2 52 35 be be 4.2 A 4.4 A 1 this (Mar VanDerhei, contrib VanDerhei, 13                                          12                                          The                                          6.2.2.2                                            similar year   conv $4.3   Summary Simulat proposal model End Intr Appe ch  after u   2012 erted trillio a. tion oduc  noccurrenc Au otes ndix    was Ja Ja  iStop this on  to )  plans n  will  tio ck. ck,   to filters    matic  for  C: results next   n   at study   and “Modifying  cun   Im  o   all at retirement en   doubtedly p   out used  t   Enrollment tak Craig this Baby ribut a                                                                                                   was ct for  eanyone   sof  time in      in voluntary place Boomers Copeland. released,   the  gMarch th   altogether   age. needs be e   financi    over Fe when  subje                                                                             dof Of   e  and Co   en  to age  ral “The 2006   course, the acngress  rbe l  to   Tax Ge   and 35 llme to    perc Im  thoroughl n to                                                                                                                                                                        awhose Trea   Xers. p  housi    cost n evalua passed this aent tct  401(k) tme   -of aco ben    nge tenure   gWhile te yDeferring ncep n   th   ma tof exami e    the of plans efit   those rk Pension  t   401   analysis  wi trillion impa does entt     ed. h crisis (k)   Ret s  th  ct tnot    ating Pl   eir                              EBRI Protection dollar i of remen beyond a  in  nadjust   curr defin                                                                                                                                                Contribu 2  they s008  tdefici udies ent t e   Age  the  for d wo and    em Ac ben ts    uld the ti assumption on  t   2009 ons: are phave  of el   oye Re re fit fact   2006, useful  duce  tireme Projected  freezes  on docu r  that is                                                                                                                           re  le  or th   me in tirem ss  that n the  e  PPA,  ton  focusing  than  amount nted In   Im   vast come e  which pact nt  T-172  their tha        t                    ? ? With Defined  the  bene  exception fit annuities  of the  and/or earliest  lump  age cohort -sum distribu  (those tions.  currently    26–35), the average reduction for  45 current participa they retirement majority atten defined eased  readi are t ion some   age on Adequa con   n saving of  on ts n    Participant ess savings  4t   minus by of  r01(k) this ibution      the  simulating or cy.”  proble    stop  parti  30. ad c  EBR o  plans m nt  Account  csaving i ribu m, n Iip  Issu i st ants the   they (and rtaieon  t  altoge  mi  iBrief,  are v  Balances,”  sdo the e  nim will  barriers   year li  IR  ttle th no. u  remain m Aer s,     rollovers to 358 employer   is if   to  Marc   help not displaye    (Employee constan providi  deca  hpolicy   they 2012, -co ddes, n tntribu gby    on makers produce)  AE   EBR  Benefit  the away  the  and tion I  amo  Notes part    und from for    Research rate are unt   of the erstand     retir th    tha partici they  e first  compo te    would ment Institu currently   ex time pants acn  tl  age.  outlined tent be y  e, and/or  where    neede Jun  have   of Mo  e retir   reo   2011).  plan safe in th d  eese to savings vment er,   harbor  sponsors. financially    defi  even    cits and   if       T aany x R  income e  f quartile o rm and T  decreases for older ax-F  age cohorts. a  v This or  is du ed R e to the fa etir ct that those ement  closest to  55 30 16 40 56    IRS This Figure It is  tax  important survey   1ta  of bles     wa Ho  from  ld sto  ce  ono nnd   20 an teuc 09 d  th  ted Va  aare tn   th online D  used erhei e an  nuitize to within  (20  com 05d  )th p.   accumulations uet eUnite  the tax d Stat  owe es  ind  by   this on  Harris  the  analysis  amounts  Inte  are ractive   from received  comm  401(  frissioned k)om  co  ntrib define  by utions  dthe  ben  Principal  efit plan  s  Even  Mr. The To summari   Chairman previous  with b.  thze, Redu e  results s  and  eit  empiri  appears ce  me  th  assumed mb ecal  aers m  fr  techni ount om  of  non   both the  you ques, e  co  of   empirical contribute mmittee,   the it is  401(k) difficult  analysis   tha  parti  n tok  obtain you c  and ipants  for  simulation  a   your  were true  invita  value automatically  results tion  of th  to  based e  testify 401(k)  enrolled  on   toda system’s  tens  in y of   on these  millions   tax potential        ? Net housing equity.   3 investments, one indemnify are security provisions   comin does retirement  tha  look g   th for  fr t e  appears om. not    automati employees at  age   the including   For  will  average  to  exa c   be have  escalation.  for m   the gple, e    the balan fe nerat  value w  F  re eigure rciduction   ng es years A  of l  th the for   3 their ough  provides   of  workers most  in   futur pri   it th  m  was eir non ary e   inf   near ex contri  -too  Social oresidence prmation e   retirement ct soon butions e d Se  retir  curi to  on  or  kn  ty subject e   the the ment ow   age, retir   average  value how   it eincome  to ment  is   plan   potentia of obviously   idefin  ndividual wealth   sponsors under eld  reduction    not  for ben various  retirement    would Baby correct efit  plans. rate   Boomers as     to a - of     -   TM Finan and exclusively  Social cial Group  Sec  andu rity do  from   not (with  May  include  th  17– e percen   projec June tage 17 ted , 2   So of 011  cial Social . It  Sec  su  Secu rveyed urityrity  re  79 tirement  benefits 8 emplo   subject be yee nefi -benefit ts.  to  This  Federal  decision  is in  co Inco ntrast  makers me Tax  to  for  other proxied  companies  EB RI as  a    On to The     of reform retirement     The                                                  individual gen        a           analysis average  caut ..  e“Tax “  and rate  The ion    pl  Reform tax   s -  EBRI in aparticipa perce ians; gnifican ry  -VanDerhei favored  note,   Re instead, n  Op tage tirem t  n                     it  tion 401 t   retirement is  or b e  e  (February admittedly they s: (k) nt ductions servations  Promoting   Readi accumulations  presum    na   for 201 es c  counts. ve (dating s  ed  the  Retire r1) Ratin y  that   difficult  was “long    over gall Imen   :am workers’   desig the -  te  Ret an  Ja tto  nure”  Se way ck   n entire ideter rement ced  VanDer u  rate  rit  back  to cohort ym   working .answer  ”of ine    In to EB hei  contribution come   how  RI 1currently ,996   research Issue   two career   those  Preparation in   qu ma Brief,   ages est  when  wor  ny dir afi  ton no. e cases), e  k36 ctor re  s:  many  the  rand -364   s45   not of   first   tha  Futur are  (Employee  the of  currently t    to  year ag  the Employee eday’s a  i n  were  largest          65 17 4 17 VanDerhei (September 2007).  c. Continue to contribute what you do now  return and income A There respond look  household   Gen only   isassumptions.   defi   ato Xers   at significant  th  cithe  is .ts is    However,  considered   new by 401(k)  the   clarity, incre   Later  nu bal  mber the aa  to se nce   EBRI that   potential  run in   of with   the  ye published   short fuar  t  self th u, re  an e  in of -   employe reported years c  update  rmoney e  aa se study  e lof idg  ein    propensity ib version ’s at  in  this le  -current  risk 2007  for  model   percenta  coverag of   employer. that  the to  if  re   mo  ashowed eggregate duce ges  in del   aresulti     savings defi was    how  ned resources  develop ng  au    for from contri tomatic  those e  (1  dbution in  )to   retirement em   in  escal enha  pl the  retirement oye a  nce lowest tion r   the are  –       result of the 20/20 caps.   9 research with function  three   of (e.g.   to th ,e 1,  Van  various 000D  em erhei  ployee retirement  and sLucas  that  income  ,do  201  offer 0)  comp th  defined at includes onents)  contri   as both  bution well  compo  as  retirement  then en indts. ividual  However,  pla naccount s. Thes  in the e  withdrawals.  decision  previous  makers  analysis,     were  all    Accounts covered retirees traditional Benefit driven for those  pri   Resea  Prospects.” Benefit only and/or  m in  (V  ar the  had E) ily rch     type Resea  lowest participa by   access In st age  EBR  of itu r-ch   40 in and t  Ito  e. t  Issu com In i1(k) n   EBRI a  st gincome   401(k) eitu in e   plan Brief,   quar  atise  defin  ,a   u November    plan characteristics  tnon nder no. ilee  (24.9  d p  344 for   artisan con th e  a(Employee   t  perc current  portion ribu 20  research 11). en tion  rather t)    set .of   plan The  Benefit  the  of   than ins    redu would tax irt itute care   tenu  incentives Research cti  er.  react othat re ns   with In for  has  to  an  Institu   has this   this  the  been atte   the ag   set current tmpt e,  efocusing   potential   cohort of July  to  in   ce employer, contr 2010).  ntiv  de on  to crease o  es, retirement   lgenerate  for  and  as  these  to   they        ® 1 A brief description of the EBRI Retirement Security Projection Model® (RSPM) is provided in Appendix A followed  1. What percentage of U.S. households became “at risk” of insufficient retirement income as a  not sufficient to meet aggregate minimum retirement expenditures, which are defined as a combination  modifications plan EBRI savings the systematic   for int e  categ Gen ract ive Xers. o  to ries.  increase   Ballpark ex  Th  iFor House Committee on Ways and Means stese in  exa g  of    Ret Epl  $m deferral ans, timate iple, rement  and  of   rates  worksheet the (2)  Sa   avings fu   over substantial ll-ti me Shortfal  tim  by  workers  eproviding  -  would portion ls (RSS)  who   render Mo   of are  are  nt low  present e  currently   even -Carlo income  more   values simulations   hous saving  favorable  at eholds   retirement for  of  retirement    the de results cre  necessary  aage sing  for   wh and   AE oro       selected workers 18 VanD  erhei were from   and simulated a Principal  Copeland  and  Fina   (2 job ncial 00  cha 8)  Group . nge was  client  allowed.  list, an  d their data10  were not weighted.  d. Increase the amount you contribute  and migh 7.2 EBRI problems, a4.1  sum   per  Average heal  tw   be that, ill cent  th  continue in  EB   ben   combined au for  account RIt e  othose and fit -enrollme  sto   for ICI    work in ba  with   pro th  th lan ee  nduce with   cpast Social  tsecond  eplan s  d  ac 34   saSe t    with ujoint -yea income cur al participant r i s. an ty publi  EBRI   be automatic  quar nefits. cation  does tile  da     would in and not ta  escalation  2002  to   tak then  be  repla etter   poli  with increase c   of assess ecy    asimulation worker  positions  sizeable  to  some  10.0  contr   portion   an of results  perce  dithe butions.  does  behavioral n   of showing t  for  not th   e  those   lemployee’s obby.  that    in     und the er     57  ?Rot Within h IRA an  each d 40 1(k of the ) acco  four unt sage  are  cohorts,  not used  the in this  hig  version hest-in  of com  thee  model quart ile bu texperie  will be n incor ces pthe orated  largest  into  average a    by a chronology of its development and utilization in Appendix B.  See Appendix C for additional detail on the                          ..  “Testim  result “The ony. Impact  of the  U.S.   financial of  Congress.  PPA on  mark  Ret  Sena iet rement  and te Finan  real  Inco  cestate eme  Committee.  for  market  401(k)   crisis  Tax Participa  Re  inform  2008 nts.”  Options:  an  EBRI d 2009?  Issu  Promoting e    Brief, no.  Ret  318, irem  ent  represent of report eliminati replacement designs  deter  that  u m nnde  inis gth   th future e  rrates tic  ey  additional a   nu expenses curr   cmber nee oently ntdribu ed   of a  m fr have   for tdifferent om iount on  specific s   less the  to  each   savings  Consumer tha  scenarios.   pr individual n obabilities $1,000,  plans  Expend      in as 71.3    that of a reaction   iture percent retirement  grou  Survey p   to indicate wo  the -uld  inc (as   o eneed  am x  they  clfunction eusion  ad  to  wo equa  have  of uld  of  cy employee    re accu income  uduce nder mula   th alternative ),   ean ted  amount d  at some  age -    tende preretire forthcoming 19 a 41 31 third  conti    Figures A These  similar -in nci nuous c o estimates em  m 6s ent    ques  and of e EBR   coverage quartile.  this  income  I7t ion publication.  of   compare group.   Van was   Th  for D situation,  aske erhei  eUntil   those  quite reductions  d  with and  this   favorably fortunate   401(k) Copela  the  type   30 increase   nd  of participa pe  to     enough (2 rcent information  those 00 8) to  governm    n in  17.1  ts  to Hol  could  have  d  percent is en e nt available,    an expect c  match odnt  Va  for innD uou   to  provision those  erhei it  replace  swill  coverage   (2 in  be 00   suggested th   quite 2) someplace e  wh highest  durin e  difficult n  th inge  - Gale, inco their   between differe  to me  Gruber,   working fu nce  quartile lly   51  assess  an  and  d.       percentage reduction from the 20/20 caps. This reaches a maximum value of 15.1 percent for the  impact of the 2008-2009 crises in66  the financial and real estate markets on retirement income adequacy.  Follow-up questions were asked of those who indicated they would either increase or decrease the  The The In an   testi exclusion  atte Security (Employee mmp onty   of to  draws    (T auto provide - 17 Be  -on 0) nenro e. f   th 15 m it llment eResear e  Sep  aningful extensive t.  2plans ch 011.   