Figure 7 (updated) Number of Workers Working for an Employer Who Does NOT Sponsor an Employment-Based Retirement Plan and Number of Workers NOT Participating in an Employment-Based Retirement Plan, by Various Demographic 49 21 ____                                          contrib                                                                                   if ____ A                                          plan not Figure A the    employers new similar   Committee at  sponsors _. _.    all   1 subroutine u“The The Expected Impa  tion sho   occurrenc relevant  w   Im were plans  swould   pthe with act    to was  to  at eme  of     gu   the tak react drop this                                                                                                   dian PPA added ide  potential an  stime    their  ct of Automatic Escalation place to  on real ce   this to    on Ret needs    replacement plan  the when   wha ne i rement for                                                                          mod  w  ma to  tdefin    lthe some e  tches be gislation, e l In   perc e to thoroughl                                                                                                                                                                     cd   rates of  o   allow and contri me  ent the    EB  401(k) for aat   lige simulations bution RI  kely y of  age 401(k)    of exami published 401(k) Con     parti those “costs” 67  plans   from n Participants.” ced.  ipa  sof t    to will   ating  a  401(k) n                             EBRI various  tstudy  ts ribution provide   be                                                                                                                                             would   they s  in  t balan udies in   styles  ter sEBRI    20  wo ains on Retire  m significan 07 culd t  es sIssu   ead  of have of   exclusively  re that  target reduced  ehave duce   Brief, ment docu  tshowed                                                                                                                          portion  - only dat th  me In no.  eretir  come (9th efor   nted amount  the  funds 318   how  of e ment 18    atha T  -   for 170  t                     ? Within each of the four age cohorts, theFigure  highest 1-income quartile experiences the largest  3.1 Social Security 1 5 9 7 8 11 Appendix Endnotes Appendix Introduction Fut References ure work A: B: A Number Brief Chronology of Workers of Without RSPM a Plan and Employer Characteristics, 2010 68 21 36 28  60 Median Real Replacement Rates at Age 67 From 401(k) Balances (Employee ed., vol. 28, p  Bepn. 1-8). EBRI Note efit Research Insts, September 2 itute, June 2008 007. ).    signific VanDer VanDer The elective d ant con he hei (S i an sequ eptemb d Cop eferral lim enc ees on la er 20 nd (Ju it incr 07 empl ). ne 20 ease oyee co d 11). by $5,5 ntributi 00 for 201 on behav 1 an ior. VanDerh d is currentl ®  ey i l and C imited to $16, opeland 5 (200 00 b 1b) estima y Internal ted a Revenue participa a worker’s benefits they perce defined automatic  comparison  are nt   con  or for  n saving pre ts   es 30  t  those currently -rcalation retirement  i  with percent bution  or  currently stop  parti   plans would   ages match  saving c  income. ipant   (and 25   in mak  provided  -  -altoge the dire 29  the e       by For the 401(k) cte  IR  th in   d example, A AE  er com by   in rollovers    is system.  results ve the  edisplaye st  quar  m govern  some e  eve  nt  they le sd  nin ment. s.   by  analysts  more  produce) 2009.  The the    favorable  amo values    will In  are unt   Ap merely   vary th  ri they  leu  2010,  nder compo  from   currently report  a  th  a  numb n  elow ent   the model    have of of  eaverage r   53 retir of  was  in   percent dif  esavings  fment completely  erent balance   for   and   the in       average percentage reduction from the 20/20 caps.  This reaches a maximum value of 15.1  The importance of Social Security retirement benefits for today’s workers is shown in October 2010  Choi,  The                                          Chairman In addi  original  James tion  Baucus,  to   vJ.e  ,the  rsion David  expansion  Ranking  of  Laibson,                      Reti remen Me  of  Brigitte mber  the  t  Se mod  Hatch, cur  C.e  iMadrian, ty l used   mem Projection  for b ers and  th eof   Mo  Andrew two  the del  analyses  commi (RSP  MetM) ttee, r ick, above  was   “Saving I am  used  to  Ja  inc   clude to For k VanDerhei,  analyze  Ret  401(k) ireme  th  plans e  nt research  future  on  with th e    for Participants Currently Ages 25–29, by Income Quartile   29  Cod behavior e Sectio al mo n 402( del that is a g)(1). A plan ma ble to contro y p l for the tende ermit participnc ants y of empl who aro e a yers to subs ge 50 or over at the e titute betwend of the cal en the amount endar they year to match 1 9 Working for an NOT investments, defined security reparameterized lowest scenarios -income   perce con tha  fortt r   appears both not inbution  tquartile  for   wi including  plan  tthe h   plans to  401(k)    spo to hi  be ghest  a    the (mo ng  high sor e  plan nerat  value -s income tand   of comm  desig  i77 ng   wor of   perc the   ntheir only quartil k  parameters er  most ent   be  the prie  havior. for TM m    for non 401(k) ary  the  those -  Social residence for   hig  subset  sponsors  hcurrently est  Securi - in of  or com  this ty   the that   ages retir e  universe   value quar  have  e36 ment t–   ile. of 4adopted 5)  defin   and and wealth  Th   attempt fall e e d  asimulated   usfor ben  tom to  Baby  e8.6  afito tic t   perce plans. assess    Boomers ratesn     tof       Path of Least Resistance,” originally prepared for Tax Policy and the Economy 2001, updated  Senate 22 69 VanDer  HELP hei an  testimony d Copeland (2  by 010). EBRI:  91 percent of the lowest-income households would be at risk of  automatic economic director of   well  enrollment, the-  Employee being of  EBR  the  Benefit I  retired plans  Resear  to  pop  conuch tlation inue  Inst  to  at itu   cond the te.   state EBRI uct research   islevel.  a non  Thp  in eartisan   Employee this area  insti  as  tBenefit ute  publ  that ic  Resea policy  has been r ch agendas       ____ _____. _.  “The “How  EBRI  Would  Re tirem Targeetnt -Date  Readi  Funnds ess  LiRatin kely gIm : p aRet ct iFuture rement  401(k)  Income  Contribu  Preparation tions.”  and  Testimony  Future  before  VanDer An emplo hei ym an ent-base d Copela d retire nd (2008). ment pla n can be sponsored by an employer or by a union. “Employer sponsored” is 90% 4. Figure make ad per do 2.1 llar of e  4ditio  Results also nm a  provides lplo electiv yfr eeom contri e deferr  the 61the butio  average 20 al contri n an 11 Retirem d the maximum p  butio percentage ns. Effectiv ent  redu Confidence e e for ye rcctions entag ars be e of com  in  401 Survey gin(k) n pens in g on or after Ja account ation the  bal y are ances nuar willin  y 1, 20 at g to match. EBRI  Soci 06, al if a plan Security  Employer NOT Participating An important policy topic resulting from an analysis of employment-based retirement plan participation  draft: July 19, 2004; and “For Better or For Worse: Default Effects and 401(k) Savings Behavior,”  the and enrollment return There   value Gen   isare   aXers   of  significant e  provisions.  xthese plained .  However,  plans   incre in mo    by  A the a re  se completely determini   potential  de in tail the  in  self n  V   gin up a  -the reported nDerhei cdated re aase m ount  of version  and   propensity at -of  risk Copeland  annual  of  percenta  the  to   nation inco   (2 re010), duce ges me al   resulti  that   model bu savings t  they this n gwas    for from lu  are m   produce those p  based   (1 sum ) em  in  a  on  dm pl the  for ount  oye a  lowest  the r  could  May      for the highest-income quartile for those currently ages 56?65.  inadequate retirement income if they had no Social Security retirement benefits, compared with 76  the Prospects.”  joint DOL/SEC  EBRI Issu  Hearing, e Brief,  Target  no. 344  Date  (Employee  Fund Publi  Benefit c Hearing,  Research  June  Institu  2009.t  e, July 2010).  dicta condu Institute te. ctin    Whil an g do re the iginal  Gale  Mil   resear (2011 bank Memo )ch  incl  onudes  retirement rial  aFu  dind, stribution  working  and health  of  with  fe dbe  ethe rnefits al  gove tax  for ch rnor ange  the  of  past  by  Orego  cash  33 nye  income , set ars.  out  EBRI  per  to  do  csee ees n tile if  not  this  in  take      used in this study for brevity, but it should be understood that it also mean 22 s union. 2 ado inten pts a Roth feature, emp ds to update that resear loch and attem yees can desi pgn t to appl ate some or al y it for more formal ev l of their electi alu vea contrib tion of this ution prop s as desi osal ign n the ated future. Roth Characteristic(s) Sponsoring a Plan In a Plan normal In 23 In A brief chrono   recent 2008,  retirement  EB  yea RIr  included ls, og  proposals y of RSPM i  age  all by  the  permane have s inclu  new Se  surfaced ded i n n  PPA nate tly n  A provisions  modifying    ppe to Com  reform ndix A. mitte   in th  the  ae  study exclusion  401(k) e on n  that system  Financ  of  co employe mpared  based ee e  on   pote contribu  thential  assumption tions  accumulations  for  that    62 Pension Research Council Working Paper, PRC WP 2002-2 (Philadelphia, PA: Pension Research  is the  80nu %mber of workers who are not participants, as50  well as the number for those that work for an  VanDerhei (April 2010). modifications stochastic savings 2010 be conv  EBRI  categ erted   pr poocess licy o   to to ries.   at  ex forum   with   retirement isFor tin   gexam aand   pl mean ans,  used p lage.  e  and e,  qui of in      Of th (2) the tye    return course, a   July full  substantial - ti2010  me of  this  8.9  workers  Issu  co  perce  portion ncep e Brief.  