For employers looking to drive more workers to sign up for HSA-eligible health plans, one of the biggest financial incentives they can offer is to reduce or even eliminate the annual premiums, according to new findings by EBRI.
This study examines whether offering a health savings account (HSA)-eligible health plan for free, alongside other health plan options with a premium, alters employee enrollment choices; and if responders differ by health status. The data for this study come from two large employers and cover the years 2011 to 2014, spanning the 2013 intervention when one of the two employers eliminated employee premiums for the HSA-eligible health plan.
Here are the key insights:
Health insurance premiums are a major driver of plan choice. After eliminating employee premiums for all coverage tiers, HSA-eligible health plan enrollment increased from 4 percent to 25 percent among individuals with employee-only coverage and from 3 percent to 28 percent among individuals with family coverage.
Healthier-than-average employees are enticed by $0 premiums for HSA-eligible health plans. Offering coverage with no payroll deduction attracted individual enrollees who were marginally healthier than those who would have enrolled without this financial incentive in place, therefore adverse selection was not mitigated as anticipated.
C a E for ob A m L H Eo ls m ut e ndno a ppendi st o r nu Sa p le e zli rloy le m u rss nge sc , c ss pD e to, rloy i ipts rv a t li r s ha e v , k Sh e e id e Re p x e s ly a w ulm s e — v M., h a g e e tina o r a B e ot ona nd ra r nin ac e . su he nd spond "Fr rb L r g y k e seq Sa og td tm gr ools ’s P is rne o a ue e o r th J e a o ut rn und on CD t s he ( ttte . ha 2010) he ly C Re ( y onsu C n $ p inc c OP b ublish ould e70 e rm D e . nt x ) ,000 e " a , P ius re v m s I a a d end y e ine ing in a b n H a ty o d C d y P e e ha e t for ov r nr s spe p a yr r r ollin e e g . tH r ss o c e N e ia 70 r on ion) Of He ae g l iss lt d 0 ( -h in t uc uni , Fi u I m a e ns g a e he on nd ost u lt or ur of r h e H e fiv la e y C H SA m 2 nc lim a e sm ep r a e -e lm e ina oy lt : ." a ligib h Se e The lle e a t H e e s ra s e ) ur le r r rTr ie v v e sk he a m xpe is of c a r e p d d sele as Re loy e lt rB - ie h p he us O e c nc ff tr sea a ine ion s la e lb td n th e h ss r. — c tta serv w I he h ta n e Re e lt offe in A e houg c v in C nv c ha ie ew r ugust ir s ut nge eom on d h si 80 a ili m pn in 199 z e m ( e 20 a 7) tH nt il it tion (t a SA 04 ion a : s w r44 . - 5, w to e C he -ligib 55 nd D he the rH hil e ( P A nu le 20 a e m s w dn v m he uni a 02 e ib n e ra ) se e r on lt .e r h of Paul Fronstin is director of the Health Education and Research Program at the Employee Benefit Research Institute The Impact of Offering Free Coverage on Enrollment Choice Figure 1 Variable Means at Baseline (2011), by Group, Coverage Type e e inpa r p m e m m la od la n p ptt Se loy loy e ie ie vl, e it nt le e e ly he the e r c ho ts w ne r w ion." se a a spit w s t nt ee ra s t ff e he a Qua t e l d not to c he only ta c s w r on y ta te s, e im ff re tp le inu y re la e c m , Jour t n o e sm e a ed nd roffe g p a unt n e tll in m a ion o tnc he l ril 199 ing of yr e dE r a e t w c he a g p 6, w ono s a nit as not ir rat u m e m c hi d m hoi e ic e c nt p .h a s a loy c 1 e lo v13 llow is e ta it e m k s, p s a ( now ong e 2) d : c rtn a i4 h h m o 33 oic eta he b rr-e e out y 4r sea 66 of cp a la t rhe rhe (e c ns 19 he p a m . 