Workers’ finances can face many challenges over their careers, potentially leading them to have to take on debt or find other sources of financing to cover the various financial challenges. This study builds on prior work done by the Employee Benefit Research Institute (EBRI) and J.P. Morgan Asset Management focused on defined contribution (DC) plan participants’ behavior when faced with one specific financial challenge: irregular expenses. Specifically, changes in credit card utilization, DC plan contributions, and/or DC plan loan use were examined after these participants experienced a significant spending “spike.” This analysis goes the next step to examine where participants who took a DC plan loan are spending that money. Key findings from this study include:
- In this sample of public-sector DC plan participants where a loan option is available, 10.9 percent took a loan during the year of interest. The likelihood of a participant taking a plan loan increased with age through the 50s then declined. Household incomes of $30,000 or above did not appear to have an impact on the likelihood of a plan loan being taken, as the percentage taking a loan in each income group was in the 10–11 percent range, and once the account balance reached $10,000, the likelihood of taking a loan was similar through balances of $100,000 or more. The percentage who took a loan increased substantially with credit card utilization, as 6.9 percent of participants in households with no outstanding credit card balances took a loan compared with 19.8 percent of those who had outstanding credit card balances equivalent to 80–100 percent of their credit card limits.
- DC plan participants in households with higher credit card usage had lower average contribution rates across all ages except for those ages 60 or older. For example, for participants in their 50s, those from households with higher credit card utilization had an average contribution rate of 4.7 percent, compared with 5.6 percent for those with lower credit card utilization. As a result, participants in households with high credit card utilization had lower average account balances.
- Among those with a new DC plan loan, health care spending was the most likely to have increased, as 58.5 percent of households where a participant took a loan saw their spending on health care increase by more than 10 percent in the year they took the loan. This was followed by travel (22.4 percent), entertainment (19.0 percent), and non-specified cash spending (18.9 percent). Comparing the spending increases by categories with those who did not take a loan, only health care spending showed a higher likelihood of having increased by more than 10 percent among those taking loans. Otherwise, spending changes were very similar between households with or without a plan loan.
- Loan usage increased among those with higher credit card utilization, which would be an indicator of households being more likely to be financially stressed. Spending increases on health care were more prevalent among the financially stressed households whose plan participants were ages 50 or older, as 63.3 percent of the households where a loan was taken had this increase compared with 56.8 percent of the households where a loan was not taken.
- In an alternative test, the share of total spending that each category represented was compared between the year prior to the loan incidence and the loan year to see if any category spending share increased by more than 5 percentage points. The spending categories most likely to have seen an increase in their share of total spending of this size were unspecified cash spending (24.9 percent of the households), health care (23.3 percent), and housing (21.0 percent).
- Only housing spending and unspecified cash spending had higher likelihoods of share increases for those taking a loan vs. those who did not. Otherwise, the likelihoods of the changes in the shares of spending in each of the other categories were either similar between the households or less for the households with a participant taking a plan loan.
- Households who started mortgage payments in the year of the loan incidence analysis were more likely to have taken a plan loan than those who did not start mortgage payments in that year — 15.6 percent vs. 10.7 percent. This was true for households with plan participants of all ages. Looking at this correlation in the opposite direction, the percentage of those having a new mortgage given that they had taken a plan loan was 5.8 percent, compared with 3.8 percent starting a new mortgage when they had not taken a plan loan. Again, a higher likelihood of starting a new mortgage for those who had taken a plan loan was found across all ages. Regardless of whether the household took a plan loan, when a new mortgage was started, the probability of having an increase in housing spending of more than 10 percent was significantly higher. For example, 60.0 percent of those who took a plan loan and started a new mortgage had a housing spending increase compared with just 6.6 percent of those without a new mortgage but having a new plan loan.
This research, like prior J.P. Morgan/EBRI studies, found that higher debt can have a long-lasting impact on retirement security, since higher credit card utilization is correlated with lower DC plan contributions and account balances. Thus, the availability of emergency savings to help cover expenses can be a critical factor in preventing or stalling a cycle of increasing debt that can significantly impact retirement readiness, wherever the individual works. Furthermore, the finding that many participants have spending increases on health care when taking a plan loan suggests that examining the health insurance available to DC plan participants could also help improve finances, showing the intersection of health and wealth. While not an emergency expense, another common reason for taking a plan loan is buying a home, which is typically a household’s largest investment. A plan loan can help reduce the borrowing costs of a home purchase or even make the home purchase possible.
Figure 8
Figure 5
Percentage of Participant Households With an Increase of More Than 10 Percent
Figure 15
Figure 17
7%
Fi gure 9, Percentage ofA Pv art erage icipantCo House ntribu holdtio s With n Rat an es, Increas by Cred e of Mit ore T Card han 10 Usag Percent e and in AH ge eal th Care Spending, by
Lo Lo Again, increasin Pub Lo an an an lic • usag -s tecto he like I Inc In g de nc e an rin idenc idenc b DC al creas lih t that ca ter ood plan nati es e e s p a o amo ve n artic f n th s tes ign d C n e s ig pan t, ifi hare thos cantly hange th tse e in incr s w h mos are of impa iteas h s in S h t e ig c cas total s s ther retirem were es are a cre pendi compa pdi end ent t c lso i a ng that ng rrd covere e red with adi uti ness lizea atio d by a those , ch cate whe n, whi de r ever th who gory r fi ch ne d wo d ben ie d not t euld be an presente ind efit iviake a loan. p dula al d was in w n, an diork cato d compare st. Fo h rFu e of r the DC rtherm house p d bet h lan is ous ore hold ween th eho , s the lbe ds ine g
in Health Care Spending, by New Loan Plan Status and Age of the Participant
Craig Cope Da Data ta Definit Sour land ce is io s Director ns Per cent of Wage ealth of Bene Par fits Researc ticipant h at Hou the seho Employ lds ee Who Bene To fit Res ok a Pla earch I n L nst oan itute, (EBRI). Michael
Where Are Households Spending Their D Figure 12 efined Contribution
Percentage of Participant Households With an Increase of More Than 10 Percent
New Loan Plan Status and Household Income ............................................................................................ 13
amount spent relative to income. See Copeland, Craig, Kelly Hahn, and Matt Petersen, “Spending and Saving Behavior of
fin cons mo with dre l ing out idere ike a parti that year pr ly to d a m sany p be up cipan ior to t plem finart t a taking nciall enta icip he loa a y l p n s a p ts n lan tres ihave ncide lato t s n ed loa he s . T nce pn, D endi her a B hnd the ealth efore, pl ng an. incr c l Th are s househ o eas an ese es o pe year pnd olds that articipants n to see in hea g was lth are i c th f an are cou l eiy cat mos kely to ld whe be eg t li n takin cov or kely be y spen ere fin to h g d by h a anciall pl di ave a nn g yb seen y st loa srid hare i res n p s its lsu an e nc g d portion o ge sreas as w wsere tse that d by more i eldlf . How enti tot exa fi aed lmining ever, to
Conrath is the Chie Per f Retirem centage ent by Strat of New Par eg Mortgag ist ticip and ant Head e Hou Statu of t seho he s R and etirem lds A W ge ent ith In of an s th igIncrease hts e Par Strat ticip egy of ant Team More for Th J.P. Morgan an Asset
To To d see etermi where ne wher those e who plan l too oans wer k plan loans e spen spen t, the partici t their fun pa ds nts , each who took loan- tak a loan ing must f participir ant st b ’se ide spend ntifie ing d. is com In thi par s sam ed ipl n e1 of 2
in Housing Spending, by New Loan Plan Status and New Mortgage Status
Plan Loans: An Examination New Lo of Public an No New Loan-Sector Participants
PRRL Spendin Dag ta — ba Total sp se — The PR ending RL Data is the a base is nnual an sum of opt- the in col mlab onthly sp orationend amiong pu ng capture blic d reti thr rement ough credit an plan sponsor d debit car s, the ds,
Public-Sector Defined Contribution Plan Participants,” EBRI Issue Brief, no 12. 570 (September 19, 2022).