nwho tt  does   and of   are low   anot   mean currently -income  adjust  fix e  hous for  saving d-income  the eholds   fact for  return retirement   that decre  the  of asing   6.3 vast  who   or       ? The finding that the highest-income quartile within each age cohort experiences the largest  percent at risk with current Social Security benefits.  The other three higher-income quartiles also  policy situation 2011,  itpositio  would  coulnd s  be be  and   extreme address  does not elyd  useful   for lobby.  Ore  to      g on. expand  The  analysis this analysis  focused  to incl  pri ude m arily both  on em  simulated ployee and  retirement  employer   (millions) contributions (which are included in gross income), rather than traditional, pre-tax elective contributions. 70 69 higher-inCouncil, come indivi  The duals Wharton  receive  School,  more  Universit  tax-relate y ofd  Pennsylv benefits from ania,  Novem these progra ber 9,ms  2001).  than    do individuals in  ____ ____ retirement under_. _.  AE   “The “The  and  savings   Im Im  VE p paa  for ct ct  plans   of of  several   Automatic Def  freom rring  different  taxabl  Retir  Enroll e  eage  income ment ment  groups.  Age  in  in  401(k)   2012 on  Aga  Retire i  n, (by Plans  we  age men   found on  and t  Fu In ct  age certain u ore m-e  specific Re  Adequacy.” tire  (high ment  sa -income la rAccumulations: y  EBRI  quartile )  groups Issues)  Brief,  bu  that t,  A     no.  45 This includes the 78.2 million who worked for employer/union that did not sponsor a plan plus 15.0 million who employer/uni 3 on who does not sponsor a plan.   Copeland (2010) investigates these numbers to show  perce report majority eliminati n  that t (expressed  nofg   4 th future 01(k) ey curr   parti c oinently nt  nom ribu cip ants have itnal ion   ter are s  less  tom   year savings  s). tha  s, n  $1,000, if  plans not deca   as 71.3  ades,  reaction  percent  away   to from indicate  the  retir  ex  they cleusion ment  wo   of uld age.  employee  re   Mo duce reo  th  veer,  amount  even if   VanDeraverage hei and Cop  percentage eland (Jul yreduction 2010).  is no surprise, given the increased likelihood that workers in this  benefit from Social Security: Comparing the at-risk percentages with and without Social Security  Total 24 This improvement took place despite the financial and real estate market cris 77.6 is of 200 91.9 8/2009. For evidence on the reactions wealth with  to  athe  comparison  proposal an  tod  ad  to  hoc simulate  thresholds  the e m for ploy  retirement ees over  ti eme xpe nditur in an attempt es, but the  to  project results  mad futuree  it     70%   VanDerhei and Lucas (2010). Simulation 358 (Employe  Study e Benefit  Based  Research  on Plan Design  Institu  Mo te, Jun difica e 2011). tions of    Large Plan Sponsors.” EBRI Issue Brief,  lower unlike The were 37  new  likely   margin Fig  mod ure  to  al 3,  edo l   tax assumes was  be  brack tter  used  un  eth  to ts der at   (as analyze  plan  VE  well  than  sponsors   as how   AE those  ,eligi  bu  complet  twho b  overall, ility  ma  for eyl  y  participa the pay  drop  AE  no  their  results  fetidoe  nplan r ain  ldomina   income a  match  definted e   ta and d .xes con    tha  in tribu  ta  e  parti m tion ploc  plan yular ees   year). are     worked for an employer that sponsored a plan but did not participate in the plan for whatever reason. where My T hese findi testimony  potential ngs  today are part of the 21st ann  legislation  will foc  may us on  exclude  retir uael Re ment  workers, tirement Co  security  or the nfid  and  ence Surv nu  th meb  po er tential of ey  (workers RCS), a surve  impact  who  of  are  yvarious that gaug  already  types es the vie  being  of tax   w s contrib one saved. impact of the recessi  does  This utions  look  va lue for  at   the retirement de on o clines  average n retirement i  to  savings 38.8  balan  perc nc com es  pla en  for en ade ts   for workers from qu  those ac  ta y,xable se  near  with e VanD   income, retirement  savings erhei (F   needs ofebru   $1,000 age,  ato r  it y   be 20  isto  11).  obviously  ana less ly  than zed   carefully not $10,  correct 000.  when    to   Copeland, cohort  Craig.  ei  “Employ ther currently ment-  e Based xceed  Retire  the $20,000 ment Plan  (ind  Participa exed) limit tio n: when   Geo  their grap hco icnt  Differences ributions are  and     retirement benefits, 24-26 percent of households in the other three higher income groups are saved  balances clear that  at  ma  rejor tir edecisions ment age.  lie    ahead if the state’s population is to have adequate resources in retirement.  46 no. 341 (Employee Benefit Research Institute, April 63 2010).  See endnote 17 of VanDerhei and Copeland (July 2010) for more detail. Some of these proposals have included modifications of the current federal income taxation treatment  left impacts 25  with 60%  retir onlye  the ment  governm  incomee  adequacy nt match of  in  30 Se pte perce mber nt.  2010.   It was also 15 used to compute Retirement  Self-Employed (Not Wage and Salary) and attitu 4.1 20/ des o 20 f w analysis orking-age and retir ed Americans regarding retirem 9.0ent, their prepar9.2 ations for retirement, their reform reached,  optio  by certain ns on retirement  demographic  inc oan me d  em adequa ployecy. r characteristics,    This draws on  an  the nual  ex te earni nsive ngs  research , employer  con  size, ducted  and  by     Trends, 2009, EBRI Issue Brief, no. 348 (Employee Benefit Research Institute, October 2010).   considering VanDerhei, look See Hol  onlyd  at e n   Ja the the an ck, d VanD   overall 401(k)  Sarah erh    bal Hol imp ei (2 adance en, c 005). t of   with and   such   Luis th eproposals.   employe Alonso. “401(k)   e’s current  Plan  employer.  Asset Allocation,       Account Balances, and  combined with employer contributionsH  or earing  are predi  on:  cted to do so in the future.  However, for  from at-risk status by Social Security.  4 36 By 2009, many of the 401(k) sponsors who previously had VE plans 64  had shifted to AE plans and EBRI was  Net Wage and Salary that The post-W  excludes orl d W some ar II demograp  or all  of the hic  con wavetribu of childre tions n born employees betwe68.6  en 1 make 948  to –19  tax 64. -qualified 82.7  defined contribution  Savings 4. confid 2.2enc  Shortfalls Impact e with regar of  for d to vari the  Boo prmop ous asp ers osal  and e on cts of retireme  Ge 4n01(k)  Xers in bal  Oc nt, atob and ncees related iss r  2010. at retireme    ues. The surve nt y was conducted in January In 47  December 2010, the National Commission on Fiscal Responsibility and Reform released their long- EBRI While work Subsequ ____ _. on  status  it  “   Loan these is R eetir nt  possible   (f to  ement Activi ull  topics  the -time/part   ty release to  In  over   in analytically co  2008.” me  the  -of time  Ade   the last  EB ).quacy     Orego  evaluate R12 He I Issue  yea  shows  for nr   study, s  Brief, Today’s   the with  that  c    no. hit it  ange in s   was  Workers:  Ret  335 2009,   ideci in remen  (Employee  incentive  78.2 ded  How t   Se mill tha  cu Certain, st  ion Benefit   rity for the  workers    an Projectio approach   How empl  Researc   worked Mu oyer n  coul  Mo ch h  to In  W del® dst   sponsor for  ill be it ut It    an as carried  eCost,  , well  Oc  a t  and  o  as to ber       VanDer 50%each hei (S  age eptemb  coheor 20 rt other 10) also dem  than th onst e oldes rates that elig t one, the ibilit  low y for a est-defin income ed contrib  quartile utio  has n retireme  the secon nt plan d-hig has a hest  26 Given able to  that  track  most  the  401(k) changes  pla  inn  plan sponsors  provisions  currently  for  hundr matcheds  at  aof  ra  th tee  gr laregaest ter  401(k) than 30  pl  perce ans. This nt,st  it  information  is not    Under 21 Years Old Copeland, Craig, and Jack VanDerhei. “The Declining Role of 6.2  Private Defined7.6  Benefit Pension Plans:  Who  See VanDerhei and Copeland (2008). 2011 through 20-minute telephone interviews with 1,258 individuals (1,004 workers and 254 retirees) age 25 and plans. In For  an  purposes  atte   Vamp nDet  of r  to hei  the  provide  (Mar  today’s ch,  m  2011) e  analysis, aningful  provid   statistics thees  (fil  antered  analysis  on )the  RCS   401(k) of  responden  two  system, new tqu s are  est EBR  ipla on I en sc ed from tere  into d  the  into  one  21  a  of  collaborative  wa  thve ree  of    the   How 2009).  Does    Eligibility “Tax  for  Refo  Participa rrm Options: tion in  P Paro  Defined moting  Contribu  Reetirement tion  Se Plan ccuri  Hel ty”p  ?” EBRI Notes, no. 9  35 awaited 5Unfortun  do atel cuy, ment  the  latest on federal  projections  debt reduc  of Social tion,  Se “The curi  Moment ty suggest  of  that  Trut  trus h.”  tAlt  fund hough  rese  th rveir es  guidi will be ng  exhausted principles   qualified other employer/uni annual  states  analys  plan  as  on isand/or   well. of  that  tens  Kans   di con  of d not t  as millions inue  and  sponsor  a  Massachusetts match  of in  adividual   retirement at the current  401  wer  pl (k) ean  chosen   partici level  and  (if  93 p  as  ants at .2  the   all) million  dating   nex as at   workers result states  back  in  of for    some did 20/20  analysis.  not  case  or  participa  the  sResul  as  Gale  far ttse   as  of  in    a  significant positive impact on reducing the additio 23 nal compensation most families need to achieve the desired level of Americans b average orn bet  pewrce een 1 nta9g6e5–1  reductions. 974 and c   uAlthough rrently betw  this een  the ag may be es of 36–4  due to several 5).  considerations,  it is  65 Year Old or Older 3.7 4.5 surprising In was  Oc  used tobIse  rin  Affected, that  41 20  an10   the April  testi  average  an  2010 mdony  How EBRI   before redu .”  