98 hy ltr. Us h p s t s ) In Se .ic ing o la ia look n n vi d s, ins pa tt es a a m it tfr s, sp t be a om e ar td r20 e e 20 of, a c03 tia m 05 or lis , e, tnt the in a tp he g hy MMA h r d oup y si d ec it aiion to, a r (n vi b ta he ed for si not 19 te rs, a e 95 a dy ny u e nd non c t ing of be - tehe n (EBRI). M. Christopher Roebuck is president and CEO of RxEconomics, LLC. This Issue Brief was written with As mentioned above, there are two types of CDHPs: HSA-eligible health plans and health plans that include an Figure 4 reports the results from the multivariate regression models of HSA-eligible health plan entry. In 2013, the Percentage of Employers Offering Consumer-directed Health Plan, and Risk Selection in an HSA-eligible Health Plan 1 Jordan, David William, and J. James Cotter. "Association Between Employee Earnings and Consumer-Directed Health uni e p p Hm a rSA e ss on p sc -loy ee rd ip ligib w e a t or e ion d ut k le p hor er r e he s) m rizi ug a ium ang ltnd h p fils t t ls a he la ) o . cn ont Se us e s w nc e pr a o ol e of rur r a e g tH a r e offe g oup SA p e rs, H op r e ( SA td te hus he , ns -te he it 19 ligib ty he y 96 sc found r le e ouni rhe w e e on a m rlt te od ha h p w no e tor ls la w H k n e he e w SA re s) n usin nr r -e e . ollm ligib The estg im e le st nt m ahe ud e , te a ta y d he su lt found for h p r ye c s of e la ould m ns e p he v lo in e inc ide a ye ltrnc e xi h r es w a st ese is e of k it nc , th e he a ful ed a ir v m l- te r H p rtese loy he S pA la sele e tc cim eon e-m only e ct.e tr io ibu ntn a: e tnd ion. nr Young oll fae m eily e s r Employee-only Coverage Family Coverage int Se ea g ess rFi ait st geur a dnc ehe 3 ea in flt rom h r http e E i:m B //RI b w ur w ’s r se we .m m sea e ent rc re c a h r.rc ra o and m/ nge ne em dw it eor snt ro ia o (l H m/ st RA a na ff )tional .s. A ny -s urve viewys e -ofxpress -employ ed e rin t -spo his ns o re re pdor -he t a arlth e -tp ho lase ns-of 2016. thehtml autho rs and increase in HSA plan enrollment due to the financial incentive was 21 percentage points for those with employee-only by Firm Size, 2010–2016, With Projections Through 2019 D The eyo, P re la su RA n C lt , s fo hoic DC rC eindiv he s." rk Jour idu in, a a na ls nd l o wMA if th e H e Ca m iol. ltp hloy c "a Ar e dee a -p Ma only ting na g c aov e C m e lin e ra nt ic ga e 61 l C w om ( e6 re )o : m r42 bost ili 0t-y ly 43 I nde m 5 (2 irrx for or 016 ed) .Use in t he w itm h I od Ce D ls- 9 for -C M indiv Adm idu ini as lst rw atit iv h f e amily 100% By Pa ul Fronstin, Ph.D. & M. Christopher Roebuck, Ph.D. c W e wov m e or rp e e k loy r e a le rge s w ss ee; s w he a itnd h c aelt rhr hy in t e found oni the hc a n vol la cond ttto erb un i tc eia ons a tm a se, ror y t e end he nr lik olle the ehose lya e lt ts. o h se w swit ho rvcus ih to le cees he utili a ss lt z h ca a ctost ion rle ym p serv e la ans su ic re te ha s m s w n o a ey rlde e fin arg d e g m e rnha ep gloy atnc e ed e e d sa. d H Thi dSA itiv s r c eont e lysu for rlibut te a dl l fa io in a ns m 16 m ilyo p re e r- should not be ascribed to the officers, trustees, or other sponsors of EBRI, Employee Benefit Research Institute- 2 Treatment Comparison Treatment Comparison coverage and 25 percentage points for individuals with family coverage. The comparable effect in 2014 was