s the not all see pend h ifealth h i ng incr public ealth than ins -ce s urance 5 are i as ecto pe e rce s by r still d availa n efined mo tage points ar ca e bt le tegory o han contr to 5 DC plan ib . pe T utio he f rce sp n pl snt en p partici ending age points di ans off ng that pan cer a ategories ts is m , l co at oa uld ore li 2n 2.2 optio mos also percent. kely to n, t li hekely to a lp nd hi mprov ave an Thi the h s pl ave was e financ i an ncr seen fol ty eas pe lo an wed b es s e that in , s incre spe howing t y mo h nd as ous re ie ng ty in in hp amo g the e ically of ( int 18 ir ng ers sh .6 t ection pe hos are of fer rc a ee nt) who l of oan tot an al d
5 Percentage Points in Housing Share of Total Spending,
s pu Managem pe blic cific -s cate ecto ent rg . DC ories Sharon C p lan betwe partic ars en on the ipan is a Retir year be ts whereme e foa re the l nt Strate oan o loan ption gi isnc t o wid as n enc the availabl e was J.P. ex e Mor , amined gan 10.9 Ass pe an et M rcd ent th anag e toyear ok em a that ent loanReti duri the rement I loan ng the was y nsigh ear o taken. ts f The
70% Figure 4
6% New Mortgage No New Mortgage
Figure 10, Percentage of Participant Households Who Are Financially Stressed With an Increase of More Than 10
New Mortgage No New Mortgage
11
5.6%
electronic Employee Ben payment trans efit Researc 13 acti h Ions nst, Chase c itute (EBRI heck ), and s, and the cas Nati h onal acrosAs s 10 sp sociati ecifi on of c spendin Governg ment D categories: efined a Contributio pparel & se n rvices,
By Craig Copeland, Michael Conrath , and Sharon Carson
option food/bever too health k p and lan are s l pe oan a age we nd 457 plan s alth. s ing of vs. (18 those w . th 0 or a n pe is size rcen hon by o d t) wer - . ERISA iOnly h New d e not. unspecif 4 Lo ous A 01 fi an ( inanciall ng k) ied cas p Plan sp lan. en yh di stres Statu n s In pe g Figure a thi nd snd ed s sing ( study h unspec and o14 usehold 24.9 perc ,A 9 ified ge 3 pe i cash of s rcent o ent defi th o e sned as fp the Par f endin th h e ticip ous p hav gub ha eho ing lic ant d higher -s lds) outs ecto , ta health r lD nd ikelihoods o C in pl g car acr n ed e partici (it 23 c f .ard s3 p hare ants
5 Strateg 70y t % eam. This Issue Brief was written with assistance from the Institute’s research and editorial staff. Any views
first tes interestt (F was igurte o see 3 Percent ). i f The like spen ag ding e lihood of in Particip ao ny of f a p t art he 12 an icipts ant c ategor T akin takinig a p es g a L increase laoa n loa n d b n ini a nc y more tha reas Year ed , b with y n 1 In a 0 co ge percent t m hroug e an . Am h d th A ong e ge 50s the thos e wit n h a
Percent in Health Care Spending, by New Loan Plan Status and Age of Participant ........................................ 14
See 30 Co % peland, Conrath, Carson, Nobile, and Petersen (September 5, 2024) (endnote 1).
25% 62.0%
October 16, 2025 • No. 645
5.3%
Administra education, tor ent sertainme (NAGDCA) nt, . fThe ood database & bevera ge include , healt s d h ata car fro e, hm ous two inghun , trans dred portatio eighty n, -on tre avel, 457( cb), 4 haritabl 01(e a)c, 4 ontributio 01(k), an nsd , and
increases bal who were ances lar pe i for t n rce ge a nt) h r than pl os Per a , and e n with tak cent 75 in housing p g a l erc age lo oan an ent (21 vs of optio of . those the Par .0n pe ir av ticip rce cr ai wed lab ho d ntant it )le . card i d not. w Hou ere limit i seho n Ot 59.0% s therwi hos anlds d e havin s pla e,W the n ith ty g p s li es pe kelihoods o an .n tIncrease more tf ha the n of 1ch 10 perc More anges in ent Th the of an the shar ir es of net income.
new de exp cli r ess ned DCed .pl Hou in t an ls oan, he his re eholdport are income alth ca re spe thos of $30,00 e nd of the ing w 0 or authors ab as the ove mos , an did d t not slik hould appear ely to not have inc to h be as avcribe e reas an i ed d mpa to , as the ct o 58 o.n ff 5 the icers percent like , trust lih of oo ee ho d sof useholds , or other a plan spo wh loaere n nso ars
New Loan No New Loan
60%
6 26.8%
Fi DATA PR Fo gure 11, r more i IVA Per nfor centag Cmatio Y: JPMorgan n on the e of P 60.0% ar Ch ticipant PRRL as e has Da House tab a ase numbe hold and the s With r offindin sec an urity p gs f Increas rom rot the d e oco of ls M a i ore T taba n plac se se han 5 e whic e TPe he h rcentag pha aresit, des S e ig amita Po ne int d , “ s to iTh n ensure th e S e tate of Share all Pu That blic-
5%
other 403(b) .16 For cas d %efined h contr and ch ibu eck tio n ( spe DC) pl nding ans that o c f o annot ver 3 be millio cate n s gorize tate, county, d, two add city itio , and sub nal categ div ories ision g areov creat ernment em ed 54.9% : unspec ploy ifie ees d cas with h
5 Percentage Points in Housing+Cash Share of Total Spending,
Am speong nding in these ea financi ch of the ally str othe esr cate sed hgories ouseho w lds ere , those eitherwho simi al lar so took betwe a en p th lan e house loan4 were .7hold % smo or re l less ike for ly to the have h households ad a sp w end ith a ing
of EBRI, EBRI-ERF, or their staff. Neither EBRI nor EBRI-ERF lobbies or takes positions on specific policy proposals.