In Issue ctions  Robert  th 37 eBrief   Sen inc  L.rae  Clark tato ese  He  show,  in  al and  Ftihgure   ,Olivia on  Edu c e4.c  ag ation,    Mitchell, For 38ain  those ,  La the bor   eds., signifi  currently  and  Reori cant  Pensi  e  ag im nting oens psa   26–35, ct Committee,  Ret  of imo rem  the ving ent  average    on Ri tosk   AE      older in the United States. Random digit dialing was used to obtain a representative cross section of the U.S. 40%(Employee 70  Benefit Research Institute, September 2010): 13-20.  51 categories: Retirement Co   nfidence Survey (RCS)  showing how workers  would likely react if they were no longer  effort with the Investment Company Institute (ICI) in 1996 known as the EBRI/ICI Participant-Directed  27 and in 2036.  values   EBR  (pIages  recently  13–14)  provided  specifi canalysis ally me ntion of a gen  thee ric nee  type d to  of keep  Socia  America l Security  sound  refo  over rm proposal  the long  that,  run  by in   proposal, 1996. the plan retireme  Kansas  (Figur   nt inco   it e  study   7) would me ade .     were Focusing  be qu  hel ac  pre yp . s  ful in ented   on (if  no em  totp   es the loye sential)  state’s es who  to  Long   did supplement  -not Term  work  Care   th for  iSe s  an with rvices  employer  a  de Task tai  Forc l ed tha  set et  sponsored on  of  July  surveys  11,  a2002,  and/or  plan,     Not Full-Time, Full-Year The full stochastic nature of the model will be included in future ana 32.1lysis. 39.5 Man almost agem  always ent (O  a result xford  Univ of their ersity  current  Press  or for  expe  the cte Pension d futu  Researc re contribu h Cotuncil, ions ex  2010): ceedi n122 g 20 -136.  perc  ent of  VanDerhei, Jack, and Lori Lucas. “The Impact of Auto-enrollment and Automatic Contribution Escalation  reduction “The pop 6 plans ulati  Wobbly  (fo on. T r  tho varies o  further incre sStool: e  currently from  Re  atire  lo as  w ag ment e repres  of es  24.6 25–29,  (In) entati  security perc  the on, a cell ph en  differe t  in for  America,”  the nce on  h e supp  iin ghest  th  the e leme -  in median  mod cnt ome weal  s  16 quartile was accumula adde  used d to the sampl  to  tions to  a  analyze high  would  of e. Starting   36.0 the  be  relative  percent with the    for  52 Retire allowed These results men  tot  de  Plan duct assume retir  Data  retirement  Colle ement at age 6 ction  savings  Project.  pl 5.an F   As  ocontri r evidence  of December butions on the impact  from  31, 2009,  taxaof deferrin ble  the  income.  data g reti base   rement ag  include e be d statistical yond that a   ge imple essence, menting  would  “policies  keep Social  today  Security  to ensu  retir re that ement  future  ben  gen efitse in rat  their ions  have current  retirement  statutory  security,  form un  affordable til 2037, and     Full-time, part-year 11.4 14.0 focus 9.2 and  million  the  30 group %  results  we  studies re  of  self  th  to -ee  Massachusetts  m incr ployed—meani ease the un  study dng erstanding  the  were  worker  pre  ofs  the ented  could  employe   on have  Dec.  started r’s  1,  likely  2002.  a  reaction pla  nWi  for th   hi to  the mself/herself  these  assistance  changes   of  the  65 _____. “on Retir  Reement tireme  Savings nt Income  Shortfalls  Adequacy.”  for Today’s  EBRI Issu  Workers.” e Brief, no.  EBRI  349  Notes,  (Employee  no. 10 Benefit  (Employee  Research  Benef  it  48 compensation when combined with employer contributions. Phrased another way, the 20/20  2001 wave of the RCS, all data are weighted by age, sex, and education to reflect the actual proportions in the adult the importance 28 approximatel  lowes ? Stop t-income  of  saving y  employer 2.39  quartil   ti for mes  retirement -provid  e. fin  al salary ed  retir altogether.  ine  an ment  AE  ben  plan   e  relative fits and  Social to a VE  Securi  plan).ty.      See VanDerhei (August 2011) for evidence of the importance of participating in a defined benefit plan. 65 see VanDerhei and Copeland (2011).. 10 It is important to note that the annuitized accumulations in this analysis are from 401(k) contributions exclusively Gale, William G. “A Proposal to Restructure Retirement Saving Incentives in a Weak Economy with Long- information Part-time, full-year  about:   10.0 12.3 at health  that  care, point  an  subject d fina ncial all So  frcial eedom,”  Securi ty th re e docume tiremennt t be  pu nefits ts forth  to  aa  pe taxrmanent  reform  plan 24 perc  that ent  would  reduction.  modify  As  without before Kansas  attempting Insurance  Institu Research the need te,    November D Institute, for  to epartm   ac quan tionet  Oc ify  nt, from  2010).  tthe  oEBRI b ehis/he  potential r   2010a): wasr  able  employer.   changes 2  - to9.  crea      tTherefore, in e  Re retir tirem ement e  the nt  income Readiness  number  resulting  of  Ra  workers tings  from  based  who  such   on worked  a  a  massive  full  for    population. Dat cap woul a for d, w  as aves of the RCS co  expected, most nd  affect ucted bef  theore  high 20est 01 hav -income e been  workers, weighted to a  but it llo  also w fo r consiste would cause nt  a  September 15, 20111  Although analysis based on financial economics suggests that higher-income employees would be the  ? 20%Reduce the amount you save.  and d Part-time, part-year o not inc Term lude pr  Deficits oject ,” ed mimeo, Social  Securit Septem y retirement be ber 8, 2011  nefits. T 10.6 his is in contrast to other EBRI resear 13.2 30 ch (e.g., 49 2 Why is it importan 24 t to promote retirement security? 7 retirement expected, the  plans  impa  byct  cappin  should g  annual be minimal  “tax-  for preferred  those currently  contribu tions on the  to  ve  [the] rge  lower of retirement—so  of $20,000   the or 20%  “at- risk” of    someone shift stochastic comparis VanDer  in incentives. ons; c hei  els  de an ecum onse  tha d Cop utq   lation di ue entl la d nd (2 not y , some data i model  sponsor 002a).  that  an   to pl thok an e 201   into totaled 1 RCS m  accou  69n.0 aty   million the differ slig  house  inhtl  2h y009. old’s with   data p longev ubity lishe  risk, d in pr  post evi -re ous tirw em aves of ent  Figure In Later  February  in  5  2010, provi  2011,  dEBRI es  th the  and e  same mod  DCIIA e  analysis l was  tea  used m  as ed   Fi to  up gure  analyze  to  do 4 bu  an  the t  this analysis  impa  time ct  that   th ofe  th   18 focused e  2008/9 perce  not nt  crisi  governmen  ons  ain  comparison  the fit nancial match  of   isand  VE     an real d   Household deficits for married couples are divid42 ed equally between the two spouses. significant reduction in retirement accumulations for the lowest-income workers.  Less than $5,000 in annual earnings most VanDerhe  likely i an  to d  Lucas, Novem be negativelyber 20  affected 10) that i  by an  proposal cludes both  tocompo  cut or nents. Ho 9.8  eliminate we  the ver, in the previ  deductibil 11.8 ity ous an  of 40 alysis, all 1(k)  ____ ? ?_.  Testi Continue 20.7m  million 1ony  to  before   4 save 01(k)   what the  plan  Senate  you  participants,  do  He  now. alth, Edu   in     cation, Labor and 2  Pensions Committee, on “The Wobbly  66 In 2010, EBRI  updated its Retirement Security Projection Model  and determined that the overall  10% level the RCS. Data income”  for  Early (page prese  Boomers  31).n  Th ted i is n  increases is tables i  oftenn  referred this rep  by only o   to rt ma 0.3  as  perc y  the not total to 100 entag  “20/20   e points.  cap.” due t   Bu o roun t Late din  Boomers g and/or missing cat  will have egor a larger ies. In   investment risk, and exposure to potentially catastrophic nursing home and home health care risks. This  Gale, modeled estate AE, bu  Willia  tmarke  rather .  As m t  expe G., s  how  on  Jonat  cte retirement  tod  im h, an this pr ove Gruber,  increases  in  plan come  and  desig   the a dPeter equa n  average  and  R. cy.  worker  Orszag.    perc    en education   2006. tage  re “Improving du  toctions  optimi  eve  Op ze nportunities th  more. e results   For   und an  thdose er  In cAE  en  plans tives   Less than $10,000 in annual earnings 17.2 21.0 50 workers wStool: ere si mulat Retire ed an ment d jo  (In) b chan security ge wa  in s allo  America” wed.  (T-166). Oct. 7, 2010b.  contrib 8 utions (at least to the point they are constrained with respect to the annual funds available to  VanDer ? 51,852 hei and Cop  empl eoyer land (2 -sponsor 001). ed 401(k) plans, holding   The Of VanDer  those  potential he  69.0 i (October 20   reaction million, 10 6.  of7 a).   emillion mployee’s  wer eno  under t currently  the age  pa  of rticipa  21, and ting  3. in6  401(k)  million  plans  were  will  age  be  65  extre  or older. mely    difficult  theor retirement y, the w  income eighted sam  adequa ple ofcy 1,258  for househ yields a statistical pr olds currently ecision  ages of plus o  36–62 rhad minus 3 perc  substant entag ially e po improved ints (wit  since h 95   percentage for  Savings of their  by expected  Middle -  Social and Lowe  Security -Inco  ben mee  House fits reduced holds.”  as     The a result  Hamilton  of this  Project,  change,  Discussion  and their   “at Paper -   Less than 100 employess 4.2 Replacing the current deduction for contribu 39.0 tions to 42.2 defined was followed by the expansion of RSPM, as well as the Retirement Readiness Ratings produced by it, to  currently with automatic  ages 26–35,  escala tion the  average of39  contrib  reduction utions. While  varies  it  fr  isom  difficult  a low  to of  det 30.6 erm  perinceent  the  for  correct  the high  “target” est-inco  forme     The    respective probabilities are comp uted for each family income category and the worker’s reaction to  0%  behavioral economics has shown that the reaction of employees in  contribute to a 401(k) plan), 67 3 4 5 Even 29 ?  if one $1.210  wer trillio e to ignore n in assets.  