identical More detailed background information about HSA-eligible health plans and health reimbursement arrangements can be July 12, 2017 • No. 435 The cove D fiv ra ag e ta e st b . a The udie ses." s fr fin Jour aonc mi n a A a l in ugus l ocf C ent t lin 20 ivic e 04 ainduc l Epid were eed m b H aiolog sed SA- e y on ligib 45 sa le (m 6 ) p he : le 61 as of lt3- h p 61 cla onv 9 (1 n e enr nie 99oll 2) nc m .e e a nt nd by only fam on iliees w hait dh low priore y r eC aha r crla lsion ms data. All HSA-Eligible Health Plans and Health Savings Accounts (HSAs) a c m e te tnt rm a c p btr e iv e re m s. Mo tiu hm an a r e inc ov rre e ed ra, u se tche tin t ion in fahe m ily p high r e cm ov -ium c eost ras, g p ea la -spe s H n in S cifi A 19 cc 96 on mod . trN ibu eon l a t- ilons c r so inc ando am lude n b de is d e us tn he e rollm d num dir ee nb ctte lcr yon of to tin p re e ue d op uc dle . e W in t out ithin he -of t hous -h proc eek e e y hold te a cos r, s, t t asnd . he P t la he n Group Group Group Group Education and Research Fund (EBRI-ERF), or their staffs. Neither EBRI nor EBRI-ERF lobbies or takes positions on (21 percentage points higher) among employee-only enrollees, but was somewhat lower (21 versus 25 percentage Number of Employees 88% Lave, Judith R., Aiju Men, Brian T. Day, Wei Wang, and Yuting Zhang. "Employee Choice of a High-Deductible Health found in the appendix. 87% of C La om v the eo , rse ebtidit a pl. ay p( e I 20 nde rs11 ex sc xa ) em xa or ine m es ine d asett n dd d ings w low ata efrr om p he re r v e nin a H le e R nc A em s w e p of loy er hy ee p roffe s in erte r e w ns de ion, st along ern P dia sib e de nns et e m s, a y or lve a nd nia tra a d s tit ha thm ion t s a at/C l he arOP teadD lt h p offe . The la ring ns int , high su era cc h a - tiv de s ed he e uc ffa teibl lt ch te s of 90% Where the world turns for the facts (n=11,711) (n=4,774 on U.S. employee benefits. ) % Bias (n=37,129) (n=14,555) % Bias C sponsor hiha ghr-ls con ost s C m pom la ay n w oarls baidit o s no consi y Ilong nde dex w re ra c offe a cs d ele re e ra d riv t ing b ee dc ta ahe us s tir e he c of ont m aa d rxi ibut ve m rse um ions sele sc tor o ct teion. he in t H he SA . fa Lm ess ily . he althy enrollees may 85% be more likely to specific policy proposals. EBRI invites comment on this research. HSA-eligible health plans were created under the Medicare Prescription Drug, Improvement, and Modernization Act of 10-499 500+ 5,000+ points) among those with family coverage (i.e., indicating some disenrollment from 2013 to 2014). EmpP loy lan a ee B creoss nefit Mul Re tiple sea rE ch I mpns loy tie tut rs." e. H Ce on alsu th Se mer r- vD icre iv s Re en H sea earlt ch B h 46 ene (1) fit: s: A 138 C -15 ont 4 (2 inu01 ing 1)E.v olution? Edited by Paul 3 m In he pra ior ain t ltr M t h p he e o ana lduc e a la ln tn h se cs (H et or ion rD v gic H ae ni Ps ut s za ) tin 200 ions ilizat io (H 6n w . MOs) The ere p a m a nd pixe ep r r d 56% d e id in sign fe rnot red spe p arnd ov cifi ide 58% vc ea rrlly y or sm m gae ani n ll in m za tion w tions -3 a.g he 5 nit (P the P ude Os) r t— he . e L spe o p79% la Sns a ciss a lly w o, e g r Ric e iv H e en th , SA 80% et- e a aligib tl. t( he 2le 00 se , 4) bm ut - 1 e r .e a 9 ins p su or tretea es d d e nr Seoll in e Figur the e 6 Hin SAhttp -eligib ://w lew he w.a m lte h p rcela r.n co if t m/he ney w s aro