partici being pan taken, t too as k a the loan perce saw nta their ge taki spe ng a l nding oan on in hea eac lth h cinc are om increase e group wa by m s i ore than n the 10 10 –11 perc percent ent in rang the year e. Partic they took ipants w the ith
60% 4.6%
4.6% 52.2%
6
Sector DC P Various S lans: 2022 pending Cat ,” PRRL Research egories R Se tud present, y (Publi b c Reti y New remen Loat Research n Plan Stat La usb, ................................ December 19, 2024)................................ 15
Introduc cus tom • er d Only h ata tio is ous n kept ing s con pend fidential ing an an d unspeci d secure. fied ca Reas sonab h spele phys nding h ical, ad hele igh ctr er lik onic and elihoods of procedural share safe increas guards es for ar those e used that
and $168 c heck s billion pe in ass ndinets g. as of yby ear - New end 2022 Loan . Plan Status and Age of the Participant 20.0%
Where Are Ho 25% useholds Spending Their Defined Contribution
increase o participant tak f more than ing a plan 10 loa percent n. than those who did not take a plan loan (Figure 10). Spending increases on health
loan EBRI in the s ( mallest Fivit gues c re a 7om cco ). Thi ment on unt balanc s was fol thes had is resear lowed the bych. trav l owest like el (22.4 lih pe oo rcd ent) of takin , enterta g a lo inment (19. an, but onc 0 pe e trcent) he acco , an unt b d no alance n-spe reach cified ca ed sh
20% 14.2%
50% 14.1%
47.2%
Incomes at Median or Below 23.0%
46.5%
are desigtak ned ing to comply a loan vswith . those wh federal o stand did not. ards to Other protec wise, th t and li e likeli mit hoods of access to p the ers changes onal in iformat n the sion. hares of There spe are ndi sev ng in eral
W 7 orkers’ finances can face many challenges over their careers including irregular expenses, buying a new home, or
care were more prevalent among the financially stressed households whose plan participants were ages 50 or older, as
See L 14 ucas % , Lori, Jack VanDerhei, Kelly Hahn, Je Oh, and Livia Salonen, “The 3% Difference: What Leads to Higher
$1 sPlan Fipe g0, ure 12, nd 000 ing , th (Loans 18 Per e .li9 centag kelihood percent) e : A of of . P ta Ho ar n Ex king wever, ticipant a l the amination oan House s wa e expens s hold similar throu s With Ne es coul w Lo an an of Public d gh Increas hav bal eNo an increas New e ces of Lo of Med aore T n $ - 1 ge Se 00 n han 5 ,erally 00ct 0 or m a Pe or Pa nd rcentag ore wer , e lan e not rticipants Po din neces int g 3.8% s in in th sHo arily spe e 1 usin 1–13 g S cific hare to
4%
3.8%
50%
C Inc hase ome 40D % ata — Si — nce JPMor all th ga e n spC ehase Ba nding data nk, N. are A. (Chas at the ho e) useh serves old 66 millio level, the i n Uncom .S. househol e used in ds w thiit sh study a broa isd aran lso ge at the of f inancial
key controls each and of th poli e ot cies in her cat place egori wes wer hich ar e e either design sed imilar to ensure c betweenust the om her d ouseho ata lis ds sor afe, less secu for t re,he anho d ano usehol nymo 3.6% ds w us: (1) ith a
financing higher education. All of these events can lead workers to have to take on debt or find other sources of
63 Sp .3 ecial percent thank osf to the Mike hous Ceros holsds w andh Katy ere a a lo Cheg an was aeva taken for th had eir contr this in ibutio creas ns e to com this par pa ed p w er.ith 5 6.8 pe r cent of the .
Retirement Spending?” Employee Benefit Research Institute & J.P. Morgan Asset Management Research Collaboration
the perce loa nt ran n. of Ther Tota ge.efore, Th l Se perce pe the nding likel nta , by ihood geN who ew s of L took oan Pl the a s l a pe oan n nd Stat iincre ng incr us an ase e dd as Ag ses of ub e of stant t th he is m iall Pa yagn w rticip ith itu ant cre de di ................................ wt ere ca rd com utipare lizatio d wit n, as h 6 thos ....................... .9 e peby rcent of 16
3.4%
Incomes Above the Median
40% 15.6% 19.2% 42.6%
20% 15.5%
s household ervices incl lev uding check el. There are ing t , wo savings, in income ves valtues use ments, cre d in di thi t cards s study f , anrom d loa tns. he Ch Chas ase data e’s scale : and wide reach allows for a
By Craig Copeland, Michael Conrath, and Sharon Carson
Before J.P. partici Morgan pant tak Assing et M a apl nag anem loa ent n. (JPMAM) receives the data, all selected data 13.7%is highly aggregated and all unique
fin ancing for them. In some cases, workers’ only source of significant savings is an employment-bas18.4% ed retirement
av households ailable at where https://am.j a loa pmor n was gan n 18.0% .c ot tak om/us/en en. /asset-management/mod/insights/retirement-insights/the-3-difference-what-
households participants in who d hous id not tak eholds w e ith a loan. outstan The dinsg ame cate -credit-card gori-es amo balance ng tho -to-cre sdi e twho -car d d-ilimit d no t take a ratios of loan 0 perc were ent the too mos k a loa t lin kely to
15% 17.7%
12% 34.5% Figure 11
13.1%
40%
Fi com gure 13, preh 35%ensiv Pere centag 2.8% view e of of ho Pusehold articipant finances House. In holdth s With is analy an sIncreas is, the Chase e of M ore T data han 5 sample Pe is rest rcentag ricte e d Poto inttshe in ho Ho usehol using S ds i hare n
3%
identifiable information, including names, account numbers, addresses, dates of birth and 33.5% Social Security numbers, is
savings plan, typically a 401(k) plan or other defined contribution (DC) plans. Thus, some DC plan participants could
leads-to-higher-retirement-savings-rates/ and VanDerhei, Jack, and Kelly Hahn, “In Data There Is Truth: Understanding How
Suggested citation: Copeland, Craig, Michael Conrath, and Sharon Carson, “Where Are Households Spending Their
have h compare ad incr d witeas h 13.5 es, in petrc he ent same ord of those with er. The ratios one cate of 20 gor –5y that sho 9 percent and wed a 19 h.8 ig h per lik ercent eli o hood f thos of e w havi ith n ra g in tios creas of e 80 d – a100 mong
Percentage of Participant Households With an
Net Income — This is the observed deposited amount from the Chase data for salaries and wages that is net
12.2%
• 30%H of ous Tota ehold l Spe s n wd ho star ing, by ted Ne mo w Lrtgage oan Pl pay an Stat meu nts s an in dt Househ he year of old tIn he loan come ................................ incidence analysis wer ............................ e more likely to have 16
2021–2022 who use Chase as their primary banking institution, and their total household spending thro 12.0% ugh all payment
r emoved. (2) JPMAM has put privacy protocols in place for its researchers. Researchers are obligated to use the data
take a loan from their plan, while others could access credit outside of a plan or use some combination of 10.0% the two.
Households Actually Support Spending in Retirement,” EBRI Issue Brief, no. 531 (Employee Benefit Research Institute, June
Defined Contribution Plan Loans: An Examination of Public-Sector Participants,” EBRI Issue FigBrie uref , n 2 o. 645 (October 16,
15%
pe thos rce e nt tak . ing loans was health care spending, with 58.5 percent vs. 51.9 percent among those who did not take a loan.
Increase of More Than 5 Percentage Points in the
of any taxes and10.7% deductions taken out before the paycheck is deposited. This income source is used for
taken a plan loan than those who di A d n T ot start A mortgag G L A e paym N Cents E i n that year — 15.6 percent vs. 10.7
10%
mechanism 30% s (select credit and debit card transactions, electronic payment transactions, check and cash payments) and
30%
solely for approved research and are ob9.2% liga10.0% ted not to re-identify any individual represented in the data. (3) JPMAM does
Balanci 24, 2021) ng thes for mo e re informati decisions is on about the EBRI/ a key component JP of Morgan participa Asset M nts’ fian nancia agement res l wellbeearch ing. collaboration.
Financial Characteristics of the Sample
202 Figure 14, 5). Percentage of Participant Households With an Increase of More Than 5 Percentage Points in Housing+Cash
Otherwise, spending changes were very similar between households with or without a plan loan.
10 2% % Share That Various Spending Categories Represent,
calculating the spending ratios.
