50 million–$250    million.   Engel eitherhardt,  thAdequa e emp  Gary loye cy.”  and r  EBR or  Anil  th I Issu e  Ku  IRS, m e aBrief,  are r. “Employer  an  no.  ex  358 tremely   (Employee matchin  important g and  Benefit   401(k) influe  Research nce  saving:  on c   Institu Evide ontribut nce te,or   Jun from  behavior. e  2011). the health        and  32 the    relative income level. For those workers assumed to be eligible (whether or not they choose to  percentage Multiple of  of  Final additional Earnings  compensation for Early Boomers desiring a 90 percent probability of   Roth IRA and 401(k) accounts are not used in this version of the model but will be incorporated into a  enrollment 401(k) plans  (86 percent versus 45 percent). However, they show that overall plan contribution rates  participaplan nts  so administrators ld out of stoc kreported s entirel yoffering  and that  a u1tomati  percent c e  trad nrollment ed to 100  in th  perce eir defi nt ned equities.  contribution  Five per  cent of  informed benefit See Holden  pe  esti nsi  and omates n  plan VanDer  in  is  this  11.6 hei   (20 regard.  percent, 05).    How  assuming ever, as  a  part  4 perc  of ent its research  rate of retur  mission, n (Figur  EBRI’s e 11).  Center   An indemnifi  for Researc cation h              . “Tax Journal  Reform  of Huma  Option n Reso s: Promoting urces, Vol.  Retire  30, No. men  2,t  Se pp.cu  311 rity?.”325  EB RI (1995).  Issue  Brief, no. 364 (Employee  T households he views ex pr that essed  can in  satisfy this statement  this staare ndar sodlel  at y  a thos  retirement e of Jack VanDer  age ofhei  84 and to approxima should not t be ely attribute  1 out of d to  7.the   VanDerhei of modifications a deferred  knowing   basis, and  the  on   Copeland, differi   as projected  under ng in   2current c 004, e401(k) nt 82iv  were es   law, balan  that  able  and  an ces   to  employe  parti at  demonstrate  retirement cipaents  eligi  would  bage,  le that  to   ass receive   for contribute u ming a household  an  the  18  to   mod pe  a rcent  401(k) seeking ificat  govern  ifaces on  as  75  took  ment at  perce  each  effect  match nt         have compensa not employed tion).  automatic enrollment, an employer match has a positive impact on plan  and In be do 999  recent   noted a   long employe  c5.2.1 atole lc-u   ter mo that  lrance atm ion nths, e sthe  Tax de  .had   or cline   average  two incentive  asset  already   of surveys   “traditional” allocati  reduction  sreinstated  ......................................................................................................................  have ons?  pr  will      pension  ovided or  be  indica  most  addi  pla te  muted nti ths; onal ey  what’s  plan   information by  previou  to unusu  reinstate asl   on accoun is th   potential their e lar t  bal gmatch e asize  responses nces   — of   for  up some   401 fr om from  of (k)   38 the   plan       13  9these ? Social  perce  Se ntages curity.  have     often increased to the high 80s or low 90s.  retirement study.” Journal of Public Economics, Vol. 91, Issue 10, pp. 1920?1943 (2007).  The participa  August te)  2for 011  more  Notes  tha  article n 30 years,  evaluated  the media  the n im TM  multiples portance  range of defin  from ed  be approxi nefit pla mately ns for  7.6–8  househ .5 times olds,  final  retirement forthcoming under automa   income EBR tic Ienrollment  publication.  adequa 4 4  fa cy  ll  would because  be  many  4.3  per newc ent, partic  to ip an accoun ts whot  for would  the  have  financial  voluntarily  and housin  choseng  a  market  higher    This is true only in those cases in which the defined benefit participant ends up with an annuity in retirement.   ? For 5.       those  $250  with  million–$500  the lowest  mi  inllion. com  e, employer matches have little or no effect on participation,  on contrib     the      Re      participa ? .tireme  “u  The (401(k) The tion   average n  EBRI nrate tts  In - ty de come   of Re pe) crease  tirem 18.8  2009  plans.  is  currently perc dent  first  equities  Howev  Readi ent -tier  would   integrating ematch n  by r, es   this smore   Ratin be  ra  datab  sufficient te  tha g:   administrative was na  se 10 Ret  87.78  does   percentag to irement  cover   ce not nts    recor  report In 75  efor come   poin perc  d each s  the  t  of ent Preparation s   duri  dollar millions year  of nthe  gthat   cthis  o  employees of n  tth  and  period ribut 401(k) e au  Futur ed tomatic   bu participa ,  while covered et  that   th  was nts  eby       Emplo 33 This was yeeBenefit   conf Beneirmed fit  Resea Researc  byr ch Ya h  kIn Institute ost boski itute  (and EBRI), , November  Vathe nDerhei EBRI  20 11). (19 Education 96)    when and  the Resyearch  analyzed Fund , the any  401(k) of its progr  parti am cipant s, officers,  data  immed perce (as contempl ntage iately.  ofated    comp  in ensati the Gale on  level proposal)  of co.ntribu    tions.   participa probability tion.  of  retirement income adequacy, the additional savings that would otherwise need to be set   See VanDerhei and Copeland (2008).  