reo le m/ ss na ctional once- rs ne urve d ayb -of out -e m not ploy ye etr 80% -h sa pv oing nsoreenoug d-hea h mone lth-plans y -in t 2016. heir html HSA to pay 2003 (MMA). To satisfy the HSA-eligible health plan requirements of the MMA, the plan must have an annual deductible Fronstin. Washington, DC: Employee Benefit Research Institute (2002). Accessed June 2015. Age (years) 44.0 43.7 2.7 47.2 47.4 -1.9 on T w A us he s m eersi d e p t et e c rnt e e onst v rd io m is icne is su t tre suc d d t h tc a p e a h a b rd tob ov w a s “ a s t e eb , rte ili he a the x tcy ond a su rof e d m vis uc b as for nt e st ting e a rd ong g e a in 200 in t d ll f e H he a vD m ide H 3. Fou il st P y nc udy g ” m e ae nd of r m r cb “ oup a o q e d m ua rvs. p e a lif a rt se ni ie Ee d m sele s w H pD loy e cH r tion a e e P rsint ,A ” e r b ss r e ot v p oie h o r ce ia w sen te f w edd t for s t hic wit he h im th C he p D rst p oH ludy, w p ie P ed sns . tha Re ityhic t sea stche h found or rc e he h ha aa nd lts c h p tis ha onsi us la t n highe es w st d ein t nt er rly e he 80% Consumer-directed health plans (CDHPs), the combination of health insurance coverage with a high deductible and an Retirement and health benefits are at the heart of workers’, employers’, and our nation’s for health care before they have had a chance to build up Fiana n a lly c, coun Figur t e b a4 d lane cm e. onst rates that the unde 75%rlying secular Figure 4 of 4 at least $1,300 for employee-only coverage and $2,600 for family coverage in 2017, and the plan’s out-of-pocket Lo Sa Coss po, yrig Ant ht In hony form T., Mona ation Sha : Thi h, s r and epor Bia t n isc a cop K. yFr right ogne edr .b " yH te he alt E h Sa mploy vings A ee Be cne count fit Re s a sea nd rH ch I ealt ns h C titu atre e ( Sp EB eRI nding ). It." m H ae y ab lteh Tenure (years) 14.2 14.0 73% 2.0 16.9 17.2 -2.9 Seeht http tp:// :/w /w w w w w.e .re br ai.or lcleg a/pd rhea f/ltph. ub co lic m/ ataions rticle /bo s/2016/ oks/c11/ ons1um 6/the er_d _arcia vs e_imp n_heaaclt t_o h.p n_e df mployer- 72% 72% m T inc found HSA ah om tc -hin e ee ligib t ha I w g or tpmp le w rk o . e he c The res w n C ss a . p ec D Sin a re H pt e P c m of rs a e or found the ree Of li offe p kool etly ha rf e tof te o d he c c ar hoo om long aing lthie pse a sirr d it Fr son g e indi he ot H v he e idua RA roup re tp yl la s w p cCo e n as. ndid e of Fow rev he m ae le tor e as, e s a ltr eh p a lik tt g E la a em l. ly ns e p (t, 2004) lo on o he yce hoo a r ltB hie E e se x wa rat m n s som he ein m r e p H o d loy D ll e d H w e a Pme e t ha a s w in b tfr sm om e ot n re a h th t a lle m Cho m r or e ta e h fir il su a lik s n th te ra ly vi nd e e c ty ha s e tsudy e n nt economic security. Founded in 1978, EBRI is the most authoritative and objective source of individual savings or reimbursement account to pay out-of-pocket costs for health care services, were first introduced trend in HSA-eligible health plan entry—common to both Difference-in-differences Model maxi us Se m eum d rv ic w e it m s Re hout ay not sea pe rr ec m xc his e 45 si eon d ( 4) $6 , b : ,550 ut 10 c 41 itfor a -10 tion o self 60 -( f th onl 201 y e 0) c sou ov . e rc ra eg is e r or eq $1 uir3,10 ed. 0 for family coverage with the deductible counting Household Size (Established in 1978, the Employee Benefit number) 1.0 1.0 0.0 3.3 3.3 1.2 provided_health_benefits_110255.html 