26.9%
percent. This was true for households with plan participants of all ages. Looking at this correlation in the
sources 20% of income including wage income, Social Security, annuity, pensions, etc. can be linked to the PRRL Database.
not allow the publication of any information about an individual or entity. Any data point included in an 9.4% y publication
Workers’ Share finances of Tota can face l Spend many ing, challen by New ges ov Loan P er th lan eir c Statareers us and , pot Age enti of ally the lea Pardticipant ing them ................................ to have to take on debt .............. or fin 17 d
8 Figure 3 25.2%
by New Loan Plan Status
Data privacy of customers and contractual relationships with recordkeepers have been carefully protected, and no data were
opposite direction, the percentage of those having a new mortgage given that they had taken a plan loan was
24.1% 8.7%
For more This study i nbfor uild mat s on ion prior about C work done hase, vis by it th the e Employee following we BFigure ebsite: nefit Rhtt 10 esear 23.9% ps:/c/h ww Inw.chase. stitute (EBRI) com/digital and J.P. /reso Murces organ /a As bout set -chase.
1 25 0 8% % %
C based opyright I 20 on % cust nom forer d maata tio ma n: Ty only re his report is c flect agg opyright regate ed i b nfo y rmat the A Em ion. (4) cco pu loyee nt B Ta he data Bene lancefit Res is stor ear ed ch In on satitut secur e (e EBRI) server . You an may d can copy, Pb ee rcentage
other so urces of financing to cover the various financial challenges. This study builds on prior work done by the
6.2%
Gross Income — This is an estimate based on net income described above with the addition of estimated
Percentage of Participants Taking a New Loan in a Year and the Average Balance of the Loans at
transferred to JPMorgan 6.9% Asset Management. EBRI has no access to personally identifiable information.
1
Figure 15, Percentage of Participant Households Who Took a Plan Loan, by New Mortgage Status and Age of the
5.8 pe Per rcent cent , cage ompared w of Par ith ticip 3.8 ant perc Hou ent sseho tartinglds a new Who mo A rtgage re Fin whe anciall n theyy h Stress ad not take ed* n a With plan an loan. Again,
Managem 1% ent focused on DC plan participants’ behavior when faced with irregular expenses. Specifically, changes in
access 10 ed % only under strict security procedures. Researchers are Le n ss ot tha pe n rmitt $2,00ed 0 to export the data outside of J.P. Morgan 8.8 %
print, or download this report solely for personal and noncommercial use, provided that all hard copies retain any and
Employee Benefit Research Institute (EBRI) and J.P. Morgan Asset Management foc Noused New on defined contribution (DC)
federal income and FeY de eral ar Insurance End, by Va Con rio tributions us Dem Act ogra (FICA phic ) C tax hes ar afc or tethe ris h tic ous s ehold.
5%
Data privacy is fully protected. No personally identifiable information is contained within the data and all spending
Participant ............................................................................................................................................... 18
9 a higher likelihood of Increa star se ting of a Mnor ew e mortga Than ge 10 for those Percent who ha in He d alth takeCa n a pl re aS n pending, loan was found across all ages.
credit card utilization, DC plan contributions, and/or DC plan loa $2,n 00use 0–$ 4 w ,9 ere 99 examined after these participants 7.2%
See th 20e %appendix in VanDerhei, Jack, and Kelly Hahn, “In Data There Is Truth: Understanding How Households Actually
all Chase’s copyr (JPMC ight and ) sys oth tems er a. T ppli he cable system com notices co plies ntaine with d al thl J erPMC In ein, and yo format u m ion Technolo ay cite or qu gL y Risk Mana ote sm oan all porti geons ment re of th qu e ir rep ements ort
Spending Category New Loan
plan partic 10 6% % ipants’ behavior when faced with one specific financial challenge: irregular expenses. Specifically, changes in
7.2%
7
6.3%
5%
and saving attributes analyzed in this research are kept completely anonymous.
by New Loan Plan Status and Age of Participant
Regardless of whether the household took a plan loan, $5 ,whe 000–n $ a ne 9,999w mortgage was started, the probability of 9.5%
experienc Support Spe edndi a ng in Retirement,” significant spendinEB g “spike RI Issu.e Bri ” Thief s ana , no. 531 ( lysis gEmpl oes the oyee Ben next s efit Research tep by examin Institute, Ju ing where part ne 24, 2021) for an icipants who took
for the monitoring and security of data. Apparel serv (5) ices JPMAM provides valuable insights to 3.5% polic 3.4 ym % akers, businesses, and financial
provide Credit 0% d Car that d U yo tiu lizat do so ver ion —batim a This is meas nd witu h red by th proper ci e tat rat ion. Any u io of the s revolv e beyo ing nd cred the it car scope d b oala f thnces i e fore n goin the glast requ mont ires EBRI’s h of the
Figure 16, Percentage of Participant Households Who Had a New Mortgage, by New Plan Loan Status and Age of the
credit card utilization, DC plan contributions, and/or DC plan loan use were examined after these participants
0%
Percentage Average Percentage Average
Figure 7
Ages <35 35–49 50+
having an increase in housing spending of more than$ 10 10,0 pe 00rce –$1 nt 4,9 was si 99 gnificantly higher. For example, 60.0 7.8%
a example schema DC plan loan are tic of how spendi n the overlap g th Caat sh mo ney of the . Chase data with data from an 2 EBR 4.9%I datab1 ase 7.9 % is determined.
prior exp profess 15ionals, %ress pe bu rmis t th 20s ese i sion. nsi Fogh r pts erm can iss not c ions, 30om s please e at t contact he expe EBRI at nse 40 of s cons permi umer ssions@ privacy. eb 50 ri.org s We tak . e every prec60 aution + to ensure
year to th Participa e credit nt limit ................................ on those cards. A r................................ atio of 0 percent means ................................ that the household ................................ has no revolving credit ............... card debt 18 ,
experienc 70% ed a significant spending “spike.” This analysis goes the next step to examine where participants who took a
Taking LoNe anw Lo A an mount No New Loan Taking Loan Amount
0% Account Balance
0% 4%
Percentage of Participant Households With an
$15,000–$19,999 6.0%
percent of tho 3.1%se who took a plan loan and started a new mortgage had a housing spending increase compared
Charitable 0.4% 0.5%
Source: Estimates from the PRRNe L Da w Lo taba an se and select Chase data. For more information, see the Data Source No s b New ox in tLo hea te nxt.
the confidence and security All of our account holders <35’ private information. 50+
whi 10 le 100 percent means that the household has used the full allowable cre 35–di 49t on their cre 63di .3t %card(s). The ratio used
0% 10.9%
DC p Alan ll loan are speF nd ig ing th ure 1 at money. Key findin $gs f 14,0 rom 14 this study Less incl tha ud n e $: 2 ,000 3.0% $3,350
The contribution rates use the income from the Chase data, not the PRRL dataset, so these numbers will be different from
Increase of More Than 1 $0 20 Pe ,00r 0c –e $n 49 t ,9 in 99 Various 24.7%
Fi Lo gure 17, ngitudina Percentag l Sa e mpl of Pare ticipant Constru House High ction hold Credit Car s With d Uti an lizatio Increas n e of More T Lower han 10 Credit Car Percent d Utilizati i on n Housing Spending, by
Although DC plan loans are a source of leakage from retirement savings if they are not paid back in full when a
with just 6.6 percent Ages 06 % e pa –19rtic % ipants in the database. 16.2%
1) Household Using the D e umo niqgrap ue partici hics ................................ pant/customer identifie ................................ r (not personally ide................................ ntifiable information) in eac ................................ h dataset, the ....... 7
loan vs. 51.5 percent. Yet, the other two income groups also saw significant differences between households with and
Other 6.5% 6.7%
The analys is presented here, which links public-sector DC plan data and banking data, builds on the cross-sectional
Source: Estimates from the PRRL Database and select Chase data. For more information, see the Data Sources box in the text.