In statistics  st re   initu se  th ar teisc,   hon analysis June  cond  the  2008  ucte 401(k)  was ).d     by necess  system,  EBRIary  on  EBR  given  these I en  the  ttop ere  current idcs  into  over  a  mode   the collaborative  last ling  13    years     health insurance and out-of-pocket health-related ex penses, plus stochastic expenses from nursing  saved. contrib risk 65 Jack  to - manage VanDerhei,  eli  This um tions ina  vment at lue efor  their  Researc   retirement de  trea clines  expe tmhents.   cted Director, to  savings 38.8  defi    perc cits  Employee  pla  en inn  tretir s  for from  e  those Benefit ment  taxable  with  (whi  Researc  income,  savings ch could h Institute   needs of  be  $1,000  a  rela to     be  tively to  ana less  short ly  than zed   carefully perio $10,000. d or  when   could    18 Orszag between 2.  (2 Of 00 nomi  6) those . nal and  who  real  are  re  at placement  risk, wh at rates  additiona  are consider l savings ed. However,  do they  ne thised  is  to to  be make  expected  each  year given  until  the    the careers. 69 In 2008,  perce  “benefi     EB nMo t RI t(d ”r  eependi included portion over, the ng  of   all on   AE the  the  in  type  cost c onew m  -of eben    PPA quarti 401(k) efit  provisions  lanalysis e)  plan  of their  wh   suggested in e npr  a  combin  estudy -retirement  above. ed  that  with  co   income EBR  automa mpared I did  assuming  provide tic  pote  escalation ntial  an  th  accumulations  ean  pu  aprlyrc ovisions shisa  of se  some of   a    highest-income quartile for those currently ages 36–45 and falls to 8.6 percent for the highest- amount 20 58  Van Capital Derhei  they  gains   (A currently  treatment pril 2010 )con   is no tribute: t used  in this version of the model.   with effort assumption The average  it  with s Ret  the i rement of -perce   no Investmen  job n  Se tage  chan curi  rtety ge.  ductions Company  Projection  It would  for  Institute   be  Mo the  very del®  “long  (ICI)  difficult  as -te   we innure”  1ll 996  as to   annual provide cohort  known   analysis currently   an as  the accurate  EBRI/ICI  of  ages  te ns analysis  46   of Participant -  millions 55  of are  the  larg  of - Di av  eindividual resetrc  afor tged e      considering last home 2 This  deca   num and des). ber  ho  the  me is  The  somewhat  overall  healt  values h  imp  ca   for sre maaller   cexpenses those t of  than  such  assumed  the   proposals. (at  $4.  least 6  to  trillion   have unti   l  re  the no po  futur  rpoint ted in e  th  yea VanD eyr sare erh  of  epi eligibili i (Oc cketdober  up ty is   by 20  approxi  10 Medi ); how ca mid). eately ver,  This  the      assumptions listed above (especially the lack of job turnover and therefore the suppression of cashouts prior to                          ..  “The “ERISA retirement  Im  At po  30: rtan  age  Th ce  eto  of  D   make eDefined cline  up of   Benefit Private for their  -Plans Se  los ctor ses  for  Defined  from  Retire  the  men Benefit  crisis? t In c  Promises ome Adequacy.”  and Annu  EBRI ity Payments:  Notes, no.  8  appears (nominal Contents under of 6 RSPM  theThe    was likely AE  to  ) Pot   and an   significantly ha  “costs” e nuity  ve nt VE  ia the  for  lat  Impac  in    pote several age  terms  enhan  65 ntial t of .     different of  Th Tax  cto ed  reduced e  produ    Ref lower for o  th age rm c  e  savings eretir   May   on even groups.  eRetir  ment 2008   larger rate  e Aga   ment benefits EB  for  retirement iRI n,  low    we poli Se-c   income found for cy ur iforum  tthose y  accumulations   certain  parti   by currently  callowing  ipan (hights  -  in  for income is   automa the  somewhat  many  401(k) ) groups  tic of  th  esystem  nose rollment  were    at    a   income quartile for those currently ages 56?65.  5 42 A similar question was asked for a 30 percent government match. However, follow-up information for those  Hearing on: Retire perce baseline those 401(k)  nmen in  p tage   assumptions ath rtteic   rPlan  lowe ip eductions an  Data tst s -da  income use  ting Colle  in d in  401(k)  back  ction the  quartile  20  in   balance Project. 10  some   analysis (21 .cases 1    un  As per  did der  ofc  as  en  not  December auto  far t).  all   Th as -oenrollment w  e1996.   for reductions  31,  the    2010,  utili  if zation the  for  the   plans this   of data   net age  inbase  cl housing  cohort uded  incl   an  equit ude decr  au dyease tomatic  par  to ensure t icipa to 9.9   n  t  - version $78,000  of  per  th iendividual.  model is  constructed The number  to de  simulate creases substantially  "basic" retirement  for those  income  with  1 ade -9 quacy; years of  howe  futurver, e eligi   bility  19 67 21 The lowest rates are experienced by employees who do not “remember” their previous contribution rates when  59 retirement).  VanDerhei    and Copeland (2001).  2. By about how much do you think you would reduce your contribution?  Would you:  What (Employee  Will It  Be  Mean?” nefit Resear  EBRI ch Issu  Inest  Brief, itute ,no.  August  269  (Employee 2011): 7–16.  Benefit    Research Institute, May 2004).   likely Septe of covered ameliorated 1   401(k) Introductio  m tober   do by  participants    b su 2  by e 011 ctter hn  th    ............................................................................................................................... aSenate  eir  un plan  der So  with  du cial   Fi VE ring nance   the Security  than   apotential  significant   Commi AE ,pay  butm  ttee  for overall, ents  portion  automatic  hearin  repres  th  of e g.  AE enting   However, their  escalation  results  working   a rela   do no  of minate tively  information  careers.  contri  large d.bution       measu No  on te s  to plan  that r  be e of    sponsor included. the ...........................  what  AE  desi th  reaction eir   gn  pre    2 -   indicating an increase or decrease in contributions is not available.  retirement As one would  income  expect,  adequa  thecy  an .  When swer  to the  the  201  first 2 analysis  question  is re pdep eate ends d with  to th  a elarge  same  ex  assumptions tent on the  as  size  use  of d in  the  20 10, the  Prior level escalation perce   informa to nt  estim for  pr  those ovision. taion ting   about: in  the  the  For  potential    second  examp-lin  ere ,c  ifdu o me acti  participant’s  oquartile ns in accumulations  and  contribution  then incr  resul ease  rat  tto ing e had  11.6  from  alrea  per  reduced cdent y bee  for  401(k) n  those  escal a  in ted  the  to  third  8  - alternative to The $55,000  findin   and ve g rthat s ieve on sthe n  of  fur   high the ther  model e sto t-income  $39,000  allow  quart   similar for those ile  analysis within  with  each   10 for -19  replacement  age  years  coh  of ort  future  experiences  rates,  eli  sgtibility. andard  the Gen  -largest of-liv  Xeing rs  average      they change jobs, have a stochastic opt-out of the automatic escalation, and participate in plans that limit the  Today’s testimony will deal with the following questions:  to retirement itself Additional  the  in  propos  no  mod  way income al  iprecl was fica  had t iavailable udes on sbe  were  these en.   at  When added   high that-    So ti income inme.  cial 2009  Conseq  Se   groups for curi  aty  uPension  e  benefits from ntly,  the increasi  Rese  (using  2011 arch ng   EBRI th   their Council e current  analysis  initia  presentation  lstatutory   defaul presente t rates.  for d  that  ther m  ulae)   inevol  was  are ve  d    60 32  Van Presumably, Derhei and  th eCopeland  $20,000  (Ju figure ly 20  would 02).   be indexed for inflation in the future similar to current treatment of IRC  Tax Reform and Tax-Favored Retirement Accounts  a. Reduce it by about a quarter  9     accoun     aggre                  References g..ate   “Can “Cappi t bala  defi  Ance cit ng m erica increases      Ta the x-  Prefer hous Afford  toehold  red $4.  Tomorrow's 8  Ret   trillion. had irement  in  defined  Retir  Contri ees:  con buti  Re trisults bution ons:  Preliminary From  plans  th eand/or  EBR  Eviden I- ER IRAs Fce  Ret  as  of  iwell rement  the  as Impa  their  Sect curity   of relative  the       2  What is the size of Americans’ retirement savings gap? ...................................................................................... 2  5 contrib perce income ?Changing ntu   quartile. 23.4 of tions  co  million mpensation ,  atax  set  Th  incentives  e  of 4  r01(k)  ebaseline du  at ctions  plan  one    results    pa increase employer, rticipants,  fir sto t  nee and  14.1  ind     upon s  per  to cbe  ent job  run   chan for  to  those  ge dete  was  rmine in  the automatically   the highe  likely st-income  values  enrolled   quartile. if the  int ovarious        fortunate automatic 43 calcula  perce The baseline tions, n tage  co enn  ough and t ributions  rreeduction sults  other  to  have in  to   ad Fi   is 6gure     at hoc percen no  least s  surprise,   thresholds. 11t  and    of twen  co  12  mpensation gi ty  were ven  years     the simulated  of  in   and future crease  increase  assuming  eligibil d likeli  the  ity h thood  contrib e  find  m itha dpoint  the uttions  worker ir  val av  by ueer a1  sfo g   perce in er   each defi thisn cits t  coho categ  per  reduced  year rt ory  either   in (the  th  to  e     38 Sec. added abased  winners/losers  41  5( on in,c)   sev the  limits. e  media ral    analysis alternative n com  of ®bined  define  scenarios.  repla d ben cement e  Mo fit freezes reo  rates ver,   in  an th  the ed  information the  first  enhan  yearc  of ed  used  retirement  employer  to mod   contri eare l pote  103 bution ntial  per c4 sent  01(k) to  for    the  exposure ? What Projection National  to fl uctuations is the  Commission  Model.  size  of in  ”Am the  EB  on e  RI housin ri  cFi Ian ssue scal ’sg    retirement Brief, Responsi  market.  no.   bility  The 263  savings   resulting  (Employ and  Ref gape o?  per erm    Benefit  cRe entages comm  Re search endations.” of households  Institu  EBR te,  tha  INovem  No t would tes, bno e rnot .  7   another “all AllianceBer By tax 61  Van 2009, -  reform pessimistic” Derhei  401(k)  man nstein  options  and y   assumption of plan,   survey. Copeland  the  are  would  401(k)   not Sensitivity   (December sc  imposed  th e  sponsors ney ar i o“remember” analysis ). In   on 20   co wh  02 the ntrast,  of ).o    previously th current is  their  the  assumption  hig   40 current h 1(k) est had  rates     is system. VE  rate  shown  plans  are  of  experience     The deferral in  had  Fig  model  ushif res  and ted  3 d   and by us   to start  ed emplo   4AE   in of   deferring   plan this Vyaees nD  sarticle  erhei  wh and o  in    do EB (Mar   isth  RI  based e ch  was      currently ? 64,455  exceed  empl  the oyer  $20,000 -sponsor  (ind edexe  401(k) d) li mit pla nwhen s, holding  their    contributions are combined with employer  only $23,000.  3 Unfortunatel b. Cut y one  it in  of  half,  the  most or   significant components of Retirement Savings Shortfalls comes 68  from an11  exposure  3  Impact of tax favored retirement accounts on retirement income adequacy ..................................................... 3  5.1 Analysis Resul  of th tse  Fr  oldest om th  coho e 2011 rt (those  Retire  curr ment ently  Con  56 fid-e65) nce  show  Survey  a mark   ed decrease in the average percentage  AllianceBerns The version oftein.  the  “Inside model used  the Minds  for th eof  an  Plan alys  Sponsors” is in this testi   Researc mony has . 2sumes 011.   all workers retire at age 65 and  lowest defined participa  inco  con nt me reac tribution  quartile tion to  plans   the and   proposal provided  between   was at  83  the   and limi  ti tme 86 ed   per th toe  “an c  defin ent  analysis  for ed  the  ben   of hieghest fi  two t plan  new  three  was  ques   in frozen. ctoions me  quarti from    the les. 21st     ® have 20 33 “remember” 12  Emplo )  bee for(Employee 2003).  ythe nees  “at   th minimum  age eir   risk”   previous 50   Be without or  reduction nover ef icto   ma Resear ntributio  the y  in be  2008/9  a  ch allowed cco n  In rates unst  tcrisis itu  balan  to  when t e co  ,that  cntribu July es,  the   ended and  2011): yte  ch   in up ang  Figures   up to  2–6. e  jo  an “at bs,   a   5dditio risk”  do  an d not   nvary 6al   of op  $5  the  tfrom -,5 ou  00 same t of   per a  the low  publication  year.  automati  of  3.8 perc  cfor  escalation,  the ent  to a   ? What is the impact of tax favored retirement accounts on retirement income adequacy?  able on new  the   to pla   401(k) track n at that  the  voluntary   rate, chan ges or  e  wou  ninrollment  plan ld their  provisions  modules  contribu 10:00 a.m. April 17, 201  for  from tion  hundr   rate theeds   EBRI drop  of  Ret   th toe  the irement  2lar  gdefault est  401(k) Security  rate  pl  of  Projection ans.  the  This  new  information   pla Model n?       that contrib   ?fa ce$ s  mos u1.414 tions t retirees,   trillio or aren   however, pre in assets. dicted  very   to  do few  so  of  in  them  the  choose future.  to  However,  actively tr  for eat  each this risk.  age  Van  cohDoerhei rt other  (Oct ob tha ern  20  th10 e )  62 VanDerhei and Copeland (2003)  reductions for the “long-tenure” cohort in the lowest-income quartile (12.7 percent), although it should  Figure immed wave of  4iately   the provi  Ret  begin des irement  similar  to withdraw  Confidence  information  money  Surv  for  from  ey the  (RCS)   impact their  reflecting  in dividual of futur  how  ea cco eli gworkers unts ibility  (defi  for  indi  ndefined ed cate  contri std  th cobution ey nt ribu would  tand ion  likely   plans cash     for  20 c. Reduce it by about three-quarters  4 and    partici The value pate  of in  pla taxn -favored s that allow  retirement  the auto  acco matic unt  co s und ntrib eru tcurrent ions to  tax increase  provisions:  to 15  perce the case nt of  of  comp  401(k) ens  plans ation  ...............  