1 c t le P second o ohor la ss 70 e n sp m % he tp, a loy onsor y tlt he ehe ay e ra s s a e s n . m a Thi atly p ls H loy si o s u s m m w m ee a a as t a s t y na t w co he he I ae nc nt d nr fi . rw toll in st in June o it p e h r xplore ap C ee p D r la H 20 tc o Pw e 0 s. (T tm 1, w a ry y e s in w n this o thic w d e lit h ha itthic hin e er rm a h th d ta i ur ne int ce e a r y lipe if su is od c p a uc rrn e c e e e h a sent q dnha ua a dn HR v l to o ee nc dr se e in m A r bsele - e le lik ne or ss ec e fit the tion d d ha s offe e atn 1 lt ah p il i is/5 rsi n th e la p dof n a in t . te e a A d a b st he p out ov a p nda e H endix.) rSA 2 t00 ri- m de e d ligib p, e e v op aia nd le le t io he e n o nr alt olle f th h pd e la n information on these critical, complex issues. 66% by employers in 2001. Today, 29 percent of workers are enrolled in some form of a CDHP. There are two types of and Risk Selection in an HSA-ee mp li loy geibl rs—w e a Hea s positivelth during Pl the a st n udy p eriod independent Charlson CE om st oirm bid a itt ye Is n do ef x C (sD co H rP e) Entry0.16 0.16 -0.1 0.35 0.38 -3.0 toward this limit. These minimum allowable deductibles and maximum out-of-pocket limits are indexed to inflation. Fendrick, A. Mark, and Michael E. Chernew. "Precision Benefit Design—Using 'Smarter' Deductibles to Better Engage 63% Conclusion 5 Research Institute (EBRI) is the only in t p b dr y eop tc e he ov rem ns C eine r D it ing y H d P s t c ar ha or s m ete tur w a d ne ny is hil td r ep ibut tr the he ev ion a e su rn etrsu a vb s elty le rs w , e serv a co nd em re ic m t he st ees a nde irta ia s p stns d ic oss b a wy lly e ible r A s w si us g tnif out e in ( re icsi 2011 a cd n om e t in b of p aa )t r he a e ot nd dh ye d to e Ad u a a uc st b rout s, t in ( tible he 61% 20 2,6 (11 m Fe 00 ab ndri g )ind n . it The cude iv kidua aa nd of na ls e C tly he he si nr s s re olle ne ffubse e w c d t 20 in m of q17 ue he or )nt . ae ly O lt h thsetra tus C D Se HeP C http s: H hronsSA ic :/ C /-g o ea n ligib ll de itry iole n .ma s he ila clt himp h pla .cns om/ —e he 399e alth p ff155a lans 589b that 9a c9a an b 45e b 37d com /file bine s/6a d w bcit9872 h he-a f0a6 lth s -494 avin4g-s a a1ac5- counts (HSAs) – and health Lo Sa Rec sso o, m Am nten hod ny ed T., CiThom tatioa ns Ric : Paul Fr e, Jon ons R. tin a Gand bel, M. and Chr Hist eidi ophe Whit r Roe mor beuc . "kTa , “le The s frIom mp a the ct of NeOffe w Frront ing ie Fr r: ee P ione Cove errs' age on of the reduction in premiums. New HSA-eligible health plan 59% 59% Certa Cin p onsu rim ma er ry s a pnd revMit entiga ivet e se C rEmployee-only v ost ice-s Re —la tytp eid c aN llona y those Family dherd enc eee m ." e d JA tMA o p rIent ve ent rna tl M he e onset of dicine 17 d7( ise 3) ase : 36 —8 ca -370 n be( 20 and 17oft ). en are By WhaPa t is ul Fro not know nst n f in, rom Ph.D. the lit & era M. tur e Chris is whe top the he r ar R nd oe how buck, adverse Ph.D. selec tion can be affected. This Issue Brief is a This st EBRI focus udy examine es solel d the exty e on emplo nt to which fin yee anc benefits research — no lobb ial incentives encouraged enrollment ying in aor advocacy n HSA-eligible he . alth plan, and r t a on r e d a t60 m a d eine it % ini nr ion H ollm s dy t a rp ta l cov he ete riv tnt efr e n e s e w a irq o nd a a n ue g s low ee nc . dThe y uc e w r a e tin st ion ight u