fin EBRI ding that and N m AGany p DCA are articip not a a nts ffili hav ated e w spit endi h JPMorgan ng increas Ch es o ase n & hea Co. lthor c an are ywhe of its n takin affiliate g as p or lan sub loa sn idiari suges ge.s ts that examining
Conclusion
50–59 60+ Percentage of Par 9.7% ticip 20ant .9$ % 1 3Hou ,254 seholds $10W 0,0 ho 00 o Had r more a New Mortgag 11e, .8% $21,595
use of plan loan in the text refers to Charitab a ne le w plan loan, not one outstandi 1ng from 3.8% a pr 14 ior year. .9%
Figure 13
balSo anc urces of e: Estim $ ate 1s00 fro Am ,0 v erage th 00 or mor e PRRL Da A tacco be a. s eTh aunt nde perc seleBalan ct Cha ent seage who dces, ata. For m by o took re 2 i0 Cred nfo % rm – a l a 5ti9 oit oan % n, sCard ee i th ncr e Dat eas Usag a So ed urc su es e bbstant oan x in td heA iall texge t.y wi th credit card 17.3%
individuals in both sets of data are established. These individuals with both the spending and the saving data are then
13 without a plan loan. Transportation 14.2% 14.8%
PRRL Lo Th an is n Inci res umbe u de lts nc a r is e nd ................................ lo the wer than evaluat the p ion of ercenta the ................................ imp ge who had a act of fina lo ncial an f outstanding, as acto ................................ rs outside a loan outstan of the DC ................................ plading co n, suchuld ha as ovverall e been ...................... spe tak ndi en n i g l n a evel 9 s
the health insurance available to DC plan participants could also help improve finances, showing the intersection of
60+ by New Plan Loan 13.2% Status and Age of the Participant
30% Checks 13.9% 12.1%
This study builds on prior J.P. Morgan/EBRI studies that looked at the link between spend8ing, credit card debt, and DC
Percentage of Participant Households 60%W –79 ith % an Increase of More Than 6.3%
utilization, as 6.9 percent of participants in households with no outstanding credit card balances took a loan
grouped into households using Tr aChase’s vel method for determining members 5.4 of % a hous7 e.hold 3% . The unit of observation in
4, 5
and differe de nt y bt acc ear b umulation ut has stil , l no on b t been ehavior paid off. inside Th the us, some l plan to s oans a ee where re in the the funds repayment sta from the ge D but n C loans ar ot take s en ipe n th ntat y . e In ar in a this ny
Credit Card Debt/Limit Ratio
health and wealth. While not an emergency expense, another common reason for taking a plan loan is buying a home,
Contr $1ibut 40,000ion Rates — The contribution rates in this study are calculated by taking the dollar amount of the DC plan
Education 12.9% 13.4%
Lo plan an loans Incide . Th nce ese an link d Change s betwseen in Sspe pen ndi din ng g and de ................................ bt suggest that r 80................................ %– etire 100% ment planning ................................ is not wholly different .............. by place of 1 12 3 . 4%
5 Percentage Points in Housing Share of Total Spending,
compared with 19.8 H pe ou rc sient ng+ of Ca t sh hose who New Lo had an outstandi No ng New cred Lo 30 a it .n5 car % d balanc 24.4 es % equivalent to 80–100 percent
this study is the household. The number of people in these households may not truly reflect the exact household size,
Endnotes Gro 7% ss Inco me — Beginning of the Year
given year.
analysis, changes in spending amounts and the share of total spending on a specific spending category are calculated
Grosswhi Incch om is e typically a household’s largest investment. A plan loan can help reduce the borrowing costs of a home
contribution from the DC plan En data terta iin nm the ent year of the loan incidence an1 d 9.di 0% viding it 2b 0y the .2% gross household income from
Figure 9
by New Loan Plan Status and Household Income
employment, even where benefits availability may be dissimilar, but part of a broader, holistic financial planning
20%
of their credit card limits.
N asew the Mor househ tgageold s ................................ size can Hi ogh nly So Cre b urc di e t Card e: app Es Uti tim ................................ roxi ate lizmat ati s f on rom ed th based e PRR L D on the ata................................ ba n se um anbe d sr of uni elect Chqu ase d e iata n ................................ di . F vid or uals who have Ch ................... ase account s17 .
$20,000–$29,999 8.0%6.3% $6,568 0% 6.9% $122,638 $15,871
$20,0 to s 00–ee $29 if ,9a 9ny s 9 pecific spending categories such 3as .3% housing, health care, travel, or transportation increase when a loan is
pu 14 rchase or Per even cent mak age e the of home Par Food ticip /B pu ev rchas eant rage e s Hou possseho ible. lds With an Increase 3.9% of 3 More .3% Than 10 Percent
the banking data plus the DC plan contribution amount. Spending-to-Income Ratio
jo urney The media where n ho aluse l fact hol ors d inco need me us to b ed f e incor or thporated ese categori . In f es was act, partici a gross pati househ ng in aol budg d inco et me of webinar $76,has 600. T been foun he income d to incl be udes
25% more information, see the Data Sources box in the text.
6.0%
As an $1 exam 20,000ple, if only one spouse has a Chase account, this will be considered a one-person household. This
$30,000–$49,999 11.0% $8,249 >0%–19% 9.7% $13,752
1 New Loan No New Loan
$30,000–$49,999 18.2%
Lower Credit Card Utilization
Conclusion tak See en Co vs 6% peland, Craig, . prior ................................ to takin Micha g the loa el Co Hea n. nra lt................................ h th, Sh Care aron Carson, Alex ................................ Nobile, and M 5att P 8.5%etersen ................................ 5 , “ 1.Ho 9% w Financial Facto .......................... rs Outside of a 19
in 5.8% Health Care Spending, 18by New Loan <0Plan S .80 tatus and Househo 23.1% ld Income 20.4%
income from • DC p all i lanndividua participan ls in tsthe h in hous ousehol ehold d,s not j with higher cr ust the partic edipant’ it card s inco usage me. ha d lower average contribution rates across
associated with higher DC plan contributions. Programs to help with workers’ overall finances — for example, financial
$ 10 5% 0,000–$74,999 10.3% $12,610 20%–59% 13.5% $14,586
household unit observation necessitates the defining of specific data variables.