perce sponsors employers participa nt ju nts  with st  that   in 10   respect th   months have is age  recently  to  group.  ea  this rlier.   type  annou Mo  Asreo   of forn  proposed v  ce em erd,  pthe pe loye nsion  lowest  rmodi s with  freezes, -fication income  fewer  and    of than quartile   th th ee1,000   freque 401(k)  no  workers, longer  ncy system.  of  has  the  Fidelity  A  the   an survey n  large o  noted un  condu cem st redu  that ent cte ction, s.  the d         Papke, Leslie, and James Poterba. “Survey Evidence on Employer Match Rates and Employee Saving  salary, depending on salary level.   assuming contribution  the  rate y retire  remain  at  at  age  the  65,  low  while  default  demonstrating  levels. Addition  the al research  impact  of in  th define is fieldd  has ben  be efit en  pla  cond nsuct  in eachieving d on a    61 crisis Those ?   The  in who  2008  a nnu hav  and eal  been  increase  2009.  cash     in ed  c o out ntri  or butions  chose  (1 to  vs. take  2  per a lump cent  sum  of co  distributi mpensation on would ).   still need to deal with (post- trustees, The Although Figure  2012 5.2.2  13 while spons    there RCS pro   v   ors, au showed ides Impact  do toma or   not evidence other  tic  of  that appear   enrollme Em staff.  workers  ployer from  Tto he  be  nVanD Employ  Ma t   often  any has tch erhei,   major dra ee  egu s Benefit m on ess  Holde ati   trends 401(k)  at c effects,  Re ho n, search w    by Alonso Saving  mu  age,  bu ch Institute t     ..................................................................   Fand they igur  Bass  ewill is  2a  shows   nee (2011) nonpr d ofi to  that  on t,  anonpartisan, ccu   th the emulat   average lower e (4 -education inc  2asset  o percent), me    and  18    with partially One Even, ?   of Defined those  Willia  the Prospects.” average enrollment  offset   of fa m  ctors  contr  IRA E.,  by  2005  an   account 4  ibution tha    dprovision per EBR  David  tfirst  cmakes Ien  Issu   -holders. balances. ttier   A. of e  was      Macphers aBrief, the match  major  adopted,    participants One   no.  rate  difference   of 344 on.   was the   so  “The (Employee    there  first 81 who   .in Effects 2 6 publica   the  increased is  ce  no   nts perc Benefit  of  wa  t  for ions Employ ent y   e to each   aquities Research fr  ge tell om er   of dollar  from   Ma this  househol  by tchi   Institu   endeavor 10 contributed. this  ng percentage  data  in ds te,   401(k) satisfying   July  source (schedul     2010). Th   points Plans.”  ehow   differ ed the      for or long    e   more. nce later    of      this from  ty  three pe of  la plrge an.   401(k) The me  sponsors. dian rate  Mo  forreo  a fivner, al-  they average  found  plan  a  signifi is larger, cant  as  amount  expected:  of clustering  13.5 percent,  around  and  the  the     Capital gains treatment is not used in this version of the model.   aside each year until retirement to achieve this objective would decrease by a median amount of 30  6.      >$500 million. 3   34 on applicabl as     the behalf  reduction e  of  perc  Theen   for Principal tage  the  was  second  Financial  over- income 46  per Grocup  ent. quartile  (2011)     is  determined  slightly larger  that  at  if  13.3  workers’  percent.  ability  The  to  reductions  deduct any  for    this      relatively        . Testim Behavior  smallony.  sample  in  U.S.  4 01(k) of  Congress.  401(  Plans.” k) plans  Sena  Economics  in Mad te Finan rian  Lan ceetter d  Committee.  Shea s 49  (2  pp. 001)  3; 13–317  Tax Cho iRe , Laibson, form  (Sept  Options:  eand mber  Mad  1995).  Promoting rian (2  004) ;Ret  andirem  Choi, ent     retirement The retirement)  parameters   investment income  of  aad  mode  equa riskth  and lcy  for     for longevity  the  Baby  first   Brisk  ioomers ncrement  on their  and   own ca  Gen n .be    Xers.  estimate   d from the entire sample by dividing it  Prior In Septe  to estim mber a2011, ting the  the  potential  U.S. Senate  imp  Fin acta  on nce  accumula  Committee tions  held  resulting  a hearing  from  on  401(k)  “Tax Reform  contrib  Op ution tions:  changes,    a  researc AnThe  important  full h orga  stochastic n icaveat zatio nnat estab  isu re tha  lis ofthed   th most e in model W  available ashin  will gton,  be  survey  in DC, clude in  da d 1 9in ta 78.  future  do EBRI  no  anal t does con ysis. tnot ain    take detailed polic yi n positi formation ons, nor  on does  plan it  lobby, rather While allocation   itthan  is  of obvious  doing  401(k)  a  that   systematic part  pe icip nsion ant sretirement   plan from  fr  th eezes e EBR  needs  affect I/ICI  calculation.  Participa  some workers nt  The -Directed   propensity negatively,  Retireme   to it  is guess nt  not  Plan   obvious or  Data do their   Collection which  own       households retirement ? Among auto  income  ar -enrollment e  workers MU  ad CHequa  more  in  had  the cy   likely  been thresholds  mid  d to impl le  be  income eme   at at  risk any nte  gr   for dretirement oup,  in  ins  a  plan employer ufficient .    