the d ay l tools s t second found o be t ha us tha y e te d e t 12% a le self r in a v va e-tll subse rre su e p ts t he orhe tc eq on d 11% ue fir h cst e e nt a p lt for ta h s of na w lt e y a or xpe ses. tus kec r3 p ts w A e .r 1 ft e dd e it he ic rh e t ta his elt d m h ca p p pla lo ron c r yc 13% e e e e ss e hoic -xpe only , the ens . c S e st ov p s m ud eec14% rifi y a a y g c g a e raoup lly . lso he The , e w mop a lp. ps -loy 1 p .osit 0ees e in p 9021b lans EBR 025e coIm Em b 485/ ine ployd up ee w d Be it ah he t ne ed fit_EO_ R aes lth r ed arra e cim h I ft_3_ ns bur titut .seme p e dIfs sue nt Brie arrfa nge (ISSN m 08 ent 87s - 13 (H 7X RA ) is s) pu . B bli ot she h ty d by p t ehe s of 10 Ema pl coy co ee unt Be s m nefita R yes re ea ce rciv h I ens tta itx ut-e, fa 11 vo 00 re d EE nr xpe ollm rie enc nte C s w hoic itindependent nonpr h e C aonsu nd Ris mke rS-e Dle riv cteion n He in a aofit, nonpartisan ltn HS h CaA re -e ."ligib Hea le lt h Se Healt rv h P icela s Re n,” sea EBRI rc hI ss 39 ue ( 4 (P Brieaf,r tno. II))4 : 35 10, 7(1 E-m 10 p89 loy e (2 e00 Be 4) ne . fit Independent Variable Coverage Coverage enrollment increased by 1?2 percentage points in 2013 and by exempted from the deductible and covered in full (Fendrick, Cliff, et al. 2014). Otherwise, all health care services must significant contribution in this regard. Data were analyzed from a large employer that attempted to encourage 13th St. NW EBRI , Suite stand 878, Ws asalone in em hington, DC, 200 pl05 oyee -4051ben , at 2 $3 efits 00 pe rese r yeaar r orch is ia nc s lu an ded a inde s par pend t of a e mn em t, bnonprofit, and no ership subscription. Pre npa sortrtisan ed Dyslipidemia 14% 13% 1.9 18% 18% 0.2 whether the 51% incentive impacted risk selection into the plan. As a result of eliminating premiums for the HSA-eligible c I e w ndivi om xc as t ep lle r r due iu sed nt a ls for or of w vindiv it e 11 h c ry,711 g hr idu ood oni aels m che w p cond a loy itlt h hefa it e w ions m -e only il re y tsi c ha a o gnd v nif te a ric 37 r ae g a ,129 e n w— te lyll m he m fa ao a lm trn hy eily a lik g fe sa e u d m lb yc ili ja e te ha n cs w ts, a c n th onfide er nd ose e m tnt he or in ly e c w e om lik or ste im se ply aa rhe tis to eon a sel tlhe th to c g eirc roup the the hoose a clt onsi H h D cH a st a rP e e C in 200 dus D of H eP 4,77 a , t7, w a e nd nr 4 hi oll those cm h e w nt a, s cTreatment ontribution Group s and x 2013 mak e tax-favored 0d .2 is 1tr*** ibutions0..2 W 5 *** hen CDHPs were first introduced, there was an expectation that Fendri Resea ck, rA ch I . Ma ns rtkit , ut Eliz e, aJuly beth 12 Q. , 2 C01 liff, M 7). . Richard McKellar, Kaden Milkovich, Kathryn Spangler, Michael Chernew, and 6 4?9 percentage points in 2014, compared with 2012. This be su standard postage rate paid in Dulles, bject to the deductible. VA. POSTMASTER: Send address changes to: EBRI Issue Brief, 1100 13th St. NW, Suite 878, organization committed 48% exclusively to data 48% For exampleorga , see nization. Buchmuell It analyzes a er (2009), Bund nd orep rf (2012) orts re , sea Buntin, rch e data t al. (without 2011), Fron spin o stin r a unde nd Ro rlying ebucka (gend 2013) a. , Fron All fin stin, dings, Diabetes 4% 4% 1.0 5% 6% -0.9 enrollment and mitigate adverse selection by eliminating premiums for its HSA-eligible health plan, which was offered he50 alt %h plan, enrollment