$106,640
$50,000–$74,999 27.2% 5.4%
Accou 70 nt % Balances — The aH cco ouunt balan sing ce is the end-of-year total amo 10unt .3% that th7 e .6 DC % plan participant has in their
Def ined Contribution Plan Can Impact Retirement Readiness: An Examination of Public-Sector Participants,” 21.8EB % RI Issue Brief,
0.80–0.94 18.9%
New Loan No New Loan
all ages except for those ages 60 or older. For example, for participants in their 50s, those from households
Endnotes wellness be ................................ nefits — could be indi ................................ spensable. The decisio ................................ n to take a plan loan................................ is not just dependent on what ............................ happens in 20
15
$75,000–$99,999 11.4% $15,088 60%–79% 18.6% $12,665
$75,0 Thi 0 0 T– she h study $99,ig 99 h i 9 -credi s part of t-card 20.6% a -util joint izatio effn categ ort betwee ory incl n the udes 18 Em .6 partici % ployee pant Bene s wifit Res th househ earch In olds h stitut aving e out (EBRI stan ) ding credit c and J.P. Mor ard ba gan Ass lances et of
63.7%
Other 5.3% 4.3%
no. pu blic 618 (Se -secto ptember 5, 2024) an r DC plan(s) in the d year Copeland, of the l Craig, oan inci Micha denc el Co e. nra 0th, and Sharon Cars .95–1.04 on, “How Financial Factors Outside of a 11.5%
$100,000
5%
with higher credit card utilization had an average contribution rate of 4.7 percent, compared with 5.6 percent
the 2) In plor ande but r to on ensure t the total hat financial the dataprof sampl ile of e only the part inclu icipa desnt. households where the Chase data have all or the majority of
$100,000 or more 11.4% $18,086 80%–100% 19.8% $12,006
$100,000 o20 r m %ore 32.7%
more than 0% 50 percent of their credit card limits, while the lower-credit-card-utilization category includes participants in
Management to deliver data-driven Transp researc ortation h to better u 19.1% 59.5% nderstand how 9 the .7%financial f 11acto .3% rs faced by defined
1.05–1.49 30.2%
401(k) Plan Ca n Impact Retirement Readiness,” EBRI Issue Brief, no. 591 ( 18.8% Employee Benefit Research Institute, September 7,
It is not sur prising that the average loan balance is lower for those in households with higher credit card ratios, as
60% 18.4%
57.5%
for those with lower credit card utilization. As a result, participants in households with high credit card
Source: Estimates from the PRR AllL Database and select Chase data. For m <50 ore information, see the Data Sources box 50 in + the text.
their spending, filters are applied to the households to meet the full (majority) spending criteria. These filters include
Source: Estimates from the PRRL Database and select Chase data.
househ Spen din olds h g Ra aving outstan tio — This din is tg credit c he Traratio vel of ard ba total lances annual of s 50 p penercent ding to a or lnnu ess a o2 lf th n 2.et 4eir credi % incom 4.2% e t card li 2 of 3.0 the % mit year s. Th ofe co the ntributio loan inc n ra idetes are nce.
Potential contributio age n pla and n pa income rticipant diff s er outs ences a ide of re th again ex ese plan amine s impd act bethe tween ir retir tho ement se with prepara and wi tions thout p . Thlan loa us, the ns. aim The is perce to prontage vide
1.50–1.99 10.2%
2023).
participants in households with higher credit card utilization are more likely to have less in their plans available to take
Ho Figures wever, t he prior studies did not identify what specific expenses were linked to the plan loans. In this study, spending
3.9%
utilization had lower average account balances.
but are not limited to 52.0% : all 12 months 3.8% of spending data, households with spending more than 50 percent of their
For more inform 4% ation, see the Data Sources box in the text. 51.5%
$80,000
calculated using the contribution amounts from the PRRL Databa2 se divided .00 or morby th e e gross household income from the Chase
8.8%
of unihouse que, fact hold -b s as with ed ins a lar ights ger t to he han Sourc l5 p e: pe poli Esrce tim cy ate nt mage point akers s from th , p e P la in RR ncr sL D peas ons ata e ors ba in se , the an and d s p pla o ele rtion of ct n Ch provi as the e d de ata rs ir . F tb ota uil or d l a spe sn tro din nge g goin r retirem g to housi ent sy ng stem .
Source: Estimates from the PRRL Database and select Chase data. For more information, see the Data Sources box in the text.
as a loan. Participants in households with higher credit card usage have lower average contribution rates (the
2
increases among households with participants taking a plan loan showed that spending changes were across many
Figure 1, Demographic Characteristics of the Sample ............................................................................................... $70,419 8
For example, 50% see Holden, Sarah, and Jack VanDerhei, “Contribution Behavior of 401(k) Plan Participants.” EBRI Issue Brief
3.4%
In estall imated gro age *Fi gro nanup csia sl i ls y exce ncom stressed pt e, hofor uan sehd the oldho s h 4 useholds av0s, e outsparti tandi nwith g cipan credit c cts red ard with bit car alan a cesbove d larg sp er end - thme an i7ng ou $68,294 di 5 pa en hous rcents t oid f th e ehold inc eiof r cre Ch $68,428 dit cas arde om li of mitses wer aless nd sp than ene d m m o 30 ore re th a p nlike ercent ly to of have taken their
more 46.8% information, see the Data Sources box in the text.
dataset plus the contribution amount.
New being higher Mort for the gages hous eholds with participants who took a loan was more pronounced among households with
15%
percentage 110 of pe rctot ent al of th gros eir nes t i h ncoo musehold e. income that contributions to the DC plan represent) across all ages except for
• Among those with a new DC plan loan, health care spending was the most likely to have increased, as 58.5
spending categories, depending on the households’ circumstances. Yet, health care and housing spending, particularly
9 14
no. 238 (October 2001). Available at https://www.ebri.org/publications/research-publications/issue-
a l overall oan than spendi pang rticipants . from households with incomes at or below the median (Figure 4). For example, among those
Specifically, public-sector DC plan participants who had taken a new plan loan are compared with those who did not
Household Demographics Contribution Rate
Figure 2, 3% Financial Characteristics of 15 the Sample ...................................................................................................... 8
partici 16 pants younger than 35, at 26.8 percent vs. 18.0 percent for those who did not take a plan loan (Figure 12).
An important component of housing expenses is mortgage payments; as a result, starting a mortgage is likely to have a
thos The ere $6 ag 0,0 h es 6 00 ave 0be or en s olde urv r e (y da Figuta re 5 on what ). For partici example, for pants report parti as cipan the rea ts in th sons eifr or t 50s ak , t ing hos plan l e from oans higher or wi-thdra crediwals. t-car Fo d-utili r exa zample, tion
percent of households where a participant took a loan saw their spending on health care increase by more
among the households starting a new mortgage, stood out as places where spending increases differed from the
briefs/content/full/contribution-behavior-of-401(k)-plan-participants-154; Munnell, Alicia H., Annika Sundén, and Catherine
ages 50–59, 14.2 percent of participants in households with incom 0% es above the median took a loan, compared with 10.6%
take a40 new % plan loan to see which spending categories experienced increases among those who took a plan loan that
Ho The useholds househo wi ldth p partici articip pan atn sts were ages wi 35 de–ly 49 dis wtributed ho took a acro loan ss ag did n es ot an h dave incom a hes igh (er Fi perce gure 1ntage ). For e ofxam an pl ine, creas 8.9e p derce share nt of than
significant impact on housing expenses. A possible source of funds for a down payment when purchasing a house is a
the EBRI/ households Gree ha nwald Reti d an avera remen ge contr t Coib nfutio idence n ra Ste urv of e y4.7 foun perc d tha ent t 35 percen , comparet dof those say with 5.6 pe ing rcent they for to those ok a lo with lower an or withdrawal from credit card
than 10 percent in the year they took the loan. This was followed by travel (22.4 percent), entertainment (19.0
households where a DC plan loan was not taken. Thus, the loan usage does not appear to be tied to spending on
3) Once Fi Tagylo ure 3, r, “W the Perc hast Determines eentag househol e of P ds a 40 art1(k) re identi icip Pa ants rtic fied, Takin ipatia on an g t le a New as d Co t on Loan ntributio e of t in he a ns?” CRR W Ye indar ivi and dualsorki the in ng P tAv he hous era aper, no. ge B ehalance old 2000 mus of -12. t t also he L Chestn be oans aut Hil public at Yea l, MA: -s r eEnd, ctor by
>0%–2% 14.5%