the potential    interaction of the proposed limitations with the present values  Submittted Testimonyy by   to Approximatel perce  model nt certain  for new ty y )33 of   incentiv million what the  were eresult  structures  not s wo  full uld b  similar -time, e if all Americ  fu  toll -those yeaarn   workers, s age 2 proposed 5 an  and  by d old   18.5 Gale. er w  million    e re sur For ex  vhad eam yed  pan lw enual ,it hdoes co  e mplete  athe rnings  current  of    An 51  April 2006 2011- 02. article   Th Lowes  eintroduced  Brookings t   aInstitution.  new method 2    of analyzing the re 3sults from the RSPM Hi.ghe stInstead of  ____ 2003. T Fewer than 10 employees hese estimat _.    “A Eve  Po n st with - eCrisis s compar  these  Assessment  improvements e quite favorab  of Re ly tireme to however, those in nt Inco  almost Holdme en  and Va  Ad one 18.3 equa -half nD cy  erh of  for  Baby ei (20  Baby  02) Boo 19.0  Boomers wm hen ers the differenc  and  and  Ge  Ge n nXers e bet  Xers.” w  were e en   9 risk” level increases by 1.6 percentage points under the baseline assumptions. Gen Xers will have even  a national model and the presentation of the first micro-simulation retirement income adequacy model  quartile retirement VanDer  to hei   asavings, an  high d Cop  of  ewe 41.4 land (Ju  use  perc dl ywhat, 2 ent 00 2for  ). by  the  most  lowest  financial -income  plannin  quartile g standards, .   appears to be quite generous:  the proposal contrib  is added ution  as plan  a stochastic s with  response a flat - rate based gove  on thernment  model descri matc bedh  abov e for the 20/20  VanDerhei (October 2010b). situations similar to this are often at odds with what would have been predicted by an objective  Income Quartile accuracy. There are other possible sources of error in all surveys, however, that may be more serious than of  10–49 employees accruals under defined benefit plans and/or the existing15.1  tax preferences  16.6 available to some individual  less experie  than nEBRI ce  $10  u nder Issue ,000.  401    Brief, Furt (k)h   no. eplan rm  354 or s allow e ,(Employee  many  researchers  of these  Benefit  workers  to  Research extrapolate  (39.4  Institu  milli  to this on) te,   worked pop Febru ulati ary  for on  2011).   employers with 53   respect  with  to:  less  simply nominal a  computing nd real rep  an lacem  overall ent rates are cons  percentageider  of th ed. Ho e simulated wever, this is to be expecte  life paths in a particul d givenar the assumptio  cohort that nwill s listed  not  determined to be at risk of not having sufficient retirement income to cover even basic expenses and  more years of their expected retirement affected by this change, and their increase in “at-risk”  built in part from administrative 401(k) data at the EBRI December 2003 policy forum.   The basic model  Gale an 80  (2011)  percen  utpdates  REAL inco  previous me repla  analysis cement  by  rate Gale,  in  Gru  retbire erm  and ent  Orszag  when  401(k) (2006)  accumulations  and analyzes a  are  plan  combined  that would     analysis. Since the    Tw inception o assumptions  of the project,  need to  average  be utilized  balan  before ces have  looking  bee nat  di  the splayed  likely  as  chan  a function ge in 401(k)  of bo  balan th the ces    as a  Holden, Sarah, and Jack VanDerhei. “Can Jack 401(k)  Vaa  AnDe ccumulations rhei, Ph.D.  Generate    Significant Income for Future  52 concerned simply with optimizing a financial strategy. In an attempt to better understand potential  theoretic 10 49–99 employees al calculations of samplin 6 g error. These include refusals to be int 5.6 erviewed an 6.6 d other forms of nonresponse, retirement abov VanDer e (esp hei ecia  account an lld Cop y the lack of j  e(IlaRA nd (D ) contributions, ob turnov ecember 20 er an 02) d therefore  this .  alternative the suppr  formulatio ession of casho n of capping uts prior  tax to retirement). -preferred   than 100  Source: EBR employees, I Retirement Security  including Projection  10.2 Model  mi Version llion110503c.  working for employers with 25–99 employees, 10.4 million  VanDerhei and Copeland (2010). uninsured have At this  sufficient  point  he  italth   may retirement  care  also  costs.  be  income  useful   The  to  results to  ana  paylyze  ,for  not   the the  surprisingly,   pote simulated ntial im   were expenses, pact  ev  ofen  the   the worse  proposal  new  for  method  low  on- income 401(k)  com putes retirement  households,  what     percentage is simulated to be 5.8 percentage points.   was thenT  he simulated modified rates  for of return  Sen for the atebaseline  Aging return  tes scenario timony are the same as in VanD  in 2004 to erhei qua and C ntif opeland y the (July  be 2010). T nefici his val ers  impa ion of the analy ct of s  is a models  mandatory  with Social Retirees?”  Security  EB . Th RIe  I ssu study e Brief,  found  no.  that  251  with  (Employee  the proper  Benefit  choi  Researc ce of plan h In  design stitute,  and  Nove  worker mber  e  2002 ducation, ).     replace     result        .   of“Re  the  the t irement ex  proposal: isting  tax Inc    ode me ductions  Adequacy:  with  Alternative  a flat-rate  re Thfundable resholds  cre andd it th  that e Im p serves ortan cas e  of a matchi  Futureng  Eli  gibility in  participant’s age and tenure with the current employer to allow more meaningful assessment of the  the effects of question wording and question order, and screening. While attempts are made to minimize these ? Initial participation decisions.  employee behavior with respect to a proposed elimination of deductions for 401(k) contributions, this  401(k) participants who are not automatically enrolled and assumes no job turnover, withdrawals or loan defaults. The full stochastic nature of the model will contributions would substantially reduce the current limits available under qualified defined  for those with 10–24 employees, and 18.8 million 54  for those with fewer than 10 employees  balances percentage  eve  ofn  the if there  househ  werolds e no  will  em  meet ployee  that  behavioral  requirement  response  more  to  than  the  achange  specifie  ind  the  per  exclusion centage of  of  times    in  as 30 53  70 percent of households in the lowest one-third when ranked by preretirement income were  be included in a future analysis. Rese earch Directorr  11 contribuDefined tion of 5  Contribution  percent of compe  Retirenmen sation. t Plans.”     EBRI Notes, no. 4 (Employee Benefit Research Institute,  Wage and Salary, Full-Year, Ages 21-64, for employees simulated to have between 31 and 40 years of eligibility, the percentage of lowest- contrib VanDer Presumab ution hei lyan  , tin he $2 d Cop  a retirement 0,00 eland (2 0 figu 003)  re savings w ould b  acco e inde unt. xe   d for inflati It reports on in the future similar to cu  estimates from the Taxrr  Pent treatment o olicy Center for f IRC Sec.  both an  factors, it is impossi accumulation potential ble to q  of ua  these ntify the err  plans. ors that ma    For the y  res lastu  10 lt from them.  years, average The RCS  balan was co- ces are spons  also ored  com by tpute he Empl d for o ya e e www.ebri.org/publications/benfaq/index.cfm?fa=retfaq14, last accessed July 26, 2011. year’s ?  RCS Decisions  include  to d two  opt  new out once  questions.  participa  Thetion  first  has  asked  begu  respondents n.   how imp 44ortant is being able to  contrib           .  u“The tion  In (401(k) fluenc -ety  of pe)  A u plans. tomati    Currently, c Enrollment,  the  Catch combination -Up, and  of  IRA  em  Contribu ployee an tions d employer  on 401(k)  con   tributions  employee the 3.2  si mulatio Defined  contribu n.  be tions. nef   iFigure t pla n 6 sanalyzes  only the financial (not behavioral ) change in match rates  classified as “at risk.”  Moreover, 41 percent of those in the lowest preretirement income quartile are  $5,000 or more in annual earnings, 10 or more employees ? April What  2011):  is the  appropriate 10?19.   percentage reduction for those 29.9  in the second 37.5  category above?  415(c) limits. Benefit Research Institute (EBRI), a private, nonprofit, nonpartisan public policy research organization, and Mathew income quartile workers achieving the 80 percent threshold was 79.2 percent, and that of the highest- 18 percent credit and a 30 percent credit. The paper includes a distributional analysis of the winners and  “consistent sample” of participants to control for the downward 31  bias that would otherwise exist from  However, many of these workers would fall into many of these categories simultaneously, such as being  54 dedu ?ct  the Contribution Accu ir re mulat tirem ions e  behavior. n tat  savings  Retire 11    m plan ent.”  co ntribu EBRI Issue tions  Brief,  from  no their  28 3ta  (Employee xable income  Benefit  in encoura  Researc ging h In  them stitut  to e,   12 is the lesser of a dollar limit of at Em  least p ployee  $49,000  Benefi  per tt Researc  year h  and  Institute  a percentage  (EBRI))   limit of 100 percent of an  VanDerhei (January 2004). Wage and Salary, Full-Year, Ages 21-64, The first major modification of the model occurred for the EBRI May 2004 policy forum. In an analysis to  (fro Accordin Fom r a historica  wha g tto  the  EBRI l  revie em  estimates, plwoye of cau r had ses of  be  the en this decl  perce  providing n ine se tage  to  e Olsen of  the  priv  18 ate and Va  perce -sector nD nterhe   govern wor i (1997). kerm s partici en t mat pating ch).   In in  th anis  em  case, ploym  for ent- predicted to run short of money within 10 years of retirement.   ? What is the appropriate portion of the household salary represented by the salary of the 401(k)  Greenwald & Associates, Inc., a Washington, DC, based market research firm. The 2011 RCS data collection was income quartile workers was 64 percent.  40 losers under the two versions of the proposal; however, the underlying analysis holds retirement saving  IRA 31  rollovers when 401(k) participants change jobs.   July 2005).  32     under        .   