age  retir   ismatches  whether ement  have   income the  substantial worker  (eve  is n  thou still  effects  participating gh we  that       in threshold  2012)6.52 Industrial  will  ra te  be ce  for  nts to   Re th  investigate for elations:   75 each  perce  dollar  A  the Jour ntile   ccontributed nal hange  increases  of E  in con  asset omy  suggests  to  allocation 21.0  and  percent. Society,  that ,at  to   retirement  Vol.  the Cash   extent 44,  balance  Issue   and that   3, pl  to   this ans pp  track .   52 have sample 5  subseq  –  a549  med  is  (2005 uent ian  ).  62 match cap. For example:   ?  Whether employees are assumed to opt out of the automatic escalation.  Looking only at those households that had exposure to the market crisis in 2008 and 2009 from all three  percent. Additional refinements were introduced in 2005 to evaluate the impact of purchasing long- VanDerhei 6  Appen  and dix  A: Lucas  Brief  (2010  Description ) demon  of st rRSP ateM  th  .............................................................................................. e profound influence of plan design variables, as well as   24  amount     While age    Van ?         cohort IRA D .  it“erhei     “The is of balances.  dif   th dec  and f  eImpact icult  r401(k)  ease Copeland   to   to of  predi contribution   11.4 PP  (2Ac00 t   on  per the 1) .Ret  c en extent  ifrom rement t for   to those taxable  which  Inco  inme  inco   the market  for  me third  401(k)  was  volatili -in ce  oParticipa lim mtyein   over quartile atedn  the ,ts.”  65   and past   per EBRI  cthen  several ent  Issu   of decrease e  Brief, the  years  pl  no. an   will to  sponsors   318, 8.7  have     a   Laibson, Ma Security drian,  (T and -17  Metri 0). 15 ck  Sep (200t6) . 2  011.  Promoting 10 into advoc set of  two ate  baseline  spec grou  Ret ifps: ic i rement results polic  those y  rec first   Security.” who ommend  needs  make at   One to  ions, the  be   of contribution  or ru  the receiv n to  pri  de e m fede termine ary  an ral  topic d fund those  the sth  ing. duri  likely  who ng   the values do not.  hearing  if  Th  the e  was parameters  various  an assess  tax  reform ofment  a model   of the     35 design. In an attempt to mitigate this problem, Mitchell, Utkus and Young (2007) use 2001 data on 500    In model workers calcula Project  February  tour  for ion,  are   ba selected    of toget affected, sic  2012,  retirement her  years   the with  nor  2IRS    to betwee current   expenses  interim whatn   fee de  199  report g  lings as ree 9  aand   they  function  of of   financial 2010, responses  are affected   of in c 67  the stress, lu sfrom  ihousehol ve  by .   may  Al its  a though pe  4 01(k) help nsion d’s expe  to   any Compliance  freeze  explain  cte time d.  There retirement - series why  Check   the are  comparison    amo ma Q  income). uny estionnaire unts  reasons   that of  The        changes Park, in a defin  Young. representative may  as ed  retirees   contribution  be Retire  larger men  age.  than  of t     In A  plan th  c  follow th oem  e  un after  eeffects iverse  Adequacy -up  age  study   of of  65.   au large    is With The toma  plan  401(k)  incre  Im tic ned mediate  enrollme ase   sponsors, that  in  will the  and n  t.begi per    those Longevity cnenta  to  sponsors  lige nk   A of  successive n  households nuities,  adopting  May  years   AE tha  2011   were of t are  the  EBRI   more       indemnifica  In recent year tion s  the contri  longevity bution  risk  rate  may  of  also 4.6  percent, be dealt with  with  vi  a 83 a  75 longevity  perc eins ntile urance  threshol  or longevity d rate  an of 28 nuities 6.3 perce .  Seent  Pa   rk    It is important to note that the annuitized accumulations in this analysis are from 401(k) contributions  fronts (defined contribution plans, IRAs, and net housing equity) shows a median percentage for Early  term care insurance on retirement income adequacy.   assumptions Facts 63 The Finally, perce responding long  authors  ter  nt from  th  m (Employee for e  imp   EB   Sthose to  ran  eof RI. pte  athe  cemployee   tregression “A mber   on  in survey  Be   History  the inn  divi 20 e  hi f iwould 11 t  ghest behavior duals’    Resear analysis of  Senate  401(k) -   have income risk ch    in on Fi    In tol Plans:   le nau  st this ass e  nce quartil itu trance o  desire  - tdatabase enrollme An  etesti ,  June  Up or e.  to   m  ass dat ony   contin 2008  neand ett.”   401 allocations,     ). analyzed  FS produ ue   (k) -185  offering  pla c e(Employee n  dthe s.   the a     Even finding their  potential  incr   with 401 easing  Benefit  that (k)   aim  : relatively   plan utiliza p  Research act. oft   ion  various  simple  of  Institu  AE  types   since te,    46 ?  Whether employees are assumed to remember/retai 15 n their previous level of contributions when they  ? Company A had a maximum pretax contribution of 9 percent of earnings and a match rate of 30  options for the second  are not  increme  imposed n ton  can  th  be e current  estima ted 401(k)  by  dividi system. ng  th The e  subsam model used ple of  in  those  this article  who make  is base  thde  on first  the     potential benefits and consequences that may result from a proposal to modify the federal tax  401(k)  VanD erhei retirement  and Copeland  plans covering  (July 2002  nearl ).  