13.7 percent of the participants in households with incomes at or below the median. The likelihoods among participants
are different from those who did not take a loan. First, th$43,746 e sample of households is described, and the likelihood of DC
the households had a DC plan participant younger than age 30, 26.6 percent ages 40–49, and 13.2 percent ages 60 or
the households who did not take a loan, but for the households with the oldest participants, those who took a loan had
DC their retirement sa plan 2% loan. In fact, vings p house lan bought a holds who home star, c ted ar, o mo r o rtgage ther large p payments urch in t ase.he Another 27 year of the pe l rcent u oan inci sed dethe nce m to analys payis off credit were more
utilization. This, in fact, leads to lower average balances among those with high credit card utilization across all age
10%percent), and non-specified cash spending (18.9 percent). Comparing the spending increases by categories
Center f DC luxury plan items or R partic Various etirement Resea biut those pan Demo t whe grap invol re a rc hving lo ic h a an Ct Bo hara their is avai sto cteris h n College, December ea lab ltics th le care in ................................ the or i ir pl nve ans.ting T 2000 he demogra in . Avail a hom ................................ abe. le at ph Th ic an is sup https:// d fipor nan crr ts cial .bc the .edu/w ................................ cha ide ract ap tha -erit stics prohi ofbi the ting p .......... pla ersn on 9
>2%–5% 23.9%
$40,000
in their 40s saw participants in at- or below-median-income households at 14.1 percent taking a loan vs. 13.1 percent
plan loan incidence across various demographic characteristics is shown. For the households where the DC plan
30%
older a signi . For ficant gro ly shig s hh ous er ehold percent incom age o es f hav , 3.3in p g had ercenta n of increas the hous e in ehold their s s ha har d e inof com spendin es of $2 g on 0,000 housin –$29 g. ,99 9, and 32.7 percent
likely The card debt, 22 percen reto we have re ag taken e and t to ai p ncom co lan ver e loan di day ffere tha -ton nces -dthos ay ex in t e pens who did aking loan es, 23 per nost star scent to pay f hown t mort previo gage or ho usly payments that may me or car repairs, and 21 in t hahve imp at year act — ed 1 p 5 the .ercent to co 6 p sercent pending vs. ver c hange 10 .7
groups (Figure 6). Looking at participants in the $31,184 ir 40s again, for example, the average account balance was $68,294 for
with those who did not take a loan, only health care spending showed a higher likelihood of having increased
loans wou content/uplo ld ads/2000 not necess /12/w arily p_200 improve 0-12.pdf partic ; aip nd US ants Gene ’ retirem ral ent Acco sun ecur ting Office ity, as the , “401(k) Pens loan usageio is mor n Plans: L e like oan ly to Provisio help wi ns th
identified as the DC plan participant in the household are thos >e us 5%–ed 10 in the % analysis. 26.4%
A spending increase may not have surpassed the 10 percent threshold, but it may have altered the composition of
of participants in households with incomes above the median.
participant has taken a plan loan, any significant increases in spending in each of 12 categories or increases in the
Figure 4, Percentage 17of Participants Taking a Loan in a Year, by Income and Age ..................................................... 10
had incomes of $100,000 or more.
medical expenses.
res perce ultnt (Fi s. When gur e look 15). ing at Th is wa the age s true for of the ho household usehold p s with p articipan lan partici t, only tho pan ste s of in al the l a tge wo s. older Look ag ing e at gro thi ups showe s correlation d a ih nig ther he
those from lower-credit-card-utilization households, compared with $43,746 for those in high-credit-card-utilization
Enhance P by artici mopatio re than n bu 10 t Ma pe yrc Aff ent ect Income Secu among those taki rity for Some. ng loans. Othe ” G Letter Report reatrwise, sp er Than 1, G endin 0%AO/ g HEHS changes -98-5 (O wercto e very ber simi 1997lar ). Washington, between2 4.5%
expens1% es in retirement — health and homes. Without the option of taking a plan loan, participants would seek loans
spending among the households. To d $2et 0,8er 04mine the second test of spending changes, the composition of total spending
s By househo hare of total ld isncom pendie, th ng that e ho thes useholds e categ with ories the repr lowest esent inc are determ omes (less ined tha , i n nclu $60di ,0 n00 g for ) wh certain o had taken demo a loan graphic were more
$20,000
4) pe opp rce Sinc osntage 20 ite e % d thir e wi ec sttion, th atus a great of the ma perce $15,781 er th ny of ntage an the 10 perc varia of those ent bles increase must be having a in he n kno ewwn alth mortga at car the e ge begi s pgi en ven di nni nthat g n g amo an the d ng tyhe had thos end of e tak who en the took a p sla tudy n a l loan oan year was com , the 5 par s .8 e ed
households.
17 DC: US Gen households eral Acco w uniting Office. th or without Ava a p ilable at lan loa ww n.w. gao.gov/assets/hehs-98-5.pdf.
outside the plan to fill spending gaps, and those loans may have terms more expensive than those of a plan loan. Yet,
The Figure 5, averaAv geer oage Contr verall outst ibut anidi on R ng loa ates n , b by alaC nce rediat t Card Us the endage of tahe nd Age year w ................................ hen the loan was tak................................ en was $14,014 (Figure 3) ....... 11 .
that Th eac e house h cat hol egory re ds whopresent were identified ed was com as starting pared bet new mortga ween the ges S year o uwere determined by rce p:r E ior t stim o the ates l froan om t i findi hnci e Pde R ng th R nc L e Dose makin a an tad batse he l an oan y g mort d sele ear ct gage C to ha see se da if ta. For
5%
As likely charact far to as erist have finics an had . T cial he a facto n spe increas rs nd , i8. ng 8 ed cate p shar ercent gory e r of elati inc treas he ve to th ho esusehol are e the ho duseholds partici n compan par witho ed ts with thos had ut a plan DC ple an l of oan ac ho count (useholds Figure 13) balanc whos . eTh s of e e les d DC ifferenc p s than articip e $2 did ant d ,0not 00 id ,
$9,746
with households perce tnt hos , com e wh m pust o d ared with h id ave not take t wo 3.8c pe ontiguous a lrce oan nt (Figure start yea inrs g 8). i an n For th ew e es mo xam amp rtple, amo gag le to e when be i nncl g th part uded. ey icip haThu ants d not s , ages tak each en 50 house a or pla older, hold n loan hav 6 (Figure 2.0 ing pe com rc16 ent pl ). et oAe fgai DC n, a
more information, see the Data Sources box in the text.
having For incom liqu eisd aand cco acco untsunt ba , such lances as health , the savi average ngs accou outst nts an (HSA ding loa s) n and bala emerge nce increas ncy sav edin as gs accounts incomes a n th d at acco can unt provi balde ances funds
3 pay any ments in t spending hcategory e loan inci ’sdence a share ina ncr lysis who eased by ha more tha dn’t beenn m 5 ak pe ing t rcenta hesge e p p aymen oints. ts in the The speyea ndin r prio g cat r. eg In other wo ories mosrds, a t likely ny to
0%
and Fo r the most r 7.2 percent ecent res had balults, ances of see Th $2 eph ,00asit, 0–$4 S,9 amita 99, while , “The St 19. ate of P 3 percent ublic h S ad ecto bal r DC P ances of lans: $1 2022 00,00 ,” 0 PRRL or m Rore ( eseaFi rch S gure tud 2). y, Over
app not take ear toa carr plan y over to t loan to se he hous e where ehold the sho wu ith sehold incom s with es ofDC $60, plan 000 loans –$100 di ,000 ffer.. Finally, the effects of starting a new
Figure 6, Average Account Balances, by Credit Card Usage and Age ......................................................................... 11
The link between credit card usage, plan loans, and plan outcomes is critical to understanding the connection between
• Loan usage increased among those with higher credit card utilization, which would be an indicator of
pl hig households an h an er ld ik cre eliho w dit iod th a partici card of s data tartin pan in g a ne 20 t thes 21 w a e a mo nd 202 ge rtg sa tge 2 aking a loa is for those an observ n who ha hation f ad their d ortak thi healt esn a pl a hnalys care an is sloa .pe Thi n wa ndi s r nesults g s in foun creas id n acro 17 e by ,720 ss this amo al house l ages. hold unt ,
All <35 35–49 50+
10%
for health care or housing, could help limit DC plan participants’ need to tap their retirement savings accounts when
househ increase old d who . The w avera as foun ge out d to snot be mak tanding loa ing n bal them in th ance increa e prisor year ed with but ma age through the king them in th 40se an befor alysis ye e decliar n were identified as ing for those in their
have seen $0 an increase in their share of total spending of this size were unspecified cash spending (24.9 percent of the
(PRRL December 19, 2024).