age “Capping  21, having  Tax- Preferred less than  $10,000 Retirement  in annual  Contributions:  earnings,  Preliminary  and not bei  Eviden ng a full ce -time,  of the  fu  Impa ll-yeactr  of worker.  the     save Finally,  for  th  retirement. e June 2011  Wh  Issue en confin  Brief allowed ed to fu lretirement l-time workers  inco  (n=591 me adequac ), the ywei  toghted  be assessed  results  at  were  retir  as ement  follows:  ages   $5,000 or more in annual earnings, 50 or more employees ? Asset allocation.  19.8 27.5 Employees age 50 or over may be allowed to contribute up to an additional $5,500 per year. employee’s compensation.      determine the impact of annuitizing defined contribution and IRA balances at retirement age, VanDerhei  those based  currently defined be  ages nefit  26–35  plan ,de  the creased  average  from  reduction  38 perc  varies ent in  from 1979  ato  lo  15 w  perce of 22.1 nt  perc  in 2008. ent for  Although  the highest  much    of  funded by grants from more than two dozen public and private organizations, with staff time donated by EBRI and 55 participant?  13 68 contrib VanDer uNational tions hei (2 0constan 05).  Commission t (page  6). on    Fi The scal  author  Responsi  me ww wntion bility w.ebri.org  sand  that  Ref   the orm  pro  Re posal comm  “could endations.”  conceivably  EBRI  No affect tes,  no. 7  Wage and Salary, Full-Time, Ages 21-64, Therefore, See VanD  the erhe  botto i (July 2 m007) f  of the or a summar  Figure 7 sh y of the respo ows the nu nmber ses of defin  of workers ed benefit spo  who would nsors  to the implem remain in a targete entation of d  later than 65.   12 13 In aggregate, the Retirement Savings Shortfalls (determined as a present value of retirement deficits at  ? Cash outs at time of job change.  and this Green   de Copeland wcrease ald. RC S materia took  (2004  pla ) wer cls a e by e nd a list of un  able  1997,  to dem  there deorn w  have s riters ma trate  been  thy at be access   afor  numb  a hous e ed at r of ehold   the EBRI W recent  see  deve king eb site:  alop  75m  perce ents www.n ebri.  that t probabili o rg/rcs have  made ty  of   income Iwry, 32 To summari  Mark  quartile  J.ze,  and  it  to   appears David  a hig h C.   of fr  John. om  23.1  both    2006. per  empirical cen   “Pursuing t for the  analysis  lowe  Universal st  and  inco  simulation  Retire me quartile. ment  results  Se   curi  based ty Throug  on tens h Au  of tomatic  millions     However, even with these empirical techniques, it was difficult to obtain a true value of the 401(k)  Section 41 (Employee Not at 5(c)all impo of the Internal  Ben rtan efit  tResear ............................................................................................ Revenue C ch Inost ditu e. te, July 2011):  2-6.  4.3% $5,000 or more in annual earnings, 10 or more employees 31.6 39.4 incentives the new fundi  for ng  firms requireme  to offer nts under the P  401(k)s ore  p nseion Protecti nsions“ (pag on Act of 2006 as e 7) but concl udes well as the p  that this otentia  seems l pe nsio unlike n expe ly.  nse He  VanDerhei (July, 2011) provides preliminary evidence of the impact of these “20/20 caps” on projected  population, if exclusions are made for age, annual earnings, work status, and/or employer size.  For  56 VanDerhei (March 2006). age 65) for these age cohorts (expressed in 2010 dollars) is $4.55 trillion, for an overall average of  For purposes    of the baseline results presented today, 50 percent is used for the first assumption and  IRAs.” Retirement Security Project, The Brookings Institution.  retirement income adequacy, the additional savings that would otherwise need to be set aside each  of defined  individual  bene  participa fit sponsors nt o  in bservations  the private  (dating  sector  all  re -the exa  way mine  back  the  cost to 1996 s and  in  ben  mae ny fits  cases),  of providing  that the  retirement     Not too important............................................................................................... 5.0% system’s    potential to generate significant 401(k) accumulations given that employees reaching  Wage and Salary, Full-Time, Ages 21-64, volatility under new FASB requirements. also dismisses as likely7  overstated the concern that the ma8tches provided in the proposal may  For more detail, see Helman, Copeland and VanDerhei (March 2011, online at www.ebri.org/surveys/rcs/2011/). retirement   33  accumulations under  a set of assumptions explained in detail later. While this provides a first  example, if the population of interest is wage and salary workers ages 21–64 who work full time, make  14 $47,732 per individual  still assumed to be alive at age 65.     Many VanDerhei, 4.3  of Caveats  EBRI  Ja’s ck,  previous   and with  Craig respect  simulation  Copeland. to  projects   Or automatic egon  (see  Future  App enrollment  Retire endix men A fort  aIn  brief c ome  chronology)  Assessment  will  Project.  be direc  A project tly17     100 See testimon  perce Somewh nt isy for this hearin  used at im  for portant  the second. g b ......................................................................................... y Jud   y Mil The le  results r for an exampl  will be e of this ana sensitive to ly sthe is.  values assumed for 27.8% these two  $5,000 or more in annual earnings, 50 or more employees benefits year unti  thro l retirement ugh the  form to ach  ofie  av equalified  this obj edefined ctive would  bene  decrease fit plan 21.9 .  by   How  a media ever, nthese 28.8  amo unt plans  of  still 30  per cover cent.  millions     traditional (VE) type of 401(k) plan under the current set of tax incentives has the potential to generate  retirement age only had the possibility to be covered by a 401(k) plan for a portion of their career.  In  57 Olsen, Kelly and Jack VanDerhei. "Defined Contribution Plan Dominance Grows Across Sectors and  discourage employer matches to 401(k) plans.   approximatio VanDerhei (Sneptemb  of thee  pote r 2006 ntial )  impact of these constraints on workers, as well as the distribution of the  14 $5,000 or of Very impo   more the EBRI  in  rt a aEducation nnual nt ..................................................................................................  earnings,  and Research  and work  F und for an  and  em  thpeloye  Milrb  with ank  Me 10 mo or more rial Fun  em dployees, , 2001a. 61.5%  31.5 million   38 applicable to such additional research and we will be happy to work with the Senate Finance Committee  Wage and Salary, Full-Year, Ages 21-64, assumptions The Th  previous is does not   and results necess  addi  assumed ation rily im alpl  sensitivity y  tnon hat man e of  the yanalysis ex isti 401(k) ng d  ise   presented parti fined cben ipants efit spons  in  were  Appendi o  automatically rs have or x C.  will terminat  enrolled e their  in 18 the existin   g Additional refinements were introduced in 2005 to evaluate the impact of purchasing long-term care  aof   In the RCS, sum  U.S.  that  workers  wh retiree e nand  combi refers to i  have ne  lon d ndivi with g bee du  Social n als  valued w hSecuri o are retired or  as ty an  ben  integral efit ws h o ar would  com e ag pon  rep e 6ent 5lac or old  of e a  re  sizeable etr an irem d not empl en  portion t income oy eof d  adequa full time.  the  cy  for  34 an attempt to control for these problems, EBRI and ICI produced a joint publication in 2002  with  Employer Sizes, While Mega Defined Benefit Plans Remain Strong: Where We Are and Where  The The reaso  study was n that the  also able youn  to ges  docume t age cohnt ort  th do e es not follo degree tow  which this trend is du  eligibility e to their rel  for partic aipation tively lo win e r curr qualient fied w  ages impact by    income, it does not tell the entire story. A follow-up study will also explore the likely impact of  $10,000 or more in annual earnings, 10 or more employees worked for an employer that did not sponso 55r a retirement 27.4 plan in 2008 (mea 35.5 ning that 46 percent of the  defin 58 ed benefit plans. Instead the process of freezing these plans for current and/or new workers has increased to retirement  provide  cost/benefit plan; instead,  assessments  workers’ escalation  of these  of types  con  of tributions  proposals  after  in the  the  fu  first ture  yea .  r are driven primarily by  insurance on retirement income adequacy.   their VanDer  househ hei an olds. d Cop     eland (2008). While employee’s  these We   Are preretirement two  Going."  papers  In pr  ovide Re income tire  an me  for  entx  tr th Prospects emely ose fortunate  in ter in ae sting Defined  enough  analysis  Contributi  to have  of a  proposal conti on World nuo  with us . Washi  coverage  profound ngton,  during   pu DC:blic   th  eir  simulation _____. A Behavior  resultsa  showing l Model for Predicti  that under ng E  a m contin ployeeuous Contri  coverage butions to 401(  situation, k) Plans. North  401(k) participants American Actu  could arial  expect  than older cohorts after adjusting for historic age/wage profiles. retirement plans (especially defined contribution plans) matters with respect to “at-risk” status.  For  Wage and Salary, Full-Year, Ages 21-64, Figure If one  we 3 provi re tod  look es th  at e  avera this from ge per  a strictly centag efinancial  reductions  perspective,  in 401(k)  one acco  would unt balances  assume  at  that  Social  the  Secu  lowerity r-income     these Worke rconstra refers to all i ints on ndiv  re idu tir als em we hn o are n t plano  sponsor t defined as retire  behavior es, regard  and esti less mate of emplo  the ex yte ment nt  status. to which fewer  total substanti  nonself ally in - recent employed years. F  working or more infor  for an em mation o ploye nr the i  that m  did pact of plan f  not spons reezes o or a plan n w ofell rkers see Van  into this grou Derhei (March p).  