y 740,000 employees  to evaluate how employer matching  for revealed workers asset 2012  This  this,   baseline  allocati result   most say  the  is  they  on numb   fr  importantly ratings om  from  ne  ae ed rregression    this of for  to  40    Early  database athe 1(k) ccu   un  on mulat  Boomers plan  iq a ue sample  n  spo eeed   characteristics fon  rranges s  sors  a  of to  comfortab   all be  that:  parti   from acco c  ipa  mpanied  and al eprojection n ts retirement  terms  (whether  by  of  the   that each   cappear o  cavea ntributi  87  pension  per t  to  that ncg  be e  or nt  plan   the  rather  not) of   the univers an  for  dlow.   lowest  whom each e  Th  of   freeze, -match irt  income data y-four   ra    t e  7 predicted (2  01 ? 1)Defined Appen  for   more tod  ix have bene  B:  detail.  Brief  adequate fit annuities    Chronology  retirement  and/or  of RSP  lump  income M -.............................................................................................. sum  as  distribu  a resultit ons. of defin   ed contribution participation varies  by 25   integrated     using exclusively         .the  “The gen Issue  cur  and  de  eIm rrous  fined Brief  ent do port  not   in to   an #3 contri t  ethe include ce 57 re st   of 401(k) b  credits.  uDefined  tion/IRA projec  par  ted Th t  Benefit icipa  ese data  Social  nts values  and   Sec Plans  when  track u rity incre  for  measure  re  the  ase Retire tirement  spend  to men d5.7  by  be - do per t  nefi this  In wn ccts. ent  ovariable  be m  This  eand havior  Adequacy.”  is  7.3 in  after   of co  perce  ntrast retirees  au tom EBRI nt  to  if,  a  other as   tic No ins  a  ten t  efunction  ead, EB srol , RI no. l  m the  en 8   of t   Boomers 5.2.1.1.2  ofParticipants  5.6 percent  for a  50 percent probability and 6.7 percent for a 90 percent probability of  of the definition It should  tax  passage -reform  be  of   not of “success,”  options  the ed,  Pension howeve  on  large  re tr, Protection i r  differ that eme nthe etnce  income   author Act s in  of  success   adequacy.  2006 use NCS  has  rat   microdata eresulted   sTh  can is  was be  in  seen,   fro e  more xpande m  de the  401(k) pen  drespon  indi  the  ng assets  dents on  Novemb  which  investe  initia er  plan te  d2011  d in   desig in  targe   EBRI 2002 nt     chanFebruary ge jobs vs  200 . reverting 5). http://www.ebri.org/pdf/publi 1  back to the plan’s initial default. catio   ns/facts/0205fact.a.pdf   percent for the first 5 percent of earnings. A total of 21 percent of participants contributed 5  treatme 401(k) incremental  voluntary nt of  contribution  401(k) -enrollme  plan  contributions in ntto  modules  those who  fr  in om   make exchange  RSP M the .   It  nex for  is  tasimilar  1  flat  per -rate c en in man t  governmen  of comp y respects ensa t match. tion  to the  con  Gale  one tribu   us (2011) tion ed  in and      64 incentives and match  level influence  information  retirement  was provided  saving.  or Their  derived.  analysis  The  regression of the impact  mod  of el  incl emude ploye d age, r matchin  tenure,g  salary,  contribu  plan tions  loan    A perce and providers 5.2.1.1.3.2  separate  the nt  sugg age of  (as    wor survey  an ests Plan  well d ke  characteristics  r  a  Size  sas by   nsay  e  plan Allia gative    th ey nceBernstei sponsors   relatio nee  of d  to the  nand ship  save n  wor   in participa  between   less 2011 kers.  than   provi VanDerhei n ts) au  $ 250,000 tomati dhas ed  gradually plan  (M c  enrollment (up  arch spo  fr n   2006) chan om sors  26  with ged   an provid  per d  over   the match cees nt   following this   in a  rates de  2007).  time tailed  and   pe qu  An  analysis riod, est  isot  ihon  ethe r  and  18  of           11 households research        . “ERISA  (e.g.  ar , Van  At e at  30: D erhei risk  Th  to  ean   Don decline  Lucas ly 13 , of  perc Novem  Private entb er for - Se 20  the ctor 10 )high  tha  Defined test  includes  income  Benefit  bot  h h o  Promises comp useholds. onents.  and   Si  Ho milar  Annu weve  tre ity r, nds in  Payments:  th  are e previous       several retirement cash balance  demo (Employee was  age,  i m plans graphic p  plem re   are Be ren et  and ni trassumed eeem fdi tportfolio   tha Resear entn  income  they  toch   characteristics. cred   were In st quartil ititu  interest  before. te,e  August  and   at    probability  the  2011):  interm  7–16.  of edi  retirement a  te long-term  inc oassumption me adequac  for y,  bu thet  this  The model84  was next used 58  in March of 2006 to evaluate 23  the impact of defined benefit freezes on  85  VanDerhei and Copeland (December 2002).  retirement income adequacy.  Younger cohorts experience a similar increase, going from the all-  Those in the lowest income quartile will be more likely to benefit from Medicaid while those in the highest  and  ?2003, Suspend  and tha edt  only or discontinue  a small per dc  matchin entage of g contribu  the plans tions  in the  in  their sample  plans  (6 perc  increased ent) we  from re governed  1 percen  by t in     Issue 8 24 date factors PLANSPONSO   ?  