and   3  maximum reduction in account balances. The average percentage reductions in account value in Figure 3 vary from  VanDerhei Choi, high of  James  14.3  (M   J. perc ,arch,  David ent  2011 .  Laibson, ) provides  Brigitte  an  C. analysis  Madrian,  of two  and  new  Andrew  quest  Met ionrsick,  from  “Saving  the 21  For  wave  Reti reme of thent    on the  provides Undou was 4.3 (RSP A new   ?The used M)  b subroutine     and te What a“con   in first dly   an is  tr -  many  is order similar  oversy” April  the  was    401(k) approximation  value 2010  in   added over  man   of EBRI  participa   a yta   to urespects  xtom Issue -  the favored  ofan   the mod  tic Brief ts   to  in enr im  retirement e   th the pl o ato  to is cllment t    one autom allow    of show,  th  us e   de  simulations  accounts st ed  aon tic ocrea  cin c he en a   Holde ag ssin rol tica  g  un linatu m n  reti  ,n of  en der the  and  re various rem t   situation  curr of signifi  VanDerhei  ethnt ent e   styles nursing cant savings  tax  fo   im llow incentives   of (2002  home p   target a  the ct )  of and   in la  -mo dat   tter that (wi  home ving et  approach h   funds it      looks health to AE  for    .   oldest one, the lowest-income quartile has the second-highest average percentage reductions.  34 Sec. 415(c) of the Internal Revenue Code.                          ..  “Retir “Kansas ement  Futur  Inec  Retir ome eAde men quacy: t Inco me Alternative  Assessment  Thresh  Project.” olds and  A  projec the Imtp  of ortan  thec  EBRI e of  Futur Educati e Eli ong ibili andty    in  be noted that the average reduction will be most muted by previous account balances for 401(k)  balance react Gen Xers  if they  plans  alth  were ough , as well  no  this  longer  as  ti m IRAs) e  allowed th ewhe  analysis never  23to defer   also the   sum controls retir eofm  their e  for nt savings   relative basic 24  expenses  plan  leve  clso  of nt  and  ribut pre  uni -re ion tnsured irse  from men  medi  tta  income. xable cal     For  Since the inception of the project, average balances have bee 69n displayed as a function of bo 27th the  63 VanDerhei (January 2004).  3.1 increase  to 19 .7 th eper  cocntributio ent (dependi ns byng  2  on per  income cent pe rquartile)  year (the  for  “all  401 -optimistic” (36 k) participants  assumption  currently  scenario).  26–35  under the minimum  care Retire  expe men nses t Co  onn fthe ide n RSce S  values Surve yby  (RCS  age)  coh  showing ort, gend  how er and  wor  marital kers  stat  would us.  Ad  likely ding  rea the cnursing t if they  home  wer eand  no  home  longer     As only aplans:  comparison  additional  at    for current Path  th ose  of  information with   401 Least  currently  parti (k)  Resistance,”  participants cipant   bec ages o -dire  m 25–29, e s  originally cte and  available d   the does  inve  differ  st  pre not  on mp e   workers’ eattempt ared nt nce s in   in for  2009.  the    to Ta behavioral   in xme  Policy c lIn dia ude  April n   eligi and accumulati  responses  2010,  bthe le non   Economy the  ons p to  model articipa  auto  at  norma 2 - 001, enrol was nts  completely update ll  ment, or retirement  workers  dEBR    I    particular emphasis on 401(k) plans)?  12 Although 3. By about  this may  how  be  mu  duch e  do to  several you 1100 L  thi nconsiderations, k ong  you wo  would rth Hou  increa se  it Office Buildin  isse  almost  your con  always gtr ibution?  a result   Wo  ofuld  their  you  current  increa  or se  it  4.1  39Average account balan56ces ............................................................................................................................ 4  The answer Research Defined  to the  Contribution  F  second und and  question  th e Re  Mil tireb  also men ank  Me dep t Plans.” mo ends rial   EBRI on  Fun  the  Notes, d,  size July   of 16, no.  account   2002. 4 (Emplo    balan yeec  Benefit es and  ex Researc posureh  to Inst  the itu te,  participa In an article nts  that in th is appeared  age group.  in the  Mo  Wall reov eSrt,reet  the  Journal lowest- income last July,  quartile  Anne Tergesen  no longer  suggested has the large  that st  redu the  ction,  participa those expenses income  in.”  th n  exceed t’s   e  lowe  age  and st the  inco   after tenu mere -tax  quartile,  with  annual  the  the  current  inc  average om  employer e from  defic  Soc it  to de ial  clines allow  Security   afr  more om  and  approxi   meani definem ngful dately  benefit  assessment  $106,  plans 000  (if  for  of  any).   those the  If   35 reduc  Thetion  reason  scenario.  that   the Fig u yoreun  5ges  shows t age 28   that  cohort  the ydoes  vary  not  from  follow  8.8 to  this  24  trend .4 per cisen  due t (de  top  th eneir din  relatively g on income  lower  quar  curr tile) ent  for     health allowed  care  to  ex depense duct  retirement increases the  savings  averag epl  individual an contri  RS buSti  fo ons r married  from tax  hoauble seholds  income.  by $2  5,317. Single males  draft: July 19, 2004;   will age re who -parameter   up would  are date  currently  be  this  iapproxima zed  analysis   with not e  401 ligi  to te bl(k) provide yle.  2.39  plan However  ti - desig ames  more  final n,  parameters  unlike robust  salary  the  model.  in  2   an for 002  AE   sponsors  model,  plan relative  this  that  analysis  have  to a  adopted VE  assumes  plan.  a uno tom  joabtic  tur - nover,  22 64  expe Results Van ? Dcte erhei How  are d future    (mig limited 200h 5)t  .con   wor  to tem ribu keployees rst  react ions  exceeding currentl  to changi y ag  ng 20 es   ta 25 perc x–29  incentives? en  an t dof  assumed  compe   n  to sation  have  when 31–40  co years mbine  of eligibility d with employer     by about  57 equity marke April t2011): ; however,  10-19.  it  is a more complicated question involving both the proximity of the  as automatic  the reduction  enrollment  for the  (A  second E) provision -income  for  quartile 401(k) plan  is slightly s, a plan  larger  desig  atn  13.3 expa  percent. nded and  The  clarifie  reductions d in the  for    this  with wages 40 there 1(k  no )   is than participants   sufficient years  older  of  co future   currentl money horts  eli af ytto g  e26– ibility r  pay adjusti 35  expens  un tonder  appr g fo  th res  oximately ehistoric   without maximum  age   ta $66,000 /  rwa epping duc ge tion profiles.  into  for  sc  ethose th n ar e itax o  with -qualifie  twendt yind  ori vidual more  years. accoun    ts, A    accu 4.2mulation   Simulation  potential  res70 u  of lts  these for 29 volunt  plans. ary    enrollment  VanDerhei,  40  H 1(ok)lden,  plans  Alonso  ...........................................................................  and Bass (2011) computed an overall  5       experience        . “Massachusetts  an average increase  Future  of  Ret  $3i2,rement 433 while  Inc  single ome  As females sessment  have  Project.” an increase  A  proje of $46c,t425  of .the  A precise  EBRI  Education evaluation   enrollment withdrawals, 6.2.1 New  provisions.   Survey or loan  Analysis defau  A completely lts.      updated version of the national model was produced for the May  contributions. Phrased another way, the 20/20 cap would, as expected, most affect the highest- ? What is the potential impact of two recent tax reform proposals on retirement security?  household Although ana  to lretirement ysis based on  ag efinancial  (the closer  economics  to retirement  suggests  age,  that  the  hi  fewer gher-income  years of  e m additional ployees  wo savings uld be    the  ____ 23 65 age 7 Pension Later   These Van  cohort Future _“For  that Derhei   Protection findings  year,   Better dec  wo  (Mrrk  arch ease  VanDerhe are   or    Act part 20   For to06   of 11.4   of Wor ).  i 2006  the and  per s e21st  :  Lu (  cDefault PPA) en c  an ats  nual for  ,(2  an 010)   Effects those  dRetiremen  de  focused signe  in  and  the dt   4Co to   third on 01(k)  nfidence broaden  how - in Savings c  to o m Su   improve pa ervey   rticipa quartile Behavior,”  (RC Splan tion ) ,and  a  survey  in  desig Pe   then these nsion n  that   decrease and  programs,  Resear  gau  wogrker esch  to  th   Council   ewas 8.7  views       average similar the excess  reduction  a. acco  isA  assumed unt  quarter   balance is found   to  be at  for   year investe  the- end high d  in 20 er  a  10 income  non  of -$60,329; tax  qua -adrvantaged tiles.  however,    ac count the average  where  vthe alue  investment  for participants  income  in  is  of the impact would involve a comparison of the values supplemen71 ted with the premia required to fully insure the            . “Aand  Post  Researc -Crisis hAssessment  Fund and the  of  Re Mitilbank remen  Memorial t Income  Fu  Adnd equa , December cy for Baby  1,  2002. Boomers    and Gen Xers.” EBRI  36 Although additional analysis needs to be performed before assessing relative importance of these factors, it  44 Under the baseline assumptions, the average percentage reduction in employee contributions for this group in  2010 4.3  EBRIThe  policy  “controver  forum sy” and  over  used  au  in to matic the July  enrollm  2010ent  Issu  decreasing e Brief.   retirement savings ........................................... 5  income workers, but it also would cause a significant reduction in retirement accumulations for the  58 most and available),  attitudes  likely  th to  of e  be   relative working  negativel  -lev agee  yand l  affected of  prere retiredt  by  irAmericans ement  a proposal  income,  regarding  to cut  and   retirement, or  the  eli de minate sired  th e  the probabi ir preparations  deductibil lity of ity  ade for  of  retir qua  40te e1(k) ment,      their  Working Paper, PRC WP 2002-2 (Philadelphia, PA: Pension Research Council, The Wharton  perce 6.2.1.1 Using actually educa ntthe ion t   for suppressing  Plan 40  to  those 1(k)  optimize  Sponsors  voluntary  in  the retirement  the   hi ghest results  enrollme -  savings. income  unnder t modules    AE quartil What  plans  that e.  fro    with  m article  RSPM  automa  failed , VanDerhei tic  to  escalation  mention  (Nov   of is em   th contrib ber at automa  2011 utions ) sh tic .o  While  w enrollment s that  it  th is e   2 taxed financial What  as  co  ordi nsequen  is nary the si  inco cze e of  of me.  nursing  Americ   The  home ans’  individual    re and tirem  home  accoun ent  he  savings alth ts  care are  gap?   tracked expenses.    un  For til  an the  exampl  point eat  of  which  this comparison  they are  with a  their 60’s was much more representative of what a participant  would have available for retirement.   66 VanDerhei Issue  (Se  Brief, ptember  no. 354  2006  (Employee )   Benefit Research Institute, February 2011).  response appears that  to the  this  proposal  result is  is ca  14 used .3 perce  by atn  leas t. Act cou twon  fa t balanc ctors. First, es will  the  also  definit  be reion duce  ofd  income  due to  quartile the plan -inspons  RSPM or  isreaction.     In addition to the expansion of the model used for the two analyses above to include 401(k) plans with  3 Impact b.  Half of tax    favored retirement accounts on retirement income adequacy  lowest-income workers.            .  “Oregon Future Retirement Income Assessment Project.” A project of the EBRI Education and  confidence with regard to various aspects of retirement, and related issues. The survey was conducted in January  contrib retirement uSchool, tions  income.  (at  University  least    to the  of  Penns point they ylvania,  are  November constraine d 9,  with  2001  respect ).   to the annual funds available to  different median The is difficult  actually 4.4  new      real to out  mod Summary   increasing de p -re ut termine e plmetric,  was lac em ...............................................................................................................................  used  savings   aesee  npre t    Van to rat c ise  analyze efor Dserhei   “target” at  many  age  ( 200 how   67 more—especially  5) for   from eligi .  retirement b  401(k) ility for  balances   participa savings,  the lowe   exclusively twe iostn  -tried  income in a  defin to  for  de  40  em participants 1(k) do  con nstrate  parti tribu c ipa these tion  currently ....................... nts.  plan  designs’       ages    7   depleted; if the Social Security and defined benefit payments are not sufficient to pay basic expenses,  Even looking at participants in this age cohort may be misleading unless one controls for tenure with the  determined in a manner similar to the average indexed monthly earnings computation for Social Security with the  6.2.1.3.2 automatic In recent  mo  enrollment, Plan nths,  Size  two      surveys EBRI plans  have  to  pr con ovided tinue  addi to cond tional uct  information  research in  on this  potential  area as publ  responses ic policy  from  evolves.  plan     25 The concept of measuring retirement security – or retirement income adequacy – is an extremely  67 VanDerhei and Copeland (230008).  72 20 4511  For  thro  example ugh 20 , a -minute  40-year  tele -oldphon  participant e interviews  would  with  need  1,25  to8  have  individuals  a tenure  (1,  of 004  at  workers least 10  and year  25 s with 4 retirees)  the current  age 25    and            . Testim Research ony.  F  U.S. und  Congress. and the Mil   Se bnate ank Me  Health, morial  Educ  Funat d,ion,  2001a.  Labor    and Pensions Committee. The Wobbly  contribute to a 401(k) plan),  behavioral economics has shown that the reaction of employees in  25–29 Previous impacts ability  to by  c. retir   research  pro in com Three educe ment e  quar  -what, by qu  income  art EB tile RI  eby rs, s.  has   most adequacy  or  The  demonstr     fina values ncial  in  ated Se vary  planni pte   that from mber ng  one   as   2010. tlow andards,  of  of the   53 It  most    was appears percent   also imp o   for used to rt an  be the  t to  quite factors  lo  compute west  generou  c -inco on tretir ribut me s: e  an quartile in ment  g80  to      to  current the 6.2  enti Modifying  ty em  isp  d loye esignated r  the .  