Yet, if  age and income characteristics as opposed to tenure with the current employer, as they would be in  Employee Benefit Research Institute, 1997: pp. 55-92  Journal, 2001b. working careers.  Moreover, the AE type of 401(k) plan when combined with automatic escalation  policy implications, it appears quite likely that some of the assumptions with respect to responses (or  to replace someplace between 51 and 69 percent (depending on income quartile) of their pre- $10,000 or more in annual earnings, 50 or more employees 59 19.7 26.0 example, the at-risk probability for Gen Xers varies from 60 percent for those with no future years of  individuals (those most likely to pay no or low marginal tax rates and therefore have a smaller financial  normal 200 35 6). F  retirement or an analysis of  age by wh ether “frozen” permanentlyw  modifying orkers have b  theen fin e exclusion ancial lof y ind  employe e  mnified vi e contribu a enhance tions d em  fo plo r yer employers Copeland  an wou d VanD ld beerhe  willing i (20  to 10).  offer qualified defined contribution plans (especially among plans offered  a The  more  model  restrictive  was next  defini  used tio  inn  March is placed  of  on 2006  the  to  targe  evalua tedte  population,  the impact  so  of  that defin  only ed ben  worekfit er  freezes s who wor  on k full- EBRI used its Retirement Security Projection Model (RSPM)® to evaluate the importance of defined  39 Although additional analysis needs to be performed before assessing relative importance of these factors, it auto-enrollment plans (especially those with automatic escalation of employee contributions).    provisions appears to have the potential to produce even larger retirement accumulations for many of  Wage and Salary, Full-Time, Ages 21-64, lack Actuall  thereof) y, the constrai  from employ nts woees uld nee  and d to be c  (more osignifi mpared to the cantly) pla 402(g) l n sponsors imit as  wwill ell as a  be th nye pl  subject an-specific con  of serious straints on   retirement income assuming the purchase of a (nominal) annuity at age 65.    33 contribution to defined contribution plans, see Copeland and VanDerhei (2010). eligibili incentive VanDerhei, ty  in to   Ja  ad e ck. defined duct  Testi  retirement  m con ontryi bution for  savings the  plan U.S.  contrib   to Senate  20 percent u  Spe tions ci afrom  lfor  Committee  those  taxable  with   income) on  20  Aging  or more  would  Hearing  years.  be  on  least   Ret  likely iremen  tot  rate     retirement _____. The F  savings uture of Reti  plans fr rem om ent Incom  taxable income e: The Cha  in 2012 nging Face of Private  (by age and age-specific Retirem  sa ent Plans lary quartile (pp. 121-147 s).  As  ). by 6  small Conclusions  employers).   time, participa  fulln -yea ts by r,  make simulating  $10,0  the 00  or mi  more nimum  in employer  annual earni  conngs, tribution  and wor  rate k  for that  an  would  empl oyer be nee  with ded  100  to  for inancially  more    benefit appears that th  plans is result is cau for households sed b  assuming y at least t  they wo factors.  retire at Fi rst, the definitio age 65 and shn of income q owed the truemendous artile in RSPM is determi  importance ned of  60 $10,000 or more in annual earnings, 10 or more employees 29.9 37.3 tax contributions (primarily for the Highly Compensated Employees).   those VanDer  covere hei (2d0  by 09).  such a plan during a significant portion of their working careers.  debate among those with first-hand knowledge of the decision-making process of employers who must  Planning: “Do We Have a Crisis in America? Results From the EBRI-ERF Retirement Security  Nation 45al Academy of Social Insurance: The Future of Social Insurance: Incremental Action or The this  as exclusion  “very im  ofp oauto rtan -t. enro ”  However, llment plans  tho sin e  in th is the  analysis  lowest  was  househol  necessdary  income  given  cat  thee gory current  ($15,0  mode 00 ling to less     expe in a man ctener si d, the milar to the av  younger coh erag orts e ind  would exed  exp monthl erie y e ncaerni  larger ngs computati  reductions on for Social Sec  given their increased urity with the fo  expllo osu wire ng  to the  15 In 2010, EBRI  documented a significant reduction in the percentage of households “at risk” for  indemnify employees, defined be  th ne  on efit  employees ly  plans  5.5 millio  in ac  for nhieving  workers  the re  retirement duction  (or 11  per in  income th cen eirt)  ex  would  adequa pecte  be d cy retir  in  for cle uded ment  Baby  among   income Boomers  those  un  and der  working   Gen various  Xe  for rs.  rate     an- of- Wage and Salary, Full-Time, Ages 21-64, When the 2002 study was performed, very few 401(k) sponsors had adopted any type of automatic  For example, an employee age 60 may have very recently changed jobs and rolled over a substantial account In fact, it can be argued that much of the problem with retirement income adequacy in this country is  decide whether Projection  or  Model,” not to sp  Jan. onso  27 r a,  2004 defined  (T- 141 contri ). bution plan and, if they do so, how to best design the  If 40 the 20/20 Funda  caps mental Reform,  are assumed 2002 to bea  imposed .   starting in 2012, the annual percentage reductions in  61 56 than modific  $25,000 ations: ()a  actually ) All earned i  have ncome is inc  the largest lud  perce ed up to the a ntage of ge  responden of retirement ts (i.e., there is no maximum ta classifying the tax deductibility xable w  of ag   e $10,000 or more in annual earnings, 50 or more employees proposal. assumption 1.4 percent re   Fo  of cusing  spon no job de  on d that the  chan  thosge. e currently    y di It would d not kno  2  be 6w–35,  .very  the  difficult  average  to  provide 20.8 percentage  a valid  reductions  analysis 27.3 19   vary of the  from  average  a low    of 11.2  employer inadequa return The VanDer  views  assumptions.  e hei (A te  xtha p re ressed tt pril 2  di ired  in m  not 010).  this en    t  spo stat  Later income enmsor  ent that   a  be are   plan. ye twee  ssolely ar,    an Of n  th  2003   ocourse, update se of  and  Jaack d  another    2version VanDerhei 010, based   way of  aa nd the   to  in shou    mo look large lddel  not  at  part  was    b bthis e attributed  on   develop la st the  nu  advent  mber to ed  the  to   Employee is   enha of  that  auto nce  89  -B  enefi the t  Overall, the presence of a defined benefit accrual at age 65 reduces the “at-risk” percentage by 11.6  balance from his previous employer to an IRA. enrollment (AE) provisions for their plans.  In a 2005 follow-up study with ICI,  we looked at the  one of whether a household is covered by an employer-sponsored retirement plan.  Appendix B  various plan parameters (including the match rate and match level) to meet their objectives.  401(k) account balances at Social Security normal retirement age are displayed in Figure 2 by age and  base constraint and the calculation terminates at retirement age); (b) Instead of indexing for changes in average Wage and Salary, Full-Time, Full-Year, Ages 21-64, ® 46 Research Institute (EBRI), th he EBRI Educatio on and Research h Fund, any of itss programs, officcers, trustees, sp ponsors, or othe er staff.  perce contrib ntage utions  re as ductions  very important  in 401(k)  (76.2 balance  pe rcent). under   auto   -enrollment because very little, if any, information  percent for the highest income quartile to a high of 24.2 percent for the lowest income quartile.  As this  enrollment _____. “Projections  in 401(k)  of  plan  Futs;u re however,  Retirem  for en tthe  Income  one- thir Security: d of the  Im  households pact of Long  with -Term  the  Care  lowest  Insuran -indecxe.” ed   pre- perce EBRI perce ____ _. in n n ttage tKa e  of rac nsas  th  points. tive ese  Futur  Ballpark  workers  Th e eRetir  def  E  with $eined timate men  those  benefi t In  worksheet c o characteristics me t plan  Assessment  advanta  by providing  worked ge  Project.  (as measu   Mo for  A  nt an  pr er  ed employer oject Carlo  by   of th simulations  ethe   ga that  p EBR  be  diItwee d  Educatio   of spons  the n the  onecessary rn   a  two and     at- risk  41 4 potential The  change pote  inntial  401(k)/IRA impact  rollover of  ta accu x mulation reform s as on  a  result retirement  of changing secur  the tradi ity tional  voluntary  62 16 EBRI plans to expand this analysis with a complete set of sensitivity analyses in the near future. presents evidence from Copeland (2010) to show where potential legislation may exclude workers, or  $5,000 or more in annual earnings, 10 or more employees VanDerhei and Copeland (2010). 25.3 31.2 age natio -specific nal wage  in s, the model in come quarti dlees xes bas  for all ed o  401 n(k) assume  partid after cipants -tax rate of retur  with salariesn  in base  exce d on ass ss of $10,000 et alloca tio and ns that are a  tenure of  The Employyee Benefit Research Institute is a nonprofit, non npartisan, educaation and researrch organization established in  Year-end 2010 data is currently being analyzed and the annual update should be available soon. American Society on Aging/National Council on Aging joint conference, March 2005.  analysis currently  Research was  exi sintended ts that  Fund  ca  to n  and   be look   used th eexclusively  Mil  tob  tra ankck  Me   what at mo the  arial  impa utom  Fun ct ad  tically of , Ju  changi ly  e16, nrolled  ng 2002.  