Net Referen  fu Brief  nds and  housing    du em  and cR.com. es ep   to  l............................................................................................................................... aoye  equity   new the e  “DC   behavior qualified  set .  D   of eferral  survey   assumptions defau  Amount  results lt investment  Wh  were  are at  use Participants  adde  alternative d:d  to the  Can   regulations. mod  Afford.” el in th  HR/Be e    Mo March reo nefits,  v2012 er,  given June  Note   23 the s ........... ,article.   2010. relati  ve 27     29 percent of pay to the plan and 45 percent contributed 9 percent of pay while 1 percent  update those Holden provision  who d  and   a (yes   20 do  V/06 a n  not. o), nDerhei  analysis  emplo  Successi  (20 y er by  02) match  ve Gale,   iterations in  tha  ra Gru tet, be  itand  looks r  are ,  and emplo  estimate  only  Orsza yer  at  match g  current  dand  un  til level analyzed  the  401(k)  variables  max   aparticipants i m plan  utom   examine that  plan  wou  limit  and  their ld   of does  change   effe all  not ma cts  tch the on  attem     pt to  on With  participati  respect o ton  potential included  two worker  important  reactions  innovations:  to this proposal,   First,  they a new  evaluated  set of ques  empl tions oyee  conc  saving ernin  behavior g     VanDerhei (September 2011).  how Gale, perce evidenced potential overall analysis,  pension  nt Willia  asset   stati What th mention  res e  for  m exp sti  p  allocations  freezes   G. onses: Will both cally e  r iTestimon a e   It n  goal  th csi  Mean?”  eare gnifican e   of   Late of   to  al thli kely l$250,000–$4 y.   equi workers   Boomers  U.S. EBRI t  to at ty    Congress. impa  the  fu Issu  (not nds   firm and ect   ju 99,999.  Brief, per  existing st -  Gen le    those  ve se Se  no. l  nate .Xers. have   Twen In    who employees 269  par  C   fluctuated  oty  were (Employee ticul mmittee  per  ar, curre c en as  match  t   as a ntly on  thi   Benefit f unction one  Finance nk  40  rates  1( they  would k)  Research    partici are of  .nee  Tax   plan expect  ab d  pReform  to ou ants)  type   Institute, tsave   7 with   was  pe and   Op $500,000– rc   changes simulated  etem in  on May tapsge ;loye   2004).   poi in  an e  the  nts d       factor alone results in at least a 10 percentage point difference in the majority of the retirement  participa interest rate nts  by of  simulating the Treasu 0  ry the  special  minim  pu umblic  employer -debt obligation -contribu  bon tiond  rate s issuable  that would  to the  be  OASDI  neede  trus d to t funds,  financially  as    65 household income quartil EBRI 110  analysis e are more  to0  the13  likely  more St. NW  be  able select #800  to  gself roup. - W insure ash   in  the gton  risk , DC 20  without 005  a  cat (202) 6 astrophic 59-0 6impact 70 w  on ww  th .eebirri.o  futrg ure  An interesting 2006 to  findi  4 perc ng of en  the t in  AllianceBe 2008. 32    rnstein25  survey of plan sponsors with33  respect to potential federal  inertia automatic  exp eenrollment rienced with  provisions.  respect to    asset 54 allocation for participants automatically enrolled in 401(k)  www.plansponsor.com/DC_Deferral_Amount_What_Participants_can_afford.aspx    VanDerhei contrib  andu  Copeland ted up to  (2  the 003)  402(g)    maximum for that year. The average deferral percentage for  participant before-tax contribution rates.   treatme formulas include participa  enn  ligible tist  of   behavior obtained.  retirement  nonpartici  in  In  res  th  saving pis pants onse  model,  in   to or  thr    the the workers ee   specific decision ways:  who   federal  of  are  an  currently  eligi  tax bmodifications le employee  not eligible  is  proposed  .ex  am However, ine d in  at  Gal   each unlike e (2 011) of  the  level  was  2002  of      25 separately job cha ? nge The   was fo  average r  NHCEs allowed.  effective  and    HCEs  match  at 1–10 the  rate  firm  level. for 2009  Second, 11–20  was  4.32 in an  perc  attempt ent 21–30  of  to  compensation,  deal with nonlinea 31–40  but only r 401(k)  4.00     demographi $999,999, equity marke lower  while cs. t s. amo       fewer Overall, ng firms  tha  53 n  with   1percent  in  10 au tomati ea  ofch  401(k)  beli c enrollme eve  assets  theynt  in  need   th the an   to EBRI/IC  amon  saveg  I$1   those database  million–$1.49  without  were  au  in  million to  equity matic  (6   funds enroll  perc  ment, at en t)    A household Promoting  is considered  Retireme  tont  run  Security,  short of  15  money  Sept. 2011,  in this  online  model  at  if   aggregate resources in retirement are  9 specified age/income   Endno  intes  the  combinations  ...............................................................................................................................  2005 Trustees  investigated.  of the OAS   DI Trust Funds Report (5.8 percent).  .............. 32  indemnify the employees for the reduction in their expected retirement income under various rate-of- retirement  Gale (201  income. 