For  exis  ex  asam ting  having p ltax e, partici   tre “run atmen  short pants t  of  of in   mone wo their rker  60’s y”  and at  with  that  employe   no tim  more e. r 40  tha 1(k) n  tw cont o years ribution  of ten s  ure had an  following modifications: (a) All earned income is included up to the age of retirement (i.e., there is no maximum  Looking at all households that would need to save an additional amount (over and above the savings  Copeland, 4 5  For Changing example  Craig, ,  tax an   emplo and ince nt Jack yivee e s  VanDerhei. ag  ............................................................................................................................... e 60 may  hav “Thee  very  Declini  recently ng Role  cha nofg ePriva d jobtse  and  Defined  rolled  Benefit over a subst  Pension antial  Plans:  account .........   Who    7   sponsors EBRI has  bee withn  respect publishin  tog  this studies  type  on  of  the  proposed  likely impa  modi ctfication  of AE for  of  sev thee  401(k) n years.  system.  In a joint  A survey  2005 study  condu  with cted    important topic.  EBRI launched a major project to provide this type of measurement in the late 1990s  employer older in th  to e Unit  be included ed States.  in  R  th andom is analysis.  digit  Alternative dialing was  specifications used to 31  obtain  of  a  minimum representative  tenure 73  cro  were ss section  used  with of the  essentially  U.S.    Stool: Retirement (In)security in America (T-166). 6 7 Oct. 2010.  The An situations  interesting  potential  similar  reaction  findi  tong  this  of  of   can the  em   be ployees AllianceBe  at odds  not  rwith nstein  currently  what  survey   mi parght  tof icipa   plan be tpredi ing  spo  inn c ted 401(k) sors  by  with   an plans   respect obj  w ective ill be  to   focused  extreme potential  strictly ly  federal difficult  on     retirement asavings perce  highn  of t  shortfalls real  77  income   perc incoent m  for  ead    for replacement Baby equa  the  Boomers cy  hig  for hest  the  rate - in and  Bo com  inomers  Ge  retirement, eneratio  quar  and tile.  nGen  Xe  when   rs  Th Xers  in e   401 simulated Oc  is teligibility (k) ober  accu  2010.  rates mulation  to parti  of     return csipate  are  comb are  in  employment expl ined ained  with  in    - 68 taxable  Copeland 13  wage  an  base d Van  constr Derhei aint  ( 201 and0)  the .   calculation terminate 14s at retirement age); (b) Instead of indexing for changes  average account balance at the end of 2010 of $26,649.   The longest tenured participants in their 60’s            . “A Behavioral Model for Predicting Employee Contributions to 401(k) Plans.” North American  balance from his previous employer to an IRA.    already facto Is Affected, red into  an  the d How.”  baseline  In Robert  model),  L.  the Clark  media  and nOlivia  perc en Mitchell, tage of  eds., additional  Reorie  compensatio nting Retirem ne  for nt  Ri Early sk    on ICI,  behalf  wed.  looked   of Double The  at  Principal  the  it   pote  Financial ntial chang  Groeu  in p  (2011) 401(k)/IRA  determined  accumulation  that if workers’ s as a result  ability  of  chan to ded ging uct  tra  any ditional     11 popula the In Se  same pte Appe tion. m  resul ber ndix  To t  further 2011, s.  B:    Brief  the  increase  Chronology  U.S.  Se representati nate  of  Fin  RSPM aon, nce  a    Committee  cell phone supplement  held a hearing  was  adde on “Tax d to  th Reform e sample.  Op ti Sta ons: rting    with  for several states that were concerned whether their residents would have sufficient income when they  21 5.1  Results From the 2011 Retirement Confidence Testimony by  Survey ...............................................................................  7  optimizi to tax  model  modificati ng  for  a  particular new ons  is incentiv  the  financial  impact e structures.  of  strategy.  plan   size  For  In  on   an exa  th  attempt meple,  exp  does ec  to te dbet  the  plan t er cur   sp und rent onerstand s oexperi r response.  ence potential  u  Th nder e  employee  reas  401(k) ons  plans to       in based more Social  average   de  Se retir tail curi  nati e ment in ty. onal  VanDerhei  Fig   wages plans. ure 8,     de the VanDerhei  and monstrates  model  Copeland  indexes  (A ugust that  (2010  based  if  2011  the ),  bu on  mo ) t  provid assumed  they st pessimistic  are es  aft information  base er-tadx  on   rate combi  a   st of on o  nre  cation ho htur aw snt  ithe  base cof  pro  plan  relativ dc  on ess  design  ass  ewith  evalue t allocations   and a mean  of    the      had been with the current employer for at least 30 years and had an average account balance of            . “Retir Actuaria ement l Jo  Savings urnal (2001  Shortfalls ).   for Today’s Workers.” EBRI Notes, no. 10 (Employee Benefit  Boomers desiring a 50 percent probability of retirement income adequacy would be 3.0 percent of  the 69 amount voluntary  Van  2001 Derhei  Man  of wave  enr  th  agem (e  o 200 of  llment 401(k)  th 9)ee.  nRCS t  .(V contribution  Ox , E) allford   401(k) data  University  are  plans   from weighted  to  taxable  Press  AE  by  plan   age for  inco s.  ,the   sex Alme though  ,Pe  and  was nsion  e de  ucatio we l im Researc  ha inat nd  to  eth dh  refle e,  C  65 advantage ouncil,  cper t the cent   2010: act  uof of al   the    using pro 122  ppl -ort 136.  an a idatabase ons  sponsors     in the    In October 2010 testimony before the Senate Health, Education, Labor and Pensions Committee on “The  reached Promoting  retir  Ret ement irement  age.  Security.”   After condu  Onect  ofing  the  studies  primary  fo rtopic  Oregon, s duri  Kansas ng the  and hearing  Massachusetts,  was an assess  wement  expande  of the d   5 Year-end 2011 data is currently being analyzed and the annual update will be available later in the year.  behavior expe allowct  researchers  an  with  increased  respect  to  sensiti ex  to tr ap a proposed volity at  by e be  sm haviors  eli alle mrination  plans  to this  to  of   population federal  deductions  tax  wi modi  for th  401(k) respect fications  contri  to:  hav   butions, e previously  the  2011 been  RCS  that 6.2.1.3 46 See  are  pages  a fu Impact nct  10io –1 n 1  on of  of  the   40 Miller  1(k) indi vidual’s (20  Balances 11) for  ag ean  at  in  exa   Re each m tireme ple.  year   ;nt  an  Age d (c)  Percentile distributions are established based on  The equity defined worker  original  return   beha bene  vvioral  fit eofrsion   accruals 8.9  assumptions   perc of RSP  en imM tp  and a  was ct  were retirement  a  used  mea  used  nto  fixed  analyze   in income  the -income   AE the  adequa  plans  futur  return  studi ecy.  economic  of  Fied  6.3 gure ,  45.7 percent  5  well cat  per e-gorizes c  being (expressed ent of  of   any the  the    lowes positive in  retir  nominal te-dinco   value  me  for     40 $202,329.  6  TheResearch  Potential  Imp Institute, act of Tax  Oc tReform ober 2010):  on Retir  2-e9. ment     Security ............................................................................... 8  compensation each year until retirement age to account for the financial and housing market crisis in  adult population. Data for waves of the RCS conducted before 2001 have been weighted to allow for consistent  responding Wobbly of tens of  Stool:  mill  toions   the Retirem   survey of 401(k) ent  would  (In  participants )securi  havety  le  inss   (going America,”  desire  ba  tock   contin th  ine  some model ue  offering cases  was  used to  their 1996),  to  401 analyze  data (k)  was plan  the  .unavailable  re   lative   with  potential the simulatio  benne fits model  and  to  c46 o  anseq  full-ublown ences  na that tional  may  mode  resultl  in from  2003,  a propo  and insal  2010  to modify  updated  the  it  fe  tod eincor ral ta pxorate     22 VanDerhei, Jack, Sarah Holden, and LuisJack VanDerhei, Ph.D.  Alonso.  “401(k) Plan Asset Allocation, Account Balances, and  70 VanDerhei (April 2010).  Gale, popula  Willia tion statistics m G.  Testimon  for each y. age  U.S.  cohort.  Congress.  Therefore,   Senate  it is  possible Committee  that  on an  individual Finance.  Tax who  Reform se preretirement  Options;  income     docume includednte  two d  by new  others.  questions.   Howe  Theve  first r, Figur  asked e 12  respon  showsde  the nts  average  how im pperc ortan ent t aisg  being e reductions  able to  in ded  401(k) uct their     terms). a 6population quartile   This define  value   dwould  ben   incr at ethe  eased fit obtain   astate ccrual  to  this  $3  lev  7, in  threshold, e56 tl.o 0  EBRI quar  for  par an tiles d ticipants    and the for   Milbank each in  in view  their  inc  of  oMemo  m 60  the e’s  grou  with  way rial  p. 2   Fu in - 5Th   nd, which yea e  largest rworking s of  Social  tenure  reduction  with  Se  and cu  the rit  $5y   in 3,  gove ben 10  at8 -erisk  rnor for fits   th  ratings are  of os   e with     74 47 comparisons;  Given the much  conseque  larger ntly  simulated , some data  acco  inunt  the  ba  2011 lance  RC  reSdu  may ctio  differ ns for  slightly  smaller  with  plans  da  show ta publ n ished in Figure  in previous  12, it is  waves of  2008 VanDerhei  and 200  (M9. arch   Similar  2012)  values  utilizes  are  th  0.9 e defin  perceent d contri  for Late bution  Boomers  particip  and ant  0.3 respons  perceens tto  for  the  Gen  RCS  Xe  questions rs.   A 90   importance respect ?  to Initial   how  of  partici employer  workers pation  -would provid  choice  ed react  s.retir    toe  AE ment  provisions,  benefits  and and  thus Social  si mulate Security. d the     likely response based on  41 While treatme several  the  signifi nt  $202,329 of 401(k) cant chan   value plan ges,  contributions  is  incl considerabl uding the  in y   impa larger exchange cts  than  of  for  defin  the  a  flat often ed -ben rate -reported e  governmen fit plan  overall  freezes, t match.  balance  automati  Gale  of  c$60,329, (2011)  enrollment    it is   6.1  Loan 20/2  Activi 0 Capsty  ...............................................................................................................................  in 2010.”  EBRI Issue Brief, no. 366 (Employee Benefit Research Institute, ....................  December 59  8   ranks             in . “Retir  the lowest ement  quartile  Income  ove  Ade r their quacy  remaining  for Today’s  work  history Workers:  may  How  indeed  Certain,  end up  How  with  Mu  an income ch Will  th It aCost, t would  and  rank     Promoting Retirement Security, 15 Sept. 2011, online at  retirement 5accoun A- 10 separate  years t bala  of   survey savings  tnces enure.  at  by    plan  Pa Social  Allia rt ico cip nceBernstei  ntribu Security ants intions  their  normal  n  fr 60  (2011) om ’s  with  retir their  provid  10 e ment ta -20 xable  years ed  age  pla  income   of du n  tenure sponsors e to  in  expected   encoura had  with  an avera   gth mo ing e difications gfollowing  ethem  acco unt to  save   que bal in response astion:  nce for  of     to  Oregon, betwee designed, n  set  the  an  out   even highest  in  the lower  and  late  perc  lowe  1990s ent st ainco  to ge  see  of me  the  if -specific  this  high  siest tuatio  defin -incneo  dcoul m  benefi e quarti d bet  addressed value le (27  quartile  perc  for en t) sOreg   takes would on  pla .  rea Thce ec  han for  th a  the ley s80 i s    71 VanDerhei and Copeland (2010)15 .  important the RCS. Data  to note  presented  that th  in e  pl taan bles -size  in  th distribution is report may  used  no  int  to thtisa lsimulation  to 100 due  model  to rou  is ndi  based ng and/or  on th o missing 1se found  cate  ingories.  the   In  percent probability of retirement income adequacy would require an even  larger increase: The median  above, as well as the plan sponsor responses to the AllianceBernstein survey, to parameterize the  6.1 the  r ?e sults 20/20 Decisions  of  Caps  academic   to opt  studies.  out once  What participa  weResearch Director  tfound ion has  was  begu  that n.  the overall expected improvement in  update provisions d a  20 for06  401(k)  analysis  plan  by s and  Gale,  the  Gru  recent ber, and  crises  Orsza  in th g eand  financial  analyzed  and  a  housing  plan that  ma  wou rkets. ld change    the 26  still higher  not  tha  a nfair  the  representation  bottom quarter  of in  on what e or  a  mo  furlel- career’s specific ye  participation ars. Second, the  in  aimpact  401(k)  of  pla  then  20 mi  per ghtc ent prod  limitation uce in terms  for    How 2011).  Does    Eligibility for Participation in a Defined Contribution Plan Help?” EBRI Notes, no. 9  $89,956 an http://finance.senate.gov/imo/medi d those with 20-30 years had an average a/doc/Testimony  account balance%  of 20of%20William%20Gale  $159,654.  .pdf   retirement. the In Fe  proposal bruary  2When  to 011,  modify   the confin  mod  the ede  federal  lto  was  full  used - time tax  tr to  workers eatmen  analyze  t(n=591),   of the  em  impa pl oye the ct  rof wei  and  thghte e  worker  2008 d results -2009  contri   were crisis butions   as in  the follows:  for  fi  4nancial 01(k)    plans  and   lowest focused percen-tincome   thres primarily h old. quartile.  