the  participa   exclusion nts  of  with  employee  automatic     replacement retirement retirement  pl income,  rates an in  2009. nee  thed  at ed   - risk for  specific percenta  prges, obabilities  while mu  ofch  retirement  smaller (th  income ey were  ad  80 equa  perc cyent  under  in 200  alternative 3) are still     percentages) is particularly valuable for the lowest-income quartile but also has a strong impact on the  Public policy consideration of this proposal will undoubtedly be subject to some type of a cost-benefit  Prior enrollment  to estim  (VE) ating  401(k)  the  potential plans to AE  re du placti ns.o ns Although  in accumulations  we had th eresul  advantage ting from  of  401 using (k)  a  c  da onttaba ribuse tion  ofs,  tens  a set  of     Wage and Salary, Full-Time, Full-Year, Ages 21-64, function of the individual’s age in each year; and (c) Percentile distributions are established based on population The the Washingto   num second ber n n, DC,   qu of  in  est workers  1978. ion     asked EEBR  who I does  of  are   those not  already  taake  currently  policy  being  positi   saving reached, o ons, nor  for  does  by  retirement   certai it lobby, n  demographi advocat  was “S ee sup pepose cific c and  policy  you  em   rrwere ecomme ployer  no  ndati  longer o ons,  or  42 at least two years.   63 57 escalation of contributions would do upon job change.  For example, if a participant has already been  contrib 17 Future ana utions lysi  from s will attempt to incl  taxable income, ude  it a fou  was rth categor  assumed y  for that those  the w total ho w  matching ould increas  con e thtribution e amount the  would y save for  remain  extremely high (70 percent in 2010).  Of course, when one limits the analysis to those who are simulated  risk middle VanDer  manage  class hei (S ment  (the eptemb   rtrea eduction er 20 tments. 10  in ).  the    at-risk percentage for the second and third income quartiles combined  $5,000 or more in annual earnings, 50 or more employees analysis beyond one that is assuming retirement saving con 17.6 tributions will remain 22.9  constant.  It is  millions The propos  of 40 ed 1(k) regu  parti latioc nipa s for 401(k) p nts goingl ans back w ein re pu  some blis  cases hed in  Novem to 1996, ber 1  we981 a  were nd muc  limite hd of the gro  in knowing wth i  how n these   of receive  baseline  fede eral  re  funding. sults first    need to be run to determine the likely values if the various tax reform options  ____ statistics for each age co _. “Defined Benefit hort. T  Plan herefore, it  Freezes: is  Who's possi ble th Affected, at an ind  How ivid  Mu ual cwhh, ose prer and Replacing etirement incom  Lost Accruals.” e ranks in the  EBRI  _____. Massachusetts Future Retirement Income Assessment Project. A project of the EBRI Education  characteristics, annual earnings, employer size, and work status (full-time/part-time).  allowed to deduct retirement savings plan contributions from your taxable income.  What do you think  43   retirement in response to this proposal. Wage and Salary, Full-Time, Full-Year, Ages 21-64, escalated constant. to    8 percent of compensation and upon job change is automatically enrolled into another  to be saving in the future, the numbers improve substantially: among Gen Xers without any future  is pla  9.7 ns took pl  percen actage e in the n  poinets xt few  which ye ars. corresponds to a 19.5 percent relative reduction).  admittedly lowest quartil Issu  ve ee over their rem  Brief, ry difficult  no. 291  to a  inin deter (Employee g wm ork h ine i stor how Benefit y ma  those y i  Research n deem edployees end u  Institute, p  not with a  currently  Mar n incom ch 2006). e  that w covered   oul d rank h and/or iparticipa gher than th ting e   are Several workers  not  im poin and  would po  ts Researc se  stand  dreact  on the  h  out to  Fun   AE current  immedia d  provisions,  and  401 the tely:  (k) Mi    and lbank syst  thus em  Memorial .    simulated The model  Fund  the  use , December  likel d iny  th  response is article  1, 2002.   us is ing based    the  on  results  the 401(k)  of    64 you (and your spouse) would be most likely to do?”  When confined to full-time workers (n=460), and  VanDerhei (October 2010a). RSPM was significantly enhanced for the May 2008 EBRI policy forum by allowing automatic enrollment  $5,000 or more in annual earnings, 100 or more employees 14.6 19.3   401(k) plan, would 20  they “remember” where they had been, or decrease contributions to the default rate  ® 43 eligibility for participation in a defined contribution plan, the at-risk percentage is 60 percent, but it  bottom quarter in one or more specific years. Second, the impact of the 20 percent limitation for the lowest-income in a defined contribution plan will react to this set of incentives, and EBRI will continue to work with  academic studies.   What we found was that the overall expected improvement in retirement  voluntary enrollment modules from the EBRI Retirement Security Projection Model  (RSPM) and is  18 For example if a 401(k) plan sponsor had been providing a 50 percent match prior to the introduction of a 30 58  4 full- Wage and Salary, Full-Time, Full-Year, Ages 21-64, eliminati  In askingn  emp g those loye  who rs and  ref  providers used to  answer  how they  or  responded thought employers  that they  wou  did lno d respond t know,  approximately to the change  in 1  in policy,  Holden and VanDerhei (2002). 34 of ____  401(k) _. "Mea  participants suring Retire  withment  the  potential Income Adequacy:  for automatic  Calculating  escalation  Realistic  of contri  Inco bution me Res place to bem  included. ent Rates.  "  _____. “Can America Afford Tomorrow's Retirees: Results From the EBRI-ERF Retire 47ment Security  65 3.3 Defined contribution plans of quarti  the ?le ma  new With y fal  pl an? the l dispr   exception oportionatel  ofy on th  the earliest e part-time  age wcohort orkers. F  (t oho r exampl se currently e, a work  26–35), er who enters the  the average wor kreduction force part   drops all the way to 20 percent for those with 20 or more years of future eligibility.     VanDerhei (October 2010b). data perce 3  Components from nt gover  sponsors nment match, it w  who of  automati retirement as assum cally ed that the pl  enroll sec  their u an sp ri  emty onsor ployees woul  in d de  401(k) crease the p  plans to la nbetter match to 20 perce  assess some nt  of  similar accumulation  in many s—especially  respects to  for the  the  one  lower  used- income in Holden  quar  antdiles—were  VanDerhei  nothing  (2002)  in less  tha  thtan  it  looks spectacular.  only at    current    $10,000 or more in annual earnings, 10 or more employees 24.9 30.8 time workers EBRI Issue  (25.6  Brief,  percen  no. t)  297 indi cated (Employee  that they  Benefit  would  Research  reduce  Institu  (in some te, Sept  cases em  completely ber 2006).)  their  it was suggested Projection  that  Model.”  they would  EBRI Iss  allow ue Brief,  the govern  no. 263 me  (Employe nt to do ethe  Benefit  match  Research  to keep  In peopl stituet econ , November tributing   Additional modifications were added in 2009 for a Pension Research Council presentation that involved  19 time whose spouse already has a full-time job may be in a better situation to attempt to maximize retirement Given the phenomenal growth of defined contribution plans (especially those 25 with a 401(k) feature) in  for any income quartile decreases for older age cohorts. This is due to the fact that those closest  under the proposal. Wage and Salary, Full-Time, Full-Year, Ages 21-64, the Hold  behavioral en and VanD  tendencies erhei (200  of 5).  this group.    401(k) In addi tion part ic toip  individual ants and do  saes vings  not  and,  attempt  to an  to  incr  include easing  eligible  extent,  nonp  partarticipan -time work ts  in  or  re  workers tiremen  who t, the  are  major     66 2003).  contributions if the ability to deduct them was eliminated.  The lowest-income category ($15,000 to less  and As additional  use their  information dollars to pay  bec  foro m the es  incre available asing  with  cost  res of p employee ect to empl  heoyees’ alth insura  behavioral nce.  Al  responses though EBRI  for  is  auto   - VanDerhei (February 201th 1). a Given contrib  winners/losers  uthat tions  the on his/her i  financial  analysis ncom  fate  of e  . Althoug define of future d ben h EBRI is in the  geenerations fit freezes  of pr  an  ocess of attempting to mod redtirees  the enhan  appeaced rs to  defin  be so ed e l  strongly contribution the impact on  tied  to em part-timers on  whether ployer    the A year  private  after EBRI  sector  this    study  1100 in the   was 13  las  St.  released, t  three NW #8 0 0deca 0   Congress des, Washi  it  appears passed n ngton,  DC th  that  e 2000  Pension  this 55    form  Protection (202)  of  65 em 99-p0670 loye  Ac   tr -of provided ww  2006 w w.ebri.org  (P re PAt),ir    ewhich ment   to retirement26  age will have fewer years of future contributions subject to potential reduction as  $10,000 or more in annual earnings, 50 or more employees _____. “Retirement Income Adequacy After PPA and FAS 158: 17.4 Part One—Plan 22.6 Sponsors' Reactions.”  componen currently no tst  of eligible  retirement .   However,  security  unlike  in this  the  country  2002  mod for several el, this  de ancades alysis  have assumes  bee n no  Social  job turnove  Security r,  and  44 20 59 48 enrollment, than $25,000  EBR ) has I will  the  up  largest date this  negative  analysis  reac  tot iprovide on to this  a  more proposal,  robust  with  model.  56.7  percent indicating a savings  currently a lon The fact that gitudin  wor al bk a wing sis, the curre e c oon nducte  a survey d this pre nt anal  toys  eli limi is filte cit nar  potential y red out an analysis  empl y w40 ith1(k) partici ou oyer t an emplo  response pant yes e w to reactio ith  this ann n  proposal, ua to the chan l income of less t  itge i  may n match rates  be han $   10,000 they Choi, La  are eligi ibson, ble Madria  to pan, and rticipat Metrick (200 e in employe 2). r-sponsored retirement plans,  the logic of modifying (either  contributions provided as a quid pro quo.   Wage and Salary, Full-Time, Full-Year, Ages 21-64, Until 67  thisEBRI  typ eIssue  of informati  Brief, no. o n337  is available  (Employee  it will  Benefit  be quite  Research  difficult  Institu  to accurately te, July 2007).  assess    the “benefit”  plan ____ eased  will _.  