1).    tax modifications is the impact of plan size on the expected plan sponsor response. 34  The reasons to  plans, this is likely to be the case even if AE utilization stays constant in the future.  Therefore, there is  66 Company A was 6.7 percent.   model, possible  this  contributions  analysis assumes  for the  no  employe  job turen.over,  Conseque  withdrawals, ntly, the  or chang  loanes  defaul  in thets.  ince   ntives of contributing an  matching perce  formulae nt of co  (empensation xplained in more  in 2005.  detail   The Num  la  ber ter), in c ofr Year e  they ase s Eligible  of  bif  0.32 urcate to Par  perce ticipat d th e in a netag  formulae 401( e points k) Plan  into  again  an  sugg “incentive ests that     not or  Van  $1.5  sufficient Derhei  millio http://finance.senate.gov/imo/medi after  (Ja  n  cto  o nuar or n  meet t  more roll y 20 in  aggr 04 g  (9  for ).  perce egate  firmnt)   mini ch aracteristi (Figure mum  6). retirement a/doc/Testimony cs.  Savings     goals expe nditur tend %20of%20William%20Gale  to es,  increase  which are  as  defined household  as.pdf   aincome  combination     rises.    Source: Source:EBRI/ERFRetirementSecurityProjectionModel,®version100205a4.Seetextforexplanationsofmodelsandassumptions.   32 19 15 20 13 35 23 26 11 24 34 33 10 18 29 37 36 16 27 31 25 17 12 30 21 28 14 22 7 6 1 2 5 8 3 4 9   Figure 18 Figure 12 Figure 11 Simulated Impact of Proposal to Modify the Federal Tax Treatment of Figure  5 Cumulative Distribution Function of the Percentage Cumulative Distribution Function of the Percentage Figure  7 Figure  6 Employer and Employee Contributions for 401(k) Plans In Exchange for Figur Figur Figur Figur Figur Figure e e e ee       2 1 3 4 9 14 of a Worker’s Annual Pay Needed to Offset the Impact Figure 8 of a Worker’s Annual Pay Needed to Offset the Impact an 18% Match From the Federal Government for Employees Currently 26–35, Workers Having Tried to Calculate How Much Money of a Pension Freeze in 2006, by Pension Plan Type Figure 19 Estimated peof rca enta Pension ge ofFreeze  consisin te2006, nt participan by Pension ts who Plan T  ha ype ve more money  TM Impact of Income and Relative Value of Defined Benefit Accrual EBRI Retirement Readiness Rating (RRR): 2003 vs. 2012  Timeframe When Retirees Began to Plan Financially for  Amount of Savings Workers Think  2012 Impact  Unconditional  of future years  Retirement  of 401(k)  Savings  eligibility  Shortfall*  on Figure  2012  numbers  at-risk* 17 by ratings  age cohort,  for Gen  martial  Xers by  status  income  and  quartile  gender by Plan Size and Age -specific Salary Quartiles: Midpoint Estimates TM (assumes 8% annual rate of return) EBRI (assumes  Retirement4%  Readiness annual  Rating rate (RRR): of return)  2003 vs. 2012  (Status Quo for Social Security, Housing Equity Used "As Needed")  2012 Conditional at Retirement Age on At-Risk* Probabilities  Retirement Savings Shortfall* numbers by age cohort, martial status and gender in their 401(k) accounts on 3/1/12 than at market high (10/9/07,) by age  Simulated Impact of Proposal to Modify the Federal Tax Treatment  $80,The 000 y Need to Save for a Comfortable Retirement (Status Quo for Social Security, Housing Equity Used "As Needed")  100.0% Percentage of population at risk* for inadequate retirement income, by age cohort and income quartile (baseline  Percentage of population “at risk” for inadequate retirement income, by age-specific remaining Assumption for this run: Employer increases or decreases to contributions are 1 Figure 10 Retirement, Among Retirees Who Planned for Retirement They Need for Retirement  $160,000 100% Predicted Employee Contributions for Selected Persons and Plans Percentage of population at risk* for inadequate retirement income, by age cohort (baseline assumptions) career income quartiles and income-sp and assumptions) ecific  tdef enur ined e benefit   value quartiles (baseline assumption) of Employer and Employee Contributions for 401(k) Plans In Exchange represented by the midpoint of the range denoted on the AllianceBernstein survey 90.0% Percentage of Various Work Forces That Participated 100% 90% for an 18% Match From the Federal Government, by Age and  $70, 1000 45 0.0 0% % in an Employment-Based Retirement Plan, 1987–2010 0.9 100.0%  $140,000 80.0% 90% Age -specific Salary Quartiles: Midpoint estimates 40% 80% 90.0% 80% 90.0% 70.0% 95% 35% Respondent 2005 2007 Respondent 2011  and/or Sp2012 ouse Income-specific  $60,000 Assumption for this run: Employer increases or decreases to contributions are   $1 0.8 20,000 12 77.6% 80% Defined Benefit 77.3% 77.1% 77.2%77.2% 75% 76.7% represented by the midpoint of the range denoted on the AllianceBernstein survey 30% 76.3% 76.1%76.1% 75.8%75.8% Value Quartiles 60.0% 75.6%75.6% 80.0% 80.0% 75.5% 75.3% 75.4% 75.2% 2% 74.8% 74.8% 70% 74.5% 74.5% 2012 53% 30% Career Average The year you retired No DB accruals 25% 73.3% 72.9%  $100,000 Career Average  $50,000 70% 0.7 71.9% 50 90.0% % 2% 10 70% 70.0% Lowest Quartile 70.0% 20% 48% Final Average 2011 47% 60% 34% 2 46% Final Average 25% 40.0% 15% 44% 44% 65% 32% 60.0% 