on    simulated  For these  retirement  households,  weal  the th  with at-risk  a  comparis ratings dron op  to  36  ad  per  hoc cen  thresholds tage points,  fo rfr  om 82  EBRI theory, 6.2 /I nvestment  theModif  weighted  yCompany ing  the sample  existing  Institute  of 1,  tax 25 8(I  tre  CI) yields a  40 tment 1(k  a statistical )  database, of worker  precision   not and  th emplo e  of universe  plus yer  40 or  of 1(  minus  k) 40  c1( o nk) 3t ributions  per plans. centa  Ev ge ................................. ide  points nce of  (with the   95   9  37 voluntary percentage Suppose  enr  ofo  additional llment  U.S.  legislation mod  compensation ule of  were RSPM  enacted  in  for  order  Early  such  to  Boomers  estimate  that empl  desirin  the oyees  like g al ywere   90 impa  perc  no ct ent  of longer  the  probability  proposed  allowed  of  to  federal   deduct -tax     72 retirement  VanDerhei  accumulations  (September 20— 10especially ).   for the lower-income quartiles—were nothing less than  treatme the lowest ? Contribution nt - income of retirement  quartile  behavior.   saving may fall     in  dis  thr pro eeportionatel  ways:   y on 75 the part-time workers. For example, a worker who  of income replacement in retirement for several reasons:  (Employee Benefit Research Institute, Sept. 2010): 13-20.  In in real   exchange December  estate mark  for  2010,  an ets  18   the on  per   retirement National cent match  Commission  inco  fro me m  ade the  on  quacy federal  Fiscal .   governme   Responsibility   nt, by  and  plan  Reform  size and  released  age-specific  their long  salary -   retirement magnitude percent certain  of  expendi  possible ty) of what tures,  statistical  the  bu  results t  the bias  results in  woul  thisd  regard   be made  if all  can  it  Ame  clear  ber icans fo  that und  ag   in major e  Van  25 an D  dec edr hei, olision der  Holde  were s lien  ,ahead surveyed  Alonso  if  and  th  with e  Bass  state’s  complete  (201   1).   perce Whenn  we t to  modeled  46 percent.  the  Households Baby Boomers  in th  and e second  Gen Xers  income  in 2012  quartile  (Figur  drop e 1) be  25tween  perce n43.3–44.3 tage points  pe  (fro rcent m  of an  the  at-  VanDerhei, Jack, Sarah Holden, Luis Alonso and Steven Bass. “401(k) Plan Asset Allocation, Account  7 EBRI is currently in the process of integrating year-end 2010 account balances of 401(k) participants with their  Gale, William G., Jonathan Gruber, and Peter R. Orszag. “Improving Opportunities and Incentives for  retirement modifications retirement Not  income  at  on  all  projected important   ad savings equa cy  plan 401(k)  ...................................................................................................... Employee Benefit Rese  would  co  balan ntribu  be 4.3 ctions es  per  at  fr retirement cent, om  th toeir  arch Institut accoun  federal  age,t  ass  for  taxable u  the ming e (EBRI)  financial   the income.  mod  and  iIn fic   addi ahousin tiontion, s took g  4.3% market  suppose  effect        enters The spectacular.  study 6.2.1  the   work found     New force  that  Su  part  with rvey  ti me  Analysis the  whose  all- optimis ..........................................................................................................................  spousetic  already  assumptio  has an  full s, the -time  per  jobce  may ntage  be  of in  alowe  better st- income situation  quart  to attempt ile    10 to   ? Asset allocation.  47 ? Some of the participants in their 60’s may still plan to work several additional years before they  73 VanDerhei (October 2010a).  The awaited ac quartiles cu  third racy.   do  pa There for ne cu  workers lment   of are  Fig  ot  uon her re  curr   federal 4  possible  (page ently   8) deb  ag  sources  inet  sth  reduc  26 at –35.  publication of terror ion ,For   “The in  all all  shows     su Moment four rv ethe  yincome s,  distribution however,  of Tru  quartiles, th.”  tha  of   tAlt   may plans  h the ough  be   in average   more the  th eir EB  serious  RI/I  guidi perCIc en n 40  tha gtage 1(  prin nk)     ciples  population simulated risk rating “First,   life of  was  58 p  unlike aths   to perce  have  for  the n tretired   ad for  current equate  those  househol  sy   resources ins tthe emds  ,lowest  workers’  were  in retirement.  defined  projected  and firms’  ben   toe  fit have contrib  value  inadequa u  quartile tions to te   em  to retirement  33 ploye  perr76 c -en based  inco t forme  401(k)  those  for   in  2010 IRA account balances.   Preliminary findings suggest the need for analyzing the combined IRA and 401(k)            . “The Balances,  Impact  and  of  Automatic Loan Activity  Enroll  in 2010 ment.”  in  EB  401(k) RI Issu  Plans e Brief,  on  no.  Fu t366 ure  (Employee Retirement  Benefit  Accumulations:  Research  A  Savings by Middle- and Lower-Income Households.”  The Hamilton Project, Discussion Paper  crisis immed maximize An April  ini ately. 2008  that 2011  retirement  the    and  article  empl  2009.  con  introduced oyee t  ribu  had tions  to   on a  pay new  his/  federal  hmethod er income.  in  of com   Alt analyzing eh ou taxgh  on  EB  the  an RI y  isre t h insults ing  the  an   pro from  em cess  pthe l oye of  RSP attempting r contribute M.  Inste  to dmodel a dto  of  th   th e e  workers achieving the 80 percent threshold increased to 79.2 percent, while that of the highest-income  ? retire. Cash outs    at time of job change.  theoretical databaseNot  in  cal  20  too c10 ulations   vs. im p 20o 08 rt ofan   sampling Det ......................................................................................................... partment  error.  of  Labor These  (D  incl OLude ) Form  refusals  5500  to  for  be  all  interviewed  401(k) plans  and  an dother  suggests  forms  an  of   5.0% un   der  - and reduction One  values point 6.2.2   for th  (pat  ages plan  Caveats had   13–14) sponsors  previo  for  Thi susly p  in ecifi s Researc   bee the cally  ntwo  ma  hme  ..................................................................................................................  sde m ntion  allest clear  the   in plan   the nee  si  academic dze  to  categories  keep  America literature  (less  sound than   was  $1   that over  million  some  the  and long  wor  $1–$10  run ker sby       12  balanc the high es accoun est when  defin  attempti tse  would d ben ng e  no to fit   assess  longer value  any quarti  be  fo exclude rm le)  of  w  com hdile  from  pthos rehensive  income e in the  retirement  subject  third and  to  income   tax highest ation,  ad equa in  contri com cy.e  b  quarti For utions  example, le  to  drop  IRA    s24     basic retirement expenses plus uninsured health care costs.  This is 5-8 percentage points LOWER than  Simulation Institute, December  Study Based  2011).  on   Plan Design www.ebri.or  Modificatiogn s of Large Plan Sponsors.” EBRI Issue Brief,  74 VanDerhei (October 2010b).  impact on2006  part--02. timers  The  on  Brookings  a longitudinal  Institution,  basis, the  2006  current .   analysis filtered out any 401(k) participants with annual  simply computing employee’s  an  re  overall tireme  percentag nt savings eaccou  of thnet  simulated  in the year  lif  ite  was path  contri s in a bu particul ted. In ar  excha  cohort nge  tha  for t will  this  not     Subsequ quartile workers ent to the  increased  release  of to  the 64  perce Oregonnt.  study,    it was decided that the approach could be applied to    ? Even though a participant has at least thirty years of tenure with the current employer, it does  representatio nonresponse,n  th  of e  effe small cts  plans  of qu  for estion  the  EB wordi RI/IC ngI  40 and 1( k) question  databa se. order,  The  and  plan  scr -size eening.  variable  While  was  attempts  specified  are  in  terms made  to of   6.2.1.3.1 Looking only Age  at  and those  Salar  households y   that had exposure to the market crisis in 2008 and 2009 from all three  imple Van million defaulte Derhei m  in enting d   assets) (A into pr il  “policies a  201  401(k)  are 2)  more anal   AE toyzed day   plan than   th to  (witho  e  1.5 en  median su  tim re uet  that   sautomati ratios  the  future  value  of com c   escalation gen ofb ined the erat   average 40 ion 1(  provi sk)  have  an  per ds ions) IR  retirement cAen  bala  would tage nce  reduction s   contin as security,  a multiple ue  for   to affordable    plans contribute of 401(   k)      what and 21  we  perc would  found en tno  aing e  longer our  points,  200  be 3  respectively   analysis. tax-deductible,     .   and any employer contributions to a 401(k) plan would be  income ofno.  less  341  than  (Employee  $10,000 as  Benefit  well as  Researc those with h In  less stit  tha uten,  two April  years  2010).  of te   nure.   Somewhat important .................................................................................................. 27.8%   7have    Future sufficient modification  work...........................................................................................................................  retirement  of the  income  current  to  tax  pa  incentiv y for thees,  simulated  assume  th expenses, e U.S. govern  the new ment  method  would..............................   match compute  18% d the  of    13  other states as well. Kansas and Massachusetts were chosen as the next states for analysis. Results of  Many 75 Van  of Derhei  EBRI  (Fe ’s previous bruary 20 11 simulation ).   projects (see Appendix B for a brief chronology) will be directly  participants minimize VanDerhei,  thes   instead Jaeck,  fact  and ors,  of  assets,  Lori it is  iLu m  bu possible cats.  a  “The similar  to  Imp qu  distribution anatify ct of  th  Auto e err  would o -enrollment rs that  be expe  mayct  r  an eed sult d  in  Au   from the tomati  latter  them. c  case.  Contribu  The  IfRC  this S was tion  is i ndeed co  Escalation -  the    not mean not the 401(k) plan has been in place for the entire period.  Moreover, there is no  Helman, fronts (d efine Ruthd , Craig  cont ribu Copeland, tion plans,  and  IR JaAs, ck  VanDerhei. and net hous  “The ing  e 2011 quity  Re ) shows tiremen  a tmedia  Confidence n percen  Sutage rvey :for    Early  balance by age and tenure for individuals with both 401(k) and IRA balances at the end of 2010.  For individuals in  sponsors health Surprisingly, at the  defau care,  in  any l an ted the d  of    cfina Wall o  the ntncial ribu  Stree  larger t  fr ion teedom,”  -Jour  size ratenal  categ  that   article th eothe  ries. docume   pla reported   n sponsor nt pu  on tsl y  forth had  the  chos   most an example en  pessimistic  (typic  that ally  in  set wou  th  of eld   range assumptions,  modify  of  re 3 per tire  cand men ent   t  treated as taxable income to the employee (just as current wages are). Second, all qualified  60 Figure 11 shows the baseline average percentage reductions in 401(k) account balances at Social  percentage whatever  of households  was contrib  thatu te would d to  ameet  retirement  that requirement  savings pla n mor . Wha e tha t do n a  you  spe  believe cified perc  wou enldtag  bee  the  of ti  most mes   Figure the While  Kansas   the 6 provi  passage  study des  were similar  of tim  pre  ein  sallowed formation ented to  mor  the  for e  state’s   eligibility funds  to Long  be  in-   saved, defined Term Care  the  con   iSe m trprovement rvices ibution  Task  plans   Forc over  fore   the  Gen on  last July  Xers   ni 11  ne in,  2002,  2012. years    isIn      sponsored case, 37 Gale  the  (201  RSP  by1) M  th .  estimates e Employee  fo rBen  overall efit Research  average  be Institute nefit re  (E ductions BRI), a private,  presented  nonpro  heref it, woul  nonpartisan d32  be expe cted publi  to c policy  be smaller     applicabl           . “Falling on e  to Re  such tireme  Sto cadditional ks: nt  What  Income  Will  research  Ade  Happe quacy.”  an n dto   we EBRI  Retirees'  will  Issu  bee   Inco happy Brief, mes?   no. to wo   The 349 rk   Worker (Employee  with the  Per  C oBenefit smmittee pective,”  Research   on Pr esentation Ways    and   guarantee Very importa  that nt  ............................................................................................................ the employee who is participating in 2010 has done so the entire ti me 61.5%  they    were  their 8   60 Conclusions ’sConfidence , a medi a............................................................................................................................... n ratio  Drop  ofs  1.2  to 3Re  was cord  found  Lows,  for  Refle individual ctings  “the with  at New  least  Normal”  thirty years   March  of tenure  2011,  with  EBRI ..........................  the  Issue  curre  Brief nt    13  Boomers of 5.6 percent for a  50 percent probability and 6.7 percent for a 90 percent probability of  plans did of 76  compe Van  no  by tD  cite erhei  capping nsation).  any  (April  of  an     20 the Tradi nual 11  other ).  t“tax ionally,  15 -preferred  combinations  and in  contrib the absence  ofu tions assumptions  of  to  these  [the]   AE in lowe   th preovisions, r  study of $20,000 . Th  many e article  or  workers  20%  also  of   reported  ha income” ve chosen   only     employer and employee contributions would be eligible for a flat-rate refundable tax credit,  61 Security normal retirement age due to expected modifications of plan sponsors and participants in  in 6.2.2  the  simulation. likely Caveats  chang  fo   r eTh  tois  your  Research  plan?    than largely this research and   case the those  du   results or  we  etha ganization,  to  se t  the would e  of  that   th fact e   be  an the   Massachusetts that  devi   Mat nu d  in enced m h  2003 ew ber   Greenw  by of  very    the  future study   full fe aldw    40  & were years  401(k)  1( Associates, k)  pre  univ the  sponsors s eented workers rse.  Inc.,      on  aused   Washington, are  Dec.   eautomati ligible  1, 2002.   for DC c , e  par based   nWi roticipa tllment h  market  the tion   assistance (A    re in E)search   aprovisions  defin   of firm. e dthe        Means to Institu for  pro  The vtide  e, Ec   November on cost/be omic n Crisis efit  2010);  assessments  of 2 008: and  DCIIA What  of   Rese  W these ill arch Hap  type p  Re en  proposals port  to Ret  (November iree  ins’  the  Inco   fu 2010). mes? ture .2009     APPAM Fall  employer. This number increased to 8.53 for those with no more than two years of tenure with the current  eligible #355  .  