some  “ERISA  provide  of  At the   a30:   substantial ad  Th mien iDecline strat27  per ive  of cbarriers en  Private tage  of to -Sector   non providi -Social  De ngfi nAE  Security ed  and  Benefit  for  retirement  the  Promises  first time  we  and al  setti  tAnnuity h for ng  Baby up  Payments:  safe  Boomers  harbor      a result of the 20/20 caps.    VanDerhei (April 2011). employer withdrawals, -sponsored  or loan  defau retirelment ts.    plans.    as well as those with less than two year of tenure. reduction. should in no   way should be interpreted that EBRI does not believe that a reduced employer match may have instructive to model this scenario to assess the additional reduction in 401(k) retirement accumulations  $10,000 or more in annual earnings, 100 or more employees 14.4 19.1 completely or marginally) the incentive structure of employees and/or employers for defined  What Will It Mean?”  EBRI Issue Brief, no. 269 (Employee Benefit Research Institute, May 2004).   portion of the cost-benefit analysis suggested above.  However, EBRI is currently in a position to provide  and provisions  Gen Xers  for.   automati Unfortuncat  escalation ely, the “success”  of employe  of these e contributions.  plans are someti  Although mes  it  measu  was too red  soon  by me  to trics know  tha  how t are     Source: Employee Benefit Research Institute estimates from the 2011 March Current Population Survey. 12 20 10 23 22 17 24 18 15 21 13 16 14 11 1 5 7 4 9 3 6 2 8                                                 Figure 2 Average Percentage Reductions in 401(k) Account Balances at Appendix C: Sensitivity analysis on baseline assumptions for Figure 3 Av Av Ave e errra a ag g ge e e   Pe Pe Perrrccce e en n nttta a age ge ge   R R Re e eductions ductions ductions   in in in   401(k) 401(k) 401(k)   Accoun Accoun Accounttt   Balances Balances Balances   at at at    Social Security NRA* by Imposing 20/20 Limits in 2012, Average Percentage Reductions in 401(k) Account Balances at  Average Percentage Reductions in 401(k) Account Balances at Social Security Normal Retirement Age by  Social Social Social   Security Security Security   Normal Normal Normal   Ret Ret Retiiire re rem m me e ent nt nt   Ag Ag Age e e   by by by   P P Pe e ermanen rmanen rmanentttly ly ly   Modifying Modifying Modifying   the the the   Ex Ex Exclusion clusion clusion   of of of   Emplo Emplo Employyyee ee ee    by Age and Age-specific Salary Quartiles Social Security Normal Retirement Age by Permanently Modifying the  Permanently Modifying the Exclusion of Employee Contributions for Retirement Savings Plans From Quartiles Con Con Contttributions ributions ributions   fo fo forrr   Ret Ret Retiiire re rem m me e ent nt nt   Sa Sa Savings vings vings   Plans Plans Plans   Fr Fr From om om   T T Ta a axxxable able able   Income Income Income   in in in   2012 2012 2012   and and and   Assuming Assuming Assuming   tha tha thattt    16% Taxable Income in 2012, by Age and Age-specific Salary  Exclusion of Employee Contributions for Retirement Savings Plans Salary   From  all all all   Plan Plan Plan   Sponsor Sponsor Sponsorsss   Dr Dr Drop op op   the the the   Plan Plan Plan   Ma Ma Matttccch h h   and and and   all all all   Employ Employ Employees ees ees   Re Re Reccce e eiiivvve e e   a a a   18 18 30   Pe Pe Perrrccce e en n nttt   Ma Ma Matttccch h h   fr fr from om om    Quartile the the the   Gov Gov Gove e ernmen rnmen rnmenttt,,,   by by by   Ag Ag Age e e   and and and   Ag Ag Age e e- - -specific specific specific   Salar Salar Salaryyy   Quartiles Quartiles Quartiles Taxable Income in 2012, by Age and Age-specific Salary Qua 14 Percentage %  reduction 0.25 0.25 0.5 0.5 0.75 0.75 Lowest Percentage of family income 0.5 1 0.5 1 0.5 1 Assumesno employee behavioral modifications.  Only the difference in employer contributions vs. Government matches  Assumptionsfor this run (sensitivity analysis  on these assumptions are available on request from the author): (1)  401(k) participant's  Assumptions for this run (sensitivity analysis  on these assumptions are available on request from the author): (1)  401(k) participant's  Assumptionsfor this run (see Appendix C for sensitivity analysis  on these assumptions): (1)  401(k) participant's share of household  2 is tracked. share share  of household  of household  income  income  = 100  = 100  percent;  percent;  (2)  401(k) (2) 401(k)  participants  participants  who  who  "reduce"  "reduce"  contributions  contributions  are  are assumed  assumed  to reduce  to reduce  them  them  by  50 by  percent 50 percent 12% income = 100 percent; (2) 401(k) participants who "reduce" contributions are assumed to reduce them by 50 percent 25% 26-35 Lowest income quartile 13.1% 15.4% 18.9% 24.2% 24.8% 33.1% 3 40% 26-35 2 8.9% 12.0% 13.2% 18.5% 17.5% 25.0% 30% 45% Highest 26-35 10 3 % 6.1% 11.6% 9.9% 15.8% 13.8% 19.9% 35% 26-35 Highest 5.8% 7.2% 9.9% 11.2% 14.0% 15.2% 40% 20% 25% 36-45 Lowest income quartile 11.6% 12.9% 16.9% 20.6% 22.1% 28.3% 8% 30% 36-45 2 6.3% 11.9% 9.8% 17.0% 13.2% 22.1% 35% 36-45 3 5.3% 8.1% 9.0% 11.3% 12.6% 14.6% 25%20% 15% 6% 3630 -45 % Highest 5.2% 5.3% 8.8% 8.7% 12.4% 12.2% 46-55 Lowest income quartile 10.1% 11.3% 14.7% 18.1% 19.4% 24.9% 20% 25% 46-55 2 5.1% 9.8% 8.1% 13.8% 11.0% 17.9% 15% Lowest Lowest Lowest   income income income   qu qu quartile artile artile 4% Lowest Income Quartile 46-55 3 4.3% 6.4% 7.3% 9.1% 10.3% 11.9% 10 15% % 2 2 2 20% 2 46-55 Highest 4.1% 4.1% 7.0% 6.9% 9.8% 9.7% 10% 3 3 3 2% 56-65 Lowest income quartile 8.6% 10.1% 12.6% 15.9% 16.5% 21.7% 3 10% 15% 56-65 2 4.7% 7.4% 7.1% 11.2% 9.5% 15.0% Highest hig high hest est Highest 56-5% 65 3 3.1% 5.1% 5.3% 6.9% 7.4% 8.8% 5% 0% 10%5% 56-65 Highest 3.0% 3.0% 5.0% 4.9% 7.0% 6.9% 26-35 36-45 46-55 56-65 0% 5% Age Source: Author's calculations based on results from EBRI Retirement Security Projection Model Version 110910c2a-j 26-35 36-45 46-55 56-65 0% 0% 26-35 36-45 46-55 56-65 Source: EBRI Retirement Security Projection Model Version 110627c1. 0% 26-35 36-45 46-55 56-65 NB: this simulation only models the financial impact of the expected reduction in 401(k) contributions for employees who are not automatically enrolled by 26-35 36-45 46-55 56-65 Source: Author's calculations based on results from EBRI Retirement Security Projection Model Version 110910c2a and  responses to the 2011 Retirement Confidence Survey. imposing the new limits and does not attempt to assess behavioral modifications on the part of either the plan sponsor nor the employees assumed to be Source: Author's calculations based on results from EBRI Retirement Security Projection Model Version 110910c2 and  responses to the 2011 Retirement Confidence Survey. eligible for participation in the plan. The simulated rates of return are the same as in VanDerhei and Copeland (July 2010). This version of the analysis Source: Author's calculations based on results from EBRI Retirement Security Projection Model Version 110910c2b and  responses to the 2011 Retirement Confidence Survey. NB: This simulation only models the financial impact of the expected reduction in 401(k) contributions for employees who are not automatically enrolled by  modifying the exclusion of  Source: Author's calculations based on resultsfrom EBRI Retirement Security Projection Model Version 110910c2b 1and  responses to the 2011 Retirement Confidence Survey. assumes no job turnover, withdrawals or loan defaults. The full stochastic nature of the model will be included in future analysis. NB:employee  This simulation   contributions  only model  for sret  the irement  financial  savings  impac  plans t of the  from  expected  taxable  reduction income and  in  401(k) does not  contributions  attempt to  assess for employees  behavioral  who  modi  arefi  not catiaons utom onatic  theally  part  enr  ofo  the lled  eligible by  modi  non fyin -par g the tic ipants. exclusion  The  of  simulated     NB: This * N  simulation ormal ret  only rment age.  models the financial impact of the expected reduction in 401(k) contributions for employees who are not automatically enrolled by  modifying the exclusion of  employee rates of    ret contributions urn are the  for same  ret ias rement  in VanDerhei  savings  and plans  Copeland  from taxable  (July  i2010). ncome  This  and  version  does not  of  attem  the analysis pt to assess  assumes  behavioral  no job  tur modi nover ficat , withdrawals ionson the part  or loan  of the  defaults.  plan sponsor  The full  nor  stochastic  the eligible  nature  non  of -  the  NB: This simulation only models the financial impact of the expected reduction in 401(k) contributions for employees who are notautomatically enrolled by  modifying the exclusion of  employee  contributions for retirement savings plans from taxable income and does not attempt to assess behavioral modificationson the part of the eligible non-participants. The simulated  parmodel ticipants.  will  The be included  simulated  in  rat a futur es ofe  ret  analysis. urn are the same as in VanDerhei and Copeland (July 2010). This version of the analysis assumes no job turnover, withdrawals or loan defaults. The full  employee  contributions for retirement savings plans from taxable income and does not attempt to assess behavioral modificationson the part of the eligible non-participants. The simulated  rates of return are the same as in VanDerhei and Copeland (July 2010). This version of the analysis assumes no job turnover, withdrawals or loan defaults. The full stochastic nature of the  stochastic nature of the model will be included in a future analysis. rates of return are the same as in VanDerhei and Copeland (July 2010). This version of the analysis assumes no job turnover, withdrawals or loan defaults. The full stochastic nature of the  model will be included in a future analysis. model will be included in a future analysis. Reduction in 401(k) Balance

Volume T-170

Pages 31

EBRI Testimony

Sept 15, 2011

Jack VanDerhei

Financial Wellbeing Retirement