1%  $80,000 43% Cash Balance–Cu43% rrent Interest Rates 3 31% 60.0%  $4 0.60,000 The year before you retired 42%42%42% 42%42% 8 60% 51% Cash Balance–Current Interest Rates 60.3%60.4% 10% 85% 59.8% Highest Quartile 59.6% 30.0% 59.5% 5% 50% 59.2% 59.0% 58.9%59.0% 59.0% 20% Cash Balance–Long-Term Interest Rates 58.4%58.3% 38% 50.0% 58.3% 60% 57.9% 57.7% 5% 57.1% 26% 56.6% Cash Balance–Long-Term Interest Rates  $60,50 00.0% 0 56.7% 45% 20.0% 55.3% 0.5 50% 6 54.8% 54.8% 401(k)  $30,000 12% 54.5% 54.4% 0% 40.0% 40% 42% 15% 2 to 4 years before you retired 55% 22% Plan Size Lowest income quartile 2 52.7% 3 Highest income quartile 80% 21% 51.6% 40.0% 10.0% 39% 9% 51.1% 20% 51.0% 20%  $40,000 <1M 36.4% 28.8% 22.8% 26.5% 19% 49.8% 37% 30.0% 48.2% 49.1% 48.3% 18% 18% 18% 0.4 48.2% 40% 4 48.4% 48.4% 50% 47.8%  $20,000 30% 10% 1 -10M 47.4% 40.1% 32.4% 26.9% 31.5% 0.0% 47.0% 47.4% 47.2% 46.8% 30.0% 46.5% 33% 46.7% 17% 46.2% 46.2% 46.1% 46.1%46.1% 46.0% 32% 32% Lowest income quartile 2 32% 3 Highest income quartile 45.5% 45.1% 31% 20.0% 5 to 9 years before you retired 44.9%  10 $20, -50M 000 22.8% 13.7% 7.4% 44.8% 12.8% 75% 44.2% 43.1% 13% 29% 43.1% 0 26-35 36-45 46-55 56-65 45% 86.8% 53.6% 13% 34.4% 16.8% 0.3 43.0% 30% 42.2% 43.2% 50 -250M 5% 20.2% 42.1% 11.4% 41.7% 3.3% 8.5% 20% 2 11% 42.0% 20.0% 41.3%  $10,000 40.9% 10% 10% 10% 40.8% 40.7% 1-4 10.0% 99% 40.4% 99% 44.4% 99% 98% 1-9 72.3% 46.8% 28.8% 40.3% 14.3% 40.0% 40.1% 39.8% 9% Age=55,Wage=45 Age=40,Wage=25 43.8% Age=35,Wage=20 Age=22,Wage=15 43.6% 39.5% 250 -500M 20.2% 10.4% 3.2% 8.3% 39.2% 43.0%8% 8%  $- 40% 27% 7% 42.0%41.9% 7% 7% 5-9 41.7%41.8% 41.8% 10-19 97% 97% 97% 95% 64.7%Early Boomers 37.0% Late Boomers 18.4% 41.5% Gen Xers 7.6% 41.0%41.0% 6% >500M 10 to23  19 .5%  years before you retired 12.2% 40.9% 6.8% 13.1% 10.0% 0% 40.4% 0.2 10% 20% 0.0% Income Group 39.7% 39.7% 39.8% 39.6% 39.4% 26 -35 36 -45 46 -55 56 -65 39.0% 39.0% 10-19 38.9% 95% 93% 92%23% 91% Sing20 le+  Male 61.1% $94,5Earl 09 y Boomers 28.8% $10Late 3,91  Boomers 8 13.2% $12 Gen 9,39  Xers 58.4%  $- 38.0% 37.6% 35% Source: Author's calculation 100%, up to 5% s based on results from EBRI Retirement Security Projection M75% up to 6% odel Version 1472, and responses to AllianceBernstein (2011) Early Boomers Late Boomers Gen Xers Lowest income quartile 22.2% 24.9% 21.1% 12.7% Lowest 20-29 income quartile 88% 88% 87% Single Female $104,799 86.8% $112,1820 3.6% $133 77,3.7% 49 0.0% and Employee Benefit Research Institute and Mathew Greenwald & Associates, Inc., 2012 Retirement Confidence Survey. *An individual is considered to be at-risk in this version of the model if their aggregate resources in retirement are not sufficient to meet aggregate minimum retirement  0% Single Male Earl $3y3,  Boomers 704 Late $3 3, Boomers 420 G $4en 1, 52 Xers 9 2 13.0% 7.2% 9.9% 34% 13.3% 0.1 10% Note: This simulation models only the financial impact of the expected reduction in 401(k) account balances for employees who are not automatically 30+ 88% 86% expenditures Fam 2 ily  defined as a combination$7  of0,  de44 te0rministic 48.0%  expenses from the Consumer Expenditure $75,84  Survey 496.9% (as a function of income) and some health $8 3, 45 insurance 97 .8% 0  and  Lowest 2 3 Highest 2050% up to 6%  years or more before you retired 75% up to 2%, 50% for 3%-5% 30% 3 enrolled by modifying the behavior of plan sponsors and participants and does not attempt Income Quartile to assess behavioral modifications on the part of eligible Sin EBRgIle  20  Female 03 RRR $651 4,74 6. .7% 1% 9 10.0% $647, 8.05 5%7 11.6% $7 51 5,.7% 82711.4% out-of-pocket health-related expenses, plus stochastic expenses from nursing home and home health care expenses (at least until the point they are picked up by Medicaid).  43% 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 3 29.3% 26.5% 29.3% nonparticipants. The simulated rates of return are the same as in VanDerhei and Copeland (July 2010). This version of the analysis assumes no job turnover, ® The resources in retirement will consist of Social Security (either status quo or one of the specified reform alternatives), account balances from defined contribution plans, IRAs  Under Source: EBRI/ERF Retirement Security  $250,000 $250,000 Projection  Model to version 110714e. $500,000 to $1,000,000 to $1,500,000 or Don't  Highest income quartile 10.8% 17.1% 14.1% 8.7% Ma EBRrried I 2012 RRR $244 1,67 .3% 2 $240, 3.73 3%8 $2 43 5,.9% 018 withdrawals or loan defaults. 