retirement income adequacy.  Younger cohorts experience a similar increase, going from the all- (page 38 to results   start The  31).  analysis  un  contr der  Thii sbuti the  fo  isr   often the threshold ng  at Sena   rae  6ferred te  perce   Finan of a  to real n cet  as  rate Comm  re  the placement  (largely  ittee “20/20  hearing  in  cap.”   rate response  modeled   of 80  to percent,  th  the e following  matching  while  scenarios:  Fig  con ure tribution  9   shows  ince  thatntive  eve n  given to the employee. Third, the credit would be deposited directly into the retirement saving  If 9  oneReferen  were to ces  look  ...............................................................................................................................  at this from a strictly financial perspective, one would assume that the ...........................  lower-income  14   77 VanDerhei and Copeland (June 2011).  reaction to the proposal to modify the federal tax treatment of employer and worker contributions for  The 2011 RCS data collection was funded by grants from more than two dozen public and private organizations,  contrib and Kansas  the  uInsurance  parti tion cplan ipation   D ma epartm  kes rates  a e  treme among nt, EBRI ndous  the  was  lowe  difference  able r income  to crea  in  ttheir employees e Re tirem at-risk e  (t  nt raho  tings. Readiness se most  For  example, li  Ra kely tings  to  be based  Gen  at  risk) Xe  onrs   a  was with  full   no  employer.  Conference (November 2009).  53 ? Many of the participants in their 60’s have already started to withdrawal money from their  Responses 48 household 6.2.2.1 As  The explored  EBRI Plan   /ICI analysis we  in r   Participant th Size e obtained e June   to the  20 -Dire   more from 11 cted Issu   1,018 select  eRetireme  Brief,   pl gran  oup. the nt  sponsors   Plan RSP    M Dat  al  agrouped lowed  Collection  retir  in  Project teoment  six size  (th -inco e  ca  EBme tego RI/ICI  adequa ries  40  based 1(k) cy  data  to  on  be ba  total  se) assessed  is   the    decreasing provided by  the  the  thr  em eshold ploye rto ).    aHowever,  70 percen  sot m real e parti  replacement cipants in  rate  AE pla  would ns –  wh incroe  otherwise ase the perc  mig ent hta ge have  of   account, as opposed to the current deduction, which simply results in a lower tax payment than  10 8 Appe Conclusi ndix ons  A:  Brief Description of RSPM    individuals (those most likely to pay no or low marginal tax rates and therefore have a smaller financial  Holden, Sarah, and Jack VanDerhei. “Can 401(k) Accumulations Generate Significant Income for Future  401(k) with ?  staff  plans Em  time ployer  in  dona  exchange  coted ntributions  by  EBR for Ian  and  are  18   Gree modified  pernwald. cent  in  match   RC sucSh  mater  a  fro  manner m ials  the  and  tha  federal  at  the list  of total  governme  underwriters  match (emplo nt,  by may  age yer  be  plus   and access  gover  aged e- specific natme  the nt      Even if one were to ignore the potential interaction of the proposed limitations with the present values  stochastic decumulation model that 77  took into account the household’s longevity risk, post-retirement  future quite low.  years   Wi  ofth  eligibili  the adoption ty are simulated  of AE in th  toe  run past  short  few  years, of money  the  60.7 parti cpercent ipation  of rates  the  have  time,  often  wher  increased eas those  to  10  Appendix A: Brief Description of RSPM .......................................................................................................... 16  78 VanDerhei (July 2011).  largest, most accoun  representative t balances.  repository Figure 7 shows  of information  the average  about  account  individual  balance  401(k)  for plan  similarly  participant  long  accounts. -tenured  See     at retirement  retirement  plan  ages  assets.  later   than 65.   8 Year-end 2010 da 43ta is currently being merged with the consistent sample.  “successful” voluntarily chosen  retirem  toe n parti t evcents ipate  by  at  19  a  hi perc gher en  ctaoge ntribut  poinion ts for  level  the  –in  lowest stead -inco  might me  si  quartile mply allow  and  their 12   savings            . Testim otherwise.” ony. Joint    DOL/SEC Public Hearing on Target Dates Funds. How Would Target-Date Funds  EBRI incentive  WebRetirees?”   site: to d e www.ebri.org/rcs. duct  EB  retirement RI Issue Brief,      savings For  more no.  contrib 251  detail,  (Employee u  see tions  Helman,  from  Benefit   taxa Cop eResearc ble land,  income)  and h  In Vastitute,  nD would erhei   Nove  be (Mar  least ch mber  20  likely 11,  2002  online  to).  rate     at   salary  Given  the quartil match)  muc es.  hremains  larger  The  avera  simulated constage nt.  per    accou centag nte  bala  redu nce ctions  reductions  for the  youn for smaller gest cohort  plans  (those shown  currently  in Figure  26–35) 12, it is   of accruals under defined benefit plans and/or the existing tax preferences available to some IRA  investment values with EBRI One   of twen has   in the   documen ex ty  ba  ce risk,  orsic ss  more    of and objectives  t80 ed   ex year   per ap  si ocsgnifican su ent.   of of re   future  RSP  to potentially tM  re   is eli duction  to gibility  simulate   catastro in  would  the  th  perce  ephic only  per  n  nursin ex ctage en pe tage ri  of gen -  home simulate of ce  the this  and  popula  sidtua   lif hoepaths tme ion tion -  18.2 heal  that  for th  per   will retired -care cent  be  risks.    “at of the  risk”  Th  is   VanDerhei, Holden, Alonso and Bass (2011).  participants who were 55-64 in 2010. However, in this case only participants with a positive  to perce  start ntage  – and  poin  remain ts for  –th  ate  the highest 78  defau -inco lt rate. me quartile   As a result,  under  they  the  were  all-pessimistic  lik 44 ely con tset ribut  ofin  assumptio g at a lower ns.  rate    than  79 VanDerhei (August 2011).  www.ebri.org/surveys/ 11  App Likely endix  Impa  B: Br ctief  rFutur cs  Chronolo /2011/ e 401(k) ). gy of  Contribu  RSPM .......................................................................................................... tions? (T-160). June 2009.    17  this as “very important.” However, those in the lowest household income category ($15,000 to less than  important 9  are In  The  a  largest July  pro  2011 posed   to fo rnote   Notes  those regulations  that   article in  the the  for   lowest plan  ,40  it1(  provide -size  k) - income plans  distribution d  we  preliminary  quartile re published  used  (22.2  evi  in  ind    this percent). November enc  simulation e of the  19  Th  impa 81 e   mod reductions anct d  mu eofl  is ch the  based  of   “20/20 for  the   the  on growt    tho ca youngest ps” hs  in e   found these      contributions, this alternative formulation of capping tax-preferred contributions would substantially  was time. Regarding  followed      the  by  proposed  the expansion  tax cre d ofit,  RSP  GalM e  (2011 and the ) repo  Retirts remen  estima t Readi tes from ness  the Rati  Tax ngs  Policy to a national  Center  model for both  and  an     households of having retirement  “at risk”  for inco  inme adequate  that is  ina retirement dequate  inco to cover me bet  basic ween  ex penses 2003 and  and  201  pay 2,  for based  uninsu  in lared rge  heal  partth  on     6.2.1.2 ? AllParticipants  plan sponsors    drop the plan match, and all employees receive a 30 percent match from the            . “The Influence of Automatic Enrollment, Catch-Up, and IRA Contributions on 401(k) Accumulations  value for the sum of employee and employer contributions in 2010 were included.  The year-end  if they been working for a plan sponsor offering a VE 401(k) plan AND had made a positive election to  While 49 See  we endn  found ote 17  no  of  signifi VanDerhei cant  and trends  Copela  by nd age  (July  demogra  2010) fo phic, r more  Figure  detail.  2 shows    that the lower-income  54 $25,000) plans took  actually  place in  th have e next  the  fe  largest w years.  per   centage of respondents classifying the tax deductibility of  in proposed cohort  the EBRI/Investment  de cre by athe se  to National  13.0  Company  perc  Commission ent for  Institute  those  on  Fiscal  in (ICI  the ) 401(k)  Res  second ponsibility  database, -income  and   quartile not  Reform  the  unive and  on  re  projected rse ac h of  a  401(k)  mini  retirement m  plans. um of  6.1     reduce 80 The  Van  other D  the erhei  sta  current  (Se tistic ptember  attri  limits but  20  aed 11 vailable ) .to    EBRI  un  in de  the r qualifie  articled  de dealt fined  with  contri  the bperc ution ent  plans. age of  C  AE urrently, -eligible  the  wor   kers who  the 18 24 care the     12         In perce   ad     presenta   costs th . v  e“App e  Re RCS n n ttfor   tiremen ecr of ,ndix  tion retiree edit   the auto  C:   of  an remainder  -Imp t enrollment   refers th d In  e acao   c30 firs tme   of to  perce t    th iAde mi  ndividuals ofe  cro  in  th fiquacy nnancial  401 eir t- simulation credi  liv (k)   who Af es  and  tplans; t.  eonce Th  are r  ho  PPA e   retirement using retired  paper  howev they  and  market  retire.    in or FAS ecr,  who l ud -for  158: income  crisis e   are   sGen  Howe a  Part     in ag d iXe -  stributional e 200 a  dver, One—Plan 65 requa  household 8  or  and  the  older cy  20  EBRI  mod  09  ana and  Sp  son  Retire  eon in l  not yl  retir s   bui sors' ithe s emplo  of ltmen e  ment two  in  the Rea  part ty   ed lowe Readiness winners c  readine tions   full from st  time. .” -s    sand  .  19       government.   at Retirement.”  EBRI Issue Brief, no. 283 (Employee  Benefit Research Institute, July 2005).  2010 average account balance for this group was $255,075.  Projections were also performed  households 4 participa The value te.  ar eof  mu  tax ch -fa  mor vore e dlikely  reti rto eme  bent  at  accounts  risk for insufficient  under curre  retir nt etax ment  provisions:  income  (eve the ncase  thou  ofgh  401  we(k)  model     49 62 contrib With respect utions to  as  potential  very important  worker  (76.2  reactions  percent).  to this     proposal, a new set of questions concerning  accu Eviden mulation ce of the s.  magnitude of possible statistical bias in this regard can be found in VanDerhei, Holden,  percent for those in the third-income quartile. The reductions increase to 10.8 percent for those in the  50 combina Worker would  Van Dbe  refers erhei t ion expecte    (Se of to  all pemployee tember d  individu  to have  20a  and 10 ls  larger  who ) also  em  are   pdemonstr tenure loyer  not defined  contri -specific atesbutio   tha as  wor reti t eligibility nsrees, k er is  th con  regar e  for lesser tribu  daless  defin ti  of on  of  ea   rates emplo d  dollar  contrib  had yment  limit u tio they   stat nof  retirement   been at us.  leas    VE t $50,000 - eligible plan has    a  losers administrativ indexed  under  pre  the e -re  401(k)  ttwo irem  versions  edata nt income  at th  ofe  the   EBRI quar  proposal;  tDe ilecs,e mb the  however, e  at r 200 -risk3  perce  policy  the nunde  tages, forum. rlying  while  The  analysis  much  basic   sholds model maller  retirement  was  (the  th y were en  saving  85     10 Rating  Holden ™EBRI  also  and   Issue provides  VanD  Brief, erhei  information  (2  no. 002  307 ).   (Employee  on the distri  Benefit bution  Research  of the li  Institu kely nu te, mber  July  of 2007).  year s before those at risk  81 VanDerhei (November 2011)  based on 2010 asset allocation and contribution behavior as well as subsequent market  plans 13 our    basic ?    End All  retirement  n plan otes  sponsors  ...............................................................................................................................  expe  drop nses  th eas  pla  a nfunction  match, and  of the  all employees  household’s  rec eexpected ive an 18  re perc tiren em t e match nt income  fro .......................... m the ). The    2012  20     33 34   participa Miller, Judy nt behavior  A. Testim  in ony  res  Spubmi onsette  tod  the  on  specific behalf of  federal  the American  tax modifications  Society of  proposed  Pension Professionals  in Gale (2011)  and  was     significant highest Alonso -and inco  posi  Ba me tive ss  quartile  (2011).  impact .  on Th    ereduci  thirdn gpane  the al dditional of Figure  comp  4 (page ensation  8) in  most that  publica families tneed ion shows  to achieve  the  distribution the desired   of  per instead. This  year  anchoring  The  and  simulation   aeffect  perce  can n tage re  be sults   limit seen 79  we   of by  provided  100  loo kperc ing  showed  at ent  the  of  to an  that p  employee’s -income  approxi  quart m  compensation. ately ile  in 60  th  pe e rcent 2005   study, of the  AE wh-eeligi re the ble   modified contrib The viewus e tions  for xpr  testimony  essed i constan n this statement are s t  for for  both the Se  em nate ploye  Special olel rsy  and thos  Com  par e of Ja m ticipants ittee ck VanDer  on  (page  Aging hei  6). and sh  in  Gale 2004 ou  men  ld not to qu tions be attribute anti fy that  the  th  dben e to the  proposal eficial    “run perce  short nt for  of  the  money,”  lowest  as income  well as  quar  thet ile perc  and en t65 ag  perc e of en compensation t for the second  they  lowest  would  inc neo ed m  in e  quartil terms eof  in  additional  2003)    The second question asked of those currently saving for retirement was “Suppose you were no longer  The 25 Actually,  August performance  th  2011 e co nstraints Notes .  