mo twortgage -fifths on (46. DC 8 pp erc lan e l nt oans ) of tand hese ho chanusehol ges in housin ds had g nos pe rev nd olving ing arcred e inve it csard de tigated bt . at the beginning of the year, while 13.4
overall finances and retirement plan accumulations. Prior research in this collaboration as well as this research reinforce
households being more likely to be financially stressed. Spending increases on health care were more prevalent
observ com pare ations d wit . h 54.9 pe 20rc s ent of the househ 30 olds w s ithout a partic 40 ip sant taking a loan. 50 For thos s e households 60+ with
Source: Estimates from the PRRL Database and select Chase data. For more information, see the Data Sources box in the text.
Fi faced 50 hagving a new sure 7, or with ages Perc he 60 mort alth entag or o gage. ev ld e ents er of . P or inves Tart he icip avera ant ting House ge i out n/re shold tandin pairin s With g g l their oa an n balanc ho Increas mes.e e ge on f er Mally ore T decreas han 10 e P d as cre ercent i din Vario t card u us Spendi tilization i ng ncreased,
households) When the ho , uhealth sing a nd care cas (2 h c 3.3 perc ategori ent), es are and combin housin ed g , th (21 e .0 like percent lihood) of (Fi a combi gure 11). ned sh Given th are inc eir reas large e b r likeli eing higher hoods o a f share mong
percent were using 80–100 percent of their credit card limit. Nearly one-half (49.3 percent) of the households had
Note: The households who were identified as starting new mortgages were determined by finding those making mortgage payments in the loan incidence
4 0%
Regardless of whether the household took a plan loan, when a new mortgage was started, the probability of having an
the importance of this relationship, but going the next step to find out what the plan loans appear to be used for has
In a prio amo r stud ngy u the sing the financially st PRRL D res atab sed as e a househol nd Chase ds w ban hos kie ng d plan p ata, articipants spending ratio were ages s of the 50 or public -older, as sector DC plan 63.3 perce partici nt of pants
participants younger than 35, the percentages between households with and without a plan loan were not significantly
Source: Estimates from the PRRL Database and select Chase data. For more information, see the Data Sources box in the text.
analysis who hadn’t been making these payments in the year prior. In other words, any household who was found to not be making them in the prior
Categories, by New Loan Plan Status ......................................................................................................... 12
going from $15,871 amo i$1 n 0ho 0,0useholds 00 with ratios
increases the househol 0%, hod us s w ing ith a a plan l nd cash were oan was ex seen amine acro d tog ss ether each a to ge de term grouip ne if , with the the re lwas argest a rel dif ations ferenhip ces amo betwng een the hous these eholds
18
spending ratios of 1.05 or more, while two-fifths (39.2 16 percent) had ratios of less than 0.95. Furthermore, nearly one-
See “FiN eld o ote: Tf D he hreams igh-credi? Meas t-card-utiliz uring th ation catege Impa ory includct o es paf Finan rticipants w ci ith al Wel househlbei olds ng I having niti outs atives on 401 tanding credit card(k ba) Plan Uti lances of more liza thtion,” an 50 perc EB enRI Issu t e Brief,
increase in ho year buusing t making s thpe emn indi thn e g anaof lys imo s yeare r w tha ere idn enti1 fi0 perce ed as havinnt g a nwas sign ew mortgageifi . cantly higher (Figure 17). For example, 60.0 percent
not been as well established with administrative data. An investigation into where the plan loan funds are being spent
who also had a DB the househol plan were ds wher co e mpared a loan wi was th th taken ose w had th ho onl is incr y had eas a DC pla e comn par . Aled so, with 56 the spend .8 p ing ratio ercent of s were co the hous mpared ehold among s where
This research, like prior J.P. <$60,00 Mor 0 gan/EBRI studies, found $60,000 that –higher $100,000debt can have a long-last >$1 in0 g impact o 0,000 n retirement
different.
of their credit card limits, while the lower credit card utilization category includes participants in households having outstanding credit card balances of
of 80–100 percent.
with categori the es yo , an unge d tog st an ether d ol the dest p y we artici re mo panre ts l ( ikely to hav Figure 14).e Th a combin e differenc ed sha e inre i the ncr likeli eashood e of m ofore than a share i 5 nc preas ercenta e wge as lpoi arg nts est at
half (49.1 per Socent) urce: Es tiof ma the tes fro part m the icip PRRL ants Data i ban set ahese ho nd select Cha usehol se datads . For ha mod re iem nform pl atioy on, ee see contr the Data ib So ut uio rcen s brates ox in th eo tef 5 xt. percent or less, and
no. 554 (Employee Benefit Research Institute, March 10, 2022).
Figure 8, 50 Perc perceentag nt or lese s oof f th P eir art credicip it card ant lim House its. holds With an Increase of More Than 10 Percent in Health Care Spending, by
of those who took a plan and started a new mortgage had a housing spending increase, compared with just 6.6 percent
is d those wi evelope th D d i B pl n an the s by ne their l xt section. evel o f tenure. DC-plan-only participants were found to spend less relative their income, on
a loan was not taken.
security, since higher credit card utilization is correlated with lower DC plan contributions and account balances. Thus,
10
amo 30.5ng percent the ho , useholds more than with in either the yo of untge hes c t ategories participan in tsdi , a vid t more th ually. an 10 percentage points.
Source: Estimates from the PRRL Database and select Chase data. For more information, see the Data Sources box in the text.
24.5 percent New had contr Loan Plan ibStat ution us rates and Age great of er than the Par 10 perc ticipantent. ................................ ........................................................ 13
of those without a new mortgage but having a new plan loan.
average, than those with a DB plan, while the tenure of the DB plan participants did not appear to have an impact on the
the availability of emergency savings to help cover expenses can be a critical factor in preventing or stalling a cycle of
ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org Is Is Is Is Is Is Is Is Is Is Is Is Is Is Is Is Is Is Is Iss s s s s s s s s s s s s s s s s s s sue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue Bri ri ri ri ri ri ri ri ri ri ri ri ri ri ri ri ri ri ri riA ef ef ef ef ef ef ef ef ef ef ef ef ef ef ef ef ef ef ef ef researc • • • • • • • • • • • • • • • • • • • • O O O O O O O O O O O O O O O O O O O Oc c c c c c c c c c c c c c c c c c c ctober tober tober tober tober tober tober tober tober tober tober tober tober tober tober tober tober tober tober tober h rep 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, ort 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • 2025 • from the E No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 No. 645 BRI Educ ation and Research Fund © 2025 Employee Benefit Research Institute 11 10 16 13 14 15 17 18 19 21 20 12 8 3 5 2 9 7 6 4