0% The full stochastic nature of the model will be included in a future analysis. Plan sponsor and participant reactions to the Highest 0%0  income quartile 12.5% 11.2% 16.7% * An individual or family is considered to be “at risk” in this version of the model if their aggregate resources in retirement are not sufficient to meet aggregate minimum retirement and/or cash balance plans, annuities from defined benefit plans (unless the lump-sum distribution scenario is chosen), and net housing equity ( in the form of a lump-sum  All Workers All Wage and Salary Workers Ages 21–64 Sources: EBRI estimates based on tabulations from EBRI/ICI Participant-Directed Retirement Plan Data Collection Project. The analysis is  19931994 proposal 1995 are explained in the tex 19961997t. 1998 Employ1999 er increases or 2000 decreases to 20012002 contribution 20032004 s are represented 2005200 by the midpoint 620072008 of the ran 2009 ge denoted on 20102011 the 2012 expenditures defined as a combination of deterministic expenses from the Consumer Expenditure Survey 6% (as a function of income) and some health insurance and out-of-pocket know/Don’t  $499,999 $999,999 $1,499,999 more distribution). Source:  This Author'  version s calculation  of the model s  is based  constructed on results  to simulate from EBR  "basic" I Retirem  retirement ent Security  income Projection  adequacy; however, Model Version  alternative 1471, and responses  versions of the model to Al  allow lianceBernst  similar analysis ein  for  0–5% 0–5% 5–10% 5–10% 10–15% 10–15% 1 15–20% 5–20% 215–20% 0–25% 220–25% 5–30% 30–35% 25–30% 330–35% 5–40% *The Retirement Savings Shortfalls (RSS) are determined as a present value of retirement deficits at age 65. Full-Time, Full-Year Wage and Salary Workers Ages 21–64 Private-Sector Wage and Salary Workers Ages 21–64 Don't know/Refused health-related AllianceBernst expenses, ein surv plus ey stoc . hastic expenses from nursing home and home health care expenses (at least until the point they are picked up by Medicaid). The resources in based on all participants with account balances at the end of 2007 and 2008 and contribution information for those years. (2011) and Employee Benefit Research Institute and Mathew Greenwald & Associates, Inc., 2012 Retirement Confidence Survey. *The replac ement Retirement  rates, standard  Savings -of  Shortfalls -living and  other (RSS)  thresholds.  are determined    as a present value of retirement deficits at age 65. Sources: retirement  EBRI w ill Retirement consist of Social  Security Security  Projection (either status  Model® quo or one of versions the specified  1501 and reform  1502. alternatives),   5% account balances from defined contribution plans, IRAs and/or cash balance Public-Sector Wage and Salary Workers Ages 21–64 Percentage of Compensation Sources: Sources:  EBRI  Retirement EBRI Retirement  Security  Security  Projection  Projection  Model®  Moversions del™ versions  1501  and 1501  1502.  and  1502.  remember Note: This simulation models only the financial impact of the expected reduction in 401(k) account balances for employees who are not automatically Source: VanDerhei and Copeland, "A Per behavioral centage of C m oodel mpensfor predicting em ation ployee contributions to Sources: plans, Source: annuities from     EBRI EBRI  Retirement Retirement defined benefit   Security Security plans   (unless Projection Projection the lump-sum   Model® Model,® distribution sc versions  Version  120201. 1501 enario is chosen), and  and    1502.  (in some cases) net housing equity (either in the form of an annuity or as a * See text enrolled for definition by modify  of ing "at the risk" behav   ior of plan sponsors and participants and does not attempt to assess behavioral modifications on the part of eligible * See text for definition of "at risk"  lump-sum distribution). This version of the model is constructed to simulate "basic" retirement income adequacy; however, alternative versions of the model allow similar analysis Source: Employee Benefit Research Institute estimates from the 1988–2011 401(k) plans." North American March Current Population Actuarial Journal Surveys. (First Quarter, 2001) Sour Sour Source: ce: ce:        Emplo Emplo Employyyee ee ee   Bene Bene Benefit fit fit   Re Re Ressse e ea a arrrccch h h   Ins Ins Institut titut titute e e   and and and   Ma Ma Mathew thew thew   Gr Gr Green een eenw w wald ald ald   & & &   Associa Associa Associatttes, es, es,   Inc., Inc., Inc.,   2011– 1993– 2005–2012 2012 2012   Re Re Retttiiire re rem m me e en n nttt   Con Con Confffidence idence idence   Sur Sur Survvve e eyyys. s. s. nonparticipants. The simulated rates of return are the same as in VanDerhei and Copeland (July 2010). This version of the analysis assumes no job for replacement rates, Source: Source: Author' Authors' sstandard-of- tabulations tabulations living, ffrom romand other ad tthe EBR he EBRII/ER /Ehoc RF F R thresholds. Ret etirement irement In Income Projection come Projection M M ode odel. l. 1 3 1 2 turnover, withdrawals or loan defaults. The full stochastic nature of the model will be included in a future analysis. Results for participants currently older than 35 are limited to high-tenure participants as explained in the text. Plan sponsor and participant reactions to the proposal are explained in the text. Employer increases or decreases to contributions are represented by the midpoint of the range denoted on the AllianceBernstein survey. Percentage At Risk of  Cumulative Percentage of Employees Average Percentage Reductions in Inadequate Retirement Income Percentage of Employees 401(k) Account Balances at Social Average Percentage Reductions in Security Normal Retirement Age 401(k) Account Balances at Social Security Normal Retirement Age Percentage of Compensation

Volume T-171

Pages 56

EBRI Testimony

March 28, 2012

Jack VanDerhei

Financial Wellbeing Retirement