Th  article  e would  estimated  nee  evaluated d to  year  be comp  -th ened  im  ared 2011 po  to rt  average an  the ce  40  of2(   defin accou g) limit end  tas   be balance  well nefit  as  pla  any forn   pl sthis  an for   -grou  specifi househ p cwas  constraints olds,       government.  11 It should be noted that this combination is in essence adding a nominal annuity for the 401(k) accumulations  baseline           . “M  ratings easuri nfor g Re  Early tirem  Boomers ent Inco  ranges me Ad equa fromcy:  a pro  Calc jection ulatin gthat  Reali  87s tic percent  Income  of  Replac the simulated ement Rat  lifepaths es.” EBR   I  82 VanDerhei (March 2012).  Actuaries. U.S. Congress. Senate Finance Committee. Tax Re 63form Options: Promoting Retirement  included level of retirement  in the 2012  income  RCS .ad  Specifi equacyc.ally,    workers currently contributing to a workplace retirement plan  plans in the EBRI/ICI 401(k) database in 2010 vs. 2008 Department of Labor (DOL) Form 5500 for all  Employee Benefit Research Institute (EBRI), the EBRI Education and Research Fund, any of its programs, officers, Given workers median  the  repla  would  phecnomenal ement  immedia  rate  gr tel  owth for y be  the   of better   top def-iincome n  off ed  in co  an nt  quar ribu  AE tpla tile ion  ndecreased   plans than  in (especially  a  VE by  4 plan,  per  those  cand entage   th with at  points  over  a 401(k)  ti  for me   th feature)  (as e  scenario  in    savings are impact “could    still   conceivably   of to ex   atha r  m em ve andatory e  aly  50,  hig  affect  70, h  con (78  or  in   t  c90 per ribution ent  per cien vectsen    for of for t   5 probability  the firms  percent  lowest  to offer  of   of income  compensation.  retirement  401(k)s  quar  ort  ile pe inc  nsions” and o  me  46  ad   perc (page equac en  7) yt.  for   but  the  concludes  second  lowest that    allowed on  tax co  to ntrib  deuduct tions  retirement  (primarily for  savings  the Hi ghl plan y Compensate  contributions d Em  from ployees  your ).   taxable income. What do you think  assuming they retire at age 65, while demonstrating the impact of defined benefit plans in achieving  Measuring $272,681  the impa  and ct  the on  ol value der cohorts  for the  end (those  of  fir over st quarter  age 35)  20 is somewhat 12 was $292,258.  probl  low e matic est income  in that  quartile  the values  under   with VanDerhei  a real an  (Jnu uly, ity  2011)  from So  provides cial Secu  preliminary rity.  Given the  evidence  larger replacement  of the imp arates ct of  for  these  the  “20/20 caps” on projected  for retired Issu  lowest e Brief, -inco  no.me  297  households  (Employee  ar Benefit e at risk  Research  to only  Institute, 13 percent  Sep  fortember  the simulated  2006).   lifepaths for 48 16 were asked: Security    .  15 Sept. 2011.  In trustees, spons  later EBRI analysis ors, or other staff.  (VanDerhei,T  Novem he Employ ber 20 ee Benefit Re 11), the follsearch Institute is a nonpr owing scenarios were added: ofit, nonpartisan,   education and 401(k) plans and suggests an under-representation of small plans for the EBRI/ICI 401(k) database.  The  the automatic with  private  a 3 per  es  sector ccalation ent contri  in the  provi bution  lasstions  three  rate  took   and deca  eff  default deects,  for it  appears investment  some of  that  the  in   this worker  a mo  form ney s)  of tha  mark  em t would petloye  fund.  rincrease -provided  Howeve  to  re  85 r, ti  rpercent. ite m would ent     this seems unlikely. He also dismisses as likely overstated the concern that the tax credit/matches called  income 51 See Van  quartile Derhei  in (A ug 2012). ust 20   Of 11)  course,  for evide  wh nce e nof  one  the  liimportance mits the analysis  of participating  to those  in  who  a defin  are ed  simulated  benefit pla  to n.  be  you (and your spouse) would be most likely to do?” When confined to full-time workers (n=460), and  retirement are Social  influe  Secunce rity  income d , the  by  disparity plan  adequa -spo  in ncy  sor favor  for  and  Baby  of  participant the  B  lower oomers  income  reactions  and  Gen would   Xers. to  increase  the    tax  as  proposal  the retirees  as well  grow  as  older.  the di   stribution  retirement accumulations. If the 20/20 caps are assumed to be imposed starting in 2012, the annual   The first major modification of the model was presented at the EBRI May 2004 policy forum. In an  26 retired Appen  1.4 per d  hig ixc  C ent h edesc s  re t inco sponde ribeme s how d  households.  tha  households t they did    not Similar  (whose  know.  tre    heads nds are  are  evidence  currently d for  ages  both  36–62)  the Late  are  Boomers tracked through  and Gen     research organization established in Washington, DC, in 1978. EBRI does not take policy positions, nor does it lobby, Since 1999, average balances are also computed for a “consistent sample” of participants to control for  plan-size variable was specified in terms of participants instead of assets, but a similar distribution  plan     appear          will . “Defined  that  provide  this  Benefit was  a substantial  more  Plan  than  Free  per  offset zes: cen ta Who's  by ge  the  of  Affected,  non increase -Social   in How  Security  parti  Mu cipa ch  retirement ,t ion and  for  Replacing  the  we  lowe al tLost hr -for income  Accruals.”  Baby quartil  Boomers  EBRI es    for saving  in the  in  the proposal  future,  may  the dis  nucm ourage bers im  em prp ove loye  substanti r matches ally:  to  among 401(k)  Gen plan s, Xer  but  house  offers holds  no su  without pporting  any  data  fut ufor re   Mitchell, ? No  Oliv  plaian  S., spons  Steoprsh e drop n P.  the  Utkus,  plan  and match,  Tongxuan  and all employees  Yang. “Turnin  receive g Workers  an 18 percen  into tSavers?  match from  Ince  the ntives,      eliminati 1. Suppose ng those  you  who  were  refu sed no longer  to answer  allowed  or responded  to deduct  that your  they  retirement  did not  savings  know,  approximately plan contributions  1 in  4fo  fu r ll- of tenure with the current employer within each age group. For example, if a 401(k) participant in the  80 perce 52 The    Van  Wall nDtage erhei  Str  re  e and et ductions  Journal  Copeland  in  article   401(k) (2002  did a )  account .  not report  bala nces the positive  at Social  impa  Secur ct iof ty  auto normal -enrollment  retirement  401  age (k)  are plans  displayed  on  7  Xers. analysis retirement    to  de age, termine  and how  the  their  impa  retirement ct of annuitizi  income/w ng definealth ed co  is nt  simulated ribution and  for  IRA  the  balan  following ces at  com  retirement ponents:  age,     advocate specific policy recommendations, or receive federal funding. 12 Tergesen (2011).  Finally, the September 2011 Senate Finance testimony  analyzed the potential impact of various types  the downward bias that would otherwise exist from IRA rollovers when 401(k) participants change jobs.    would be expected in the latter case. If this is indeed the case, the RSPM estimates for overall average  and due 27 See   Gen to  Hol auto Issu  Xers den -ee   .Brief, and n   rollment, Unfortu  Va  no. nDer  n291  at hei reely, sulting   (Employee (20  the 05) . in  “success”  subst  Benefit an tof ial   Research these  incre aplans ses  Institute,  in  are  th eir someti   re Mar tirmes echm  2006). e  measu nt accumulations   red by metrics  (for  tha thet  are  this assumption  eligibilitygovernment.  for participa   tion in a defined contribution plan, the at-risk percentage is 60.7 percent, but it  time workers federal  (25.6  inco  percen me taxt)  pu  indi rposes cated  and  that  that  they  anything  would  reduce your employer  (in some  contributed  cases completely  to your ) th  retir eir ement  oldest cohort (those currently 56-65) has recently changed jobs and has a rela vely low account balance  in many  Figu  worker re 10 by s  who age  and bega  age n to -specific  save for  inco  retirement me quartiles  due  for to AE.  all  401(k) As with  participa  any chan nts ge  with , some  salaries  peopl e in  will  excess  not  of  VanDerhei and Copeland, 2004, were able to demonstrate that for a household seeking a 75 percent  Principal Financial Group. “New Data from The Principal Reveals Power of Employer Match.” News  50 of VanDerhei,  tax-reform  Hol  options den and  on  Alonso  retire m (2010) ent income  report  adequacy.  on the average   This  was acco  eunt xpande  balances d in the  among  Novemb  401(k) er  2011 EBRI  53 benefit  This material  reductions  first  presented appeared in  here  VanDer  would hei an  be d  Co expecte pelandd  (July  to be  20  smaller 10).   than those that would be evidenced  not same  at  scenario  all relevant  as mentioned  to the potential  above,  for  the  defin  third ed- income contribution  quartil  plans e’s me  todian  provide  repla  ac  significan ement rate t portion  increased  of a  2     Knowing 13  ?Holden ?Social No  the  and   Se pla  perce  curi Van n spons ty. Derhei n  tage ors  (2 drop  of 005  households  the ).   plan match,  that  and will  all be  employees  at risk for  receive  inadequate  a 30 pe  retirement rcent match  income  from the  is   drops all the way to 18.2 percent for those with 20 or more years of future eligibility.     contribusavings tions if  the this  abi year lity  on  to  your  ded ubeh ct them alf was  was  also  eli  treated minated.  as  The  part  low  of eyour st-income  taxable  category  income.  ($15,  Suppose 000 to  th  less e    in 28  his See  or  Van  her Derhe  current i and  401(k) Copela nd plan,  (200  any 8).  reported decrease in contributions would have a much larger  $10,000 have the  and  mos 81  tenure t desirable  of at  results;  least two  bu tyear  if the s.    focus of auto-enrollment is to increase participation among 82           . Testimony. U.S. Congress.  Senate Special Committee on Aging. Do We Have a Crisis in America?  probability These two  pa ofpers  retirement  provide  in  an co  interesting me adequac  analysis y, the addi  of ational  proposal  savings  with  that  profound  would  publi otherwise c-policy  ne ed implica  to betions.  set    Room: News Release Archive, November 30, 2010.  www.principal.com/about/news/2010/ris- Issue participa  Brief nts  present and a new  from  set  ye  ofar  survey -end 1999  results  throug  wereh  ayear dde-dend  to  the 2009  mod  ande lfind  in th  the e March  overall  2012  average  Note  increases s article.     by the full 401(k) universe.   important worker’s percentage government.  pre  for  points, -retirement  public  the     policy  second  income.  analysis; -income   For  however,  example,  quartile  equally   incr some ea   analysts sed important  7 per  will ce  isn   merely tage knowing  points,  report  just  and  how  the  the   average large  lowe  thst  ebalance - income  in     54 than  The  $25,000  nomi governmen nal)  has cost  the of t  matched th  largest ese expendit   ne 18% gative  uof res  contri   re increases actbuti ion o  to with ns  this  so  co   that proposal, mponen  for tevery -speci  with  f$100  ic56.7  inflation   you perc  or  ent ass  your u  imptions. ndi  em catin pl oye gSee  a savings r  the       impact 14 IRA rollovers  than it  would that origi  onn ated the same  from  40 indivi 1(k)dual  plans  (with  are included  the same  in th  survey e projec  response) ted accumulati  had tha ons.t  worker not  lower-income Results  parti  From cipan  thets  EBR  (and, I-ER  as F  Ret a res irem ult,e nt their  Se cretir uritye ment Projection  financial  Mod  preparedness), el, 27 Jan 2004 8  objectiv (T-141).e    analysis  The aside Given ?  assumptions Defined   each that  the year  contr  financial  until  based ibution  retir   fate one  balances.  ment responses  of future  to ac     h(o ge iernerations v  lack e this  ther  obje  eof), ofctive  re tirees both  would  from  appea  dec  individual rrsease  to be  by   so workers  a  strongly median  and   amo tie  the d unt to  plan  whether  of  30    match-stats113010.htm   29 The full stochastic nature of the model will be included in future analysis.  th from $67,420 at year-end 1999 to $131,438 at year-end 2009, an increase of 95.0%.   Two points stand out immediately:  defined accu quartile mulate   con incrdter  adefici ibution sed 14 ts   plans are per clientage kely  (mo  to st   comm be. points  Theonly ).   aggrega In  the sum,  401(k) t  w e h defic ile  subset  some it number   wor of this k ers with  universe  saved  the curr  le) ss and ent  than  attempt  Social  they  Se  migh  to cu  assess ritt y    EBRI 1100 13 St. NW #800 Washington, DC 20005 (202) 659-0670 www.ebri.org appendix for more details.  reduction. contrib   uted to your retirement savings plan this year, the govern 51ment would contribute $18.  recently 39 VanDerhei  chan  (Se ged ptember  jobs and  20 11 had ). The  a significantly  2011 RCS questio  largerns  4  were 01(k)  fielded  balan cine .January  Therefore,  2011  the  and  analysis  therefore  in  did  VanDerhei  not ask    suggests auto-enrollment does indeed achieve that goal.    sponsors perce they are nt.  eligi Addi  themselves, btional le to p  refine a rt will icip  ments likely ate in  be   were em  the pl oye i foc ntroduced ru-ssponsor  of serious  ine  2005 d retirement  deba  to te. evaluate  Moreover,  plans,  the  im  the publ pa  logic ct ic  of poli   of pcy u  modifyi rchasi  consideration ng n glong  (either -  of    15 Choi, Laibson, Madrian, and Metrick (2001) and Choi, Laibson, Madrian, and Metrick (2004).  42 retirement the otherwise, ?  value IRA balances.  of   more ben  these efit  workers   plans s and  by the  saved.  determini  assumption   ng  the tha ta  net mount  housing  of annual  equity  inco  is utili me zed that  “as  this  neede  lumpd  sum ” is  es am timate ount d could  to    Tergesen, Anne.  “401(k) Law Suppresses Saving for Retirement,” Wall Street Journal, July 7, 2011.   64 401(k) participants What do  you abo uthink t the  specific you wo  provisions uld be most  used  likely  in the  to  Sep  do?tember      2011 Gale proposal.  55 Net housing equity is introduced into the model in three different mechanisms (explained below).  term comple  care tely  in  or su marginally rance on retirement ) the incentive  inco me structure  adequacy  of e.m ployees and/or employers for defined  1  EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t EBRI t estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H estimony for H ous ous ous ous ous ous ous ous ous ous ous ous ous ous ous ous ous ous ous ous ous ous ous ous e e e e e e e e e e e e e e e e e e e e e e e e Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co Co mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt mmitt ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W ee on W a a a a a a a a a a a a a a a a a a a a a a a a ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an ys an d d d d d d d d d d d d d d d d d d d d d d d d Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr Means, Apr i i i i i i i i i i i i i i i i i i i i i i i i l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, l 17, 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 201 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 012 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa Pa ge ge ge ge ge ge ge ge ge ge ge ge ge ge ge ge ge ge ge ge ge ge ge ge 1 1 2 1 1 2 2 1 2 1 2 1 1 1 1 7 4 3 6 2 9 5 1 8 6 3 4 4 8 1 2 1 7 2 9 3 0 0 5                            Figure 12 Figure 10 Simulated Impact of Proposal to Modify the Federal Tax Treatment of Average Percentage Reductions in 401(k) Account Balances at Employer and Employee Contributions for 401(k) Plans In Exchange for Social Security NRA* by Imposing 20/20 Limits in 2012, Figur Figur Figur Figur Figur Figure e e eee       2 7 6 134 Figure 5 Figure 8 an 18% Match From the Federal Government for Employees Currently 26–35, by Age and Age-s Figure pecific 9 Salary Quartiles Impact of future years of 401(k) eligibility on 2012 at-risk* ratings for Gen Xers Average account balances for 401(k) participants 55-64 with at least  TM Success Rates of Achieving an 80 Percent Real Replacement Rate From Social Impact of Income and Relative Value of Defined Benefit Accrual EBRI Retirement Readiness Rating (RRR): 2012  Figure 11 16% 2012 by Unconditional Plan Siz Retirement e and Savings Age -Sspecific hortfall*numbers Salary for GenXers Quartiles:  byyears ofMidpoint  future eligibilityEstimates   for participation  70.0% CDFs* 2012Unconditional of the Tw Retirement o Extreme SavingsShortfall* Combinations numbers for TM GenXers of Design  byincomequartile Variables  and  yearsand  of future  EBRI Retirement Readiness Rating (RRR): 2003 vs. 2012  Security and 401(k) Accumulations Combined Under Various Assumptions (Status Quo for Social Security, Housing Equity Used "As Needed")  at Retirement Age on At-Risk* Probabilities thirty years of tenure Salary  in a defined contribution plan  eligibility  for participation in a defined contribution plan  Simulated Impact of Proposal to Modify the Federal Tax Treatment (Status Quo for Social Security, Housing Equity Used "As Needed")  Percentage of population at risk* for inadequate retirement income, by age cohort and income quartile (baseline  Employee Response Assumptions for Employees Currently Ages 25–29 and Percentage of population “at risk” for inadequate retirement income, by age-specific remaining Assumption for this run: Employer increases or decreases to contributions arQuartile e Lowest Quartile 60.7% Percentage of population at risk* for inadequate retirement income, by age cohort (baseline assumptions)  $300,000.00 career income quartiles and income-sp assumptions) ecific defined benefit value quartiles (baseline assumption) of Employer and Employee Contributions for 401(k) Plans In Exchange 14% represented by the midpoint of the range denoted on the AllianceBernstein survey  $120,000 6.0% Assumed 31–40 6.0% Years of 6.0% Eligibility 6.0% , High- vs. 9.0% Low-salary 9.0% Quartiles 9.0% 9.0%  $90,000 60.0% Lowest 90% for an 18% Match From the Federal Government, by Age and 1045 0.0% % 100% 10 Don't Remember 0.0% Don't Remember Remember Remember Don't Remember Don't Remember Remember Remember 2 Age -specific Salary Quartiles: Midpoint estimates 12%40% Auto-Escalation  $80,000  $290,000.00 3 90%  $100,000 90.0% Delta 80% Don't Opt Out Opt Out Don't Opt Out Opt Out Don't Opt Out Opt Out Don't Opt Out Opt Out 90.0% 50.0% 35% Income-specific Assumption for this run: Employer increases or decreases to contributions are  46.5% 1.0% 45.7% 48.5% 47.5% 59.2% 56.4% 63.2% 59.4% Highest Defined Benefit 2.0% 47.5% rep 47.0% resented by th 48.9% e midpoint of48.3%  the range denoted on th 62.1% e AllianceBernstein 60.6% survey 64.2% 62.5% 30% 10%  $70,000 80% Value Quartiles 80.0% 80.0% 41.1% 70% Lowest Quartile  $280,000.00 30% No DB accruals 25% 40  $8 .0% 0,000 12.0% 12.0% 12.0% 12.0% 15.0% 15.0% 15.0% 15.0% 70.0% 70% Lowest Quartile 70.0%  $60,00020% 8% 60% 2 25% Don't Remember Don't Remember Remember Remember Don't Remember Don't Remember Remember Remember 15% 30.6% 60.0% 60% 3  $270,000.00 Auto-Escalation 60.0% 30.0%  $50,000  $60,000 10% Delta Don't Opt Out Opt Out Don't Opt Out Opt Out Don't Opt Out Opt Out Don't Opt Out Opt Out Highest Quartile 6% 50% 20% 1.0% 50.0% 66.7% 61.0% 71.8% 65.1% 70.4% 62.1% 76.6% 66.8% 50% 5% 50.0% 2.0% 70.6% 68.0% 70.6% 75.5% 71.4% 79.2% 74.7% 401(k)  $40,000  $260,000.00 0% Highest Quartile 40.0% 40% 15% 18.2% 20.0% 4% Plan Size Lowest inc40 ome % quartile 2 3 Highest income quartile  $40,000 6.0% 6.0% 6.0% 6.0% 9.0% 9.0% 9.0% 9.0% 40.0% <1M 36.4% 28.8% 22.8% 26.5% 30.0%  $30,000 30% 30% Don't Remember 10% Don't Remember Remember Remember Don't Remember Don't Remember Remember Remember 1 -10M 40.1% 32.4% 26.9% 31.5% 30.0% 2%  $250,000.00 Auto-Escalation 10.0% 20.0% 10 -50M 22.8% 13.7% 7.4% 12.8% Delta  $20,000 Don't Opt Out Opt Out* Don't Opt Out Opt Out Don't Opt Out Opt Out Don't Opt Out Opt Out 20%  $20,000 50 -250M 5% 20.2% 11.4% 3.3% 8.5% 20% 27.4% 1.0% 27.0% 28.6% 28.2% 35.9% 34.1% 39.4% 37.1% 20.0% 10.0% 0% 2.0% 27.9% 27.6% 28.9% 28.6% 38.6% 37.8% 41.0% 39.9% 250 -500M 20.2% 10.4% 3.2% 8.3%  $240,000.00 10% 26-35 36-45 46-55 56-65 Highest Quartile 0.0%  $10, >500 000 M 23.5% 12.2% 6.8% 13.1% 10.0% 0% 10% 0.0% Income Group 12.0% 12.0% 12.0% 12.0%  $- 15.0% 15.0% 15.0% 15.0% 01 26 -35 -91 36 -45 0-19 46 -55 20 56 + -65 Early Boomers Age Late Boomers Gen Xers lowest income quartile 2 3 highest income quartile 0% Source: Author's calculations based on results from EBRI Retirement Security Projection Model Version 1472, and responses to AllianceBernstein (2011) Lowest income quartile 22.2% 24.9% 21.1% 12.7% Future years of 401(k) eligibility  100 105 110 115 120 125 130 135 140 145 Lowest income quartile 86.8% 83.6% 77.7% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 0 0.0% and Employee Don't Remember Benefit R $106,0 esearch 26 Don't Remember Institute and Mathew Remember Greenw $78,al 36 d & Associates 6 Remember , Inc., 2012 R Don't Remember etirement $51,77 C8 onfidence Don't Remember Survey. Remember $28,908 Remember  $- % % % % % % % % % %  $230,000.00 0% Early Boomers Late Boomers Gen Xers 2 Source: EBRI Retirement Security Projection 13.0% Model Version 110627c1. 7.2% 9.9% 13.3% 01-910-19 20+ Note: This simulation models only the financial impact of the expected reduction in 401(k) account balances for employees who are not automatically Auto-Escalation 2 48.0% 46.9% 45.8% 12/31/2010 12/31/2011 estimated 3/31/2012 estimated Lowest 2 3 Highest *An 1-9  Low individual est–inc  is om considered e quart$9 i lto e, 8, be al15 l  at pes 7 -risk si m ini st this ic version 0% 0%  of the 0%  model 0%  if their 0%$  aggr 62% 9,5 egate 06% 9  resources 9% 13%  in 20% retirement 26%  35% are not 44%  sufficient $4 54% 5,80 64%  to1  meet 71%  aggr 77% egate 80%  mini 83% mum 86%  retirement 88% 90% $23 92% ,52293% 94% 95% 95% NB: this simulation only models the financial impact of the expec ted reduction in 401(k) contributions for employees who are no t automatically enrolled by 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 DeltaEBRI 2003 RRR Don't Opt Out Opt Out* 516. .7% 1% Don't Opt Out Opt Out 10.0% 48.5% Don't Opt Out 11.6% Opt Out Don't Opt Out 51.7% 11.4% Opt Out Total $78,407 $54,656 $38,496 $22,786 expenditures defined as a combination of deterministic expenses from the Consumer Expenditure Survey (as a function of income) and some health insurance and Highest-income quartile, all pessimistic 0% 1% 2% 4% 7% 12% 19% 28% 37% 46% 55% 63% 69% 73% 76% 80% 83% 85% 87% 89% 90% 91% 92% 93% 94% 94% 95% imposing 3 average the new limits and does not attempt $255,074.90 29 to assess .3% behavioral m odifications $2 on 72 the part of either the plan ,680.81 26.5% sponsor nor the employ $2ees 92,2 29 assumed 58 .3% .22 to be 10 nonparticipan -19 ts. The $8 simulated 7,212 rates of return are the same $4 as 8, 7 in 04 VanDerhei and Copeland (July 2010). T $22,962his version of the analysis assumes $10,16 no job tur 3 nover, 43.7% ® 1.0% 38.8% 50.1% 43.6% 50.0% 41.1% 58.6% 47.1% 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). Source: EBRI/ERF Retirement Security Projection Model version 110714e. Highest income quartile 10.8% 17.1% 14.1% 8.7% EBRI 2012 RRR 44.3% 43.3% 43.9% eligible for participation in the plan. The simulated rates of return are the same as in VanDerhei and Copeland (July 2010). Th is version of the analysis Lowest-income quartile, all optimistic 0% 0% 0% 0% 0% 2% 5% 6% 7% 9% 11% 14% 17% 21% 25% 31% 37% 43% 50% 57% 64% 69% 73% 77% 78% 81% 83% withdrawals or loan defaults. The full stochastic nature of the model will be included in a future analysis. Plan sponsor and participant reactions to the Highest 2.0% income quartile 49.1% 46.9%12.5% 50.4% 11.2% 57.5% 52.9% 64.0% 16.7% 58.4% * The An  individual resources in or  ret family irement is considered  will consist of to  be Social “at risk” in this version  Security (either status of quo the model  or one of  if the their  spec aggregate ified reform resources  alternatives), in retirement  account balances are not sufficient  from defined to  contribution meet aggregate  plans,minimum  IRAs retirement 20+ $65,700 $36,690 $13,574 $5,430 assumes no job turnover, withdrawals or loan defaults. The fu ll stochastic nature of the model will be included in future anal ysis. Source: EBRI/ERF Retirement Security Projection Model, versions 100810a1 proposal Highest-in are com ex e plained in the tex quartile, all optimistt. icEmploy 0% 1% er increases or 2% 2% -3% 100810a16. decreases to 4% 6% 8% contribution 10% 14% 19% s are represented 24% 30% 36% by 43% the midpoint 50% 55% 60% of the ran 64% 68% ge denoted on 71% 73% 76% the 78% 80% 82% 84% expenditures defined as a combination of deterministic expenses from the Consumer Expenditure Survey (as a function of income) and some health insurance and out-of-pocket 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 Source: Author's calculations based on results from EBRI Retirement Security Projection Model Version 1471, and responses to AllianceBernstein * Normal ret ir ement age. * See VanDerhei (2007) for distribution of opt -out rates from the Retirement Confidence Survey. 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 distribution). This version of the model is constructed to simulate "basic" retirement income adequacy; however, alternative versions of the model allow similar analysis for Sources: 2010 Account Balances: Tabulations from EBRI/ICI Participant-Directed Retirement Plan Data Collection Project; 2011 and (2011) and Employee Benefit Research Institute and Mathew Greenwald & Associates, Inc., 2012 Retirement Confidence Survey. retirement will consist of Social Security (either status quo or one of the specified reform alternatives), account balances from defined contribution plans, IRAs and/or cash balance replacement rates, standard-of-living and other thresholds. Sources: Sources:  EBRI  Retirement EBRI Retirement  Security  Security  Projection  Projection  Model®  Moversions del™ versions  1501  and 1501  1502.  and  1502.  Note: This simulation models only the financial impact of the expected reduction in 401(k) account balances for employees who are not automatically *The Retirement *The 2012 Retirement  Account Savings  Balances: Savings Shortfalls Shortfalls  EBRI (RSS)  estimates. are (RSS) determined are  The determined  analysis asa present isa  based sa present value  on allof value  participants retirement ofretirement  with deficits  account deficits at age  balances at 65. age65.S  at the end of 2010 and plans, annuities from defined benefit plans (unless the lump-sum distribution scenario is chosen), and Combin(in some cases) net ed Real Replacement hous ing equity Rate (either in the form of an annuity or as a Source: EBRI Retirement Security Projection Model,® Version 120201. * 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 Sources: ources: positive EBRIERetirement BRIvalues Retirement  forSecurity  theS  sum ecurity Projection  of employee Projection Model®versions  and Model®versions  employer 1501 contributions 1501 and 1502. and1502.  for that year. 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, standard-of-living, and other ad hoc thresholds. Source: EBRI/ERF Retirement Security Projection Model, versions 100810a1–100810a16. turnover, withdrawals or loan defaults. The full stochastic nature of the model will be included in a future analysis. Results for participants currently * Cumulative distribution functions. 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  Reduction in 401(k) Balance Average Percentage Reductions in Inadequate Retirement Income 401(k) Account Balances at Social Average Percentage Reductions in Security Normal Retirement Age 401(k) Account Balances at Social Security Normal Retirement Age

Volume T-172

Pages 37

EBRI Testimony

April 17, 2012

Jack VanDerhei

Financial Wellbeing Retirement