increased by 21 percentage points among individuals with employee-only coverage and 25 per- Mc e r a not Treatment e nd m pD p or ob e loy cta ving serv n b ite, eRola -g e Group only er m d e aor nd in 200 t e a e x rnd D e 2014 he ., a14 6 si a A. lly m t ,555 Thi h se e con lia s p r s v v M. ubje ia ic nc p eH e e s rd a c ut tv a s. Fin of il ls ili 0 ao z .nd, t2a he found 1tion w a *** ll Ry fin y, a atn Lor n te he ha c ria e 0 t rl v .si e t 2 e he su g 1, anif lue L *** l td a ing ic is ur a p ta r nt rc ibut ov L ov ly aide ude a le ion o ria ss d nbe t b e lif t y kb e r a a he g ly n HSA la e t n rhe o , cMa e c a hoose lt - ae t hy ctligib rhe os aw s t c le a r o E he C ss he is D e a t HP tnbe w he lth p o . r tP g w ga la r, oup o rn. e ahe nt nd Fi s a e ana , N lttFe h p e lb ly eldma a r, la a seline m jn Sood o s b n a ving e w nd c. a a" tm s o Riesk a Washington, DC, 20005-4051. Copyright 2017 by Employee Benefit Research Institute. All rights reserved. No. 435. providing participants with a savings or reimbursement account and subjecting claims to high deductibles—requiring Stephen Parente. Health Savings Account-Eligible High Deductible Health Plans: Updating the Definition of confirms the upward sloping trends for large employers during A T A G L A N C E Asthmawheth /Chronic e r Oon fi bstrunan ctivecial data, options, or trends, are revealing and reliable — the reason EBRI information is Sepulveda and Roebuck (2013a), and Fronstin, Sepulveda and Roebuck (2013b). alongside other insurance options. In addition to estimating the impact of this financial incentive on enrollment, this Male -0.001 0.01 *** dissemination, research, and education on cent Ra Se ep ge le o c p rtoint t Availabili ion int s ao mong Consu ty indiv :m Thi eir- ds r ua Dire ls ep cor w teit td h f is H a a em v aa lt ililh P ya b cov le la ns e on r: ag A the e n . A C int na lee a ly rrne si lys of , t w aor t Fa w ke w m rw s a ily .e nd C bhoi ri.or the ce gis w r fait m hin ilie s a Larrg ee highly Employ sensi ers." tivH ee tao lth health a full C m ss hr or -e ir e st s e sed ib p aa la ns la c a on e nc nd me( ed 2004 is nt ov ppre la e a rsent n )t he e (w xa e he td m w r o in Figur in ee y only d e ad ra s. te a an 2, w on HSp A hic la -en ch h show ligiboi lec e he s t fr a h om lt ah p t a tll v he lan a fir w ris a atb s offe y lee a m r e rte a hd ns a)t w tw he ould er e Univ w ee lim e ll w rsi ina tityhin t e of r itMinne sk he sele 10so p ct t eion a r cof e, nt fe a s rb etid a he s arn HRA e A out n e -of m-p ploy ock ee r tm pa ay y m offe ent r b an H eforSA e t-he eligib y arle e he cov ae ltrh p ed la bn, y ins but ur is anc not e— re wq ould uire d induc to cont e enr ribut olle ee ts t o o an HSA make. m W or he e n cos an e t- a m nd ploy qua er lity- Prevention. Center for Value-Based Insurance Design, University of Michigan (2014). Accessed May 2014. Pulm the onargold sta y Disease ndard for priva0 te .12 analysts a 39% % 0 nd .05de t % hecisi saon me make 2 ti.m 0 e fr rs, am gove e derp nme 0 ic.t 1e 2d % n t in Figur policym e 0 .ake 2 1. 4 %rs, the media, and -2.8 Age -0.003 *** -0.003 *** a 7 nalysis investigates whether less-healthy individuals were responsive to the $0 premium. Data from a second large ins40 ur %ance premiums. 36% t w phr la ould e nsh . Se They old b rv eic no ree found s Re cot om he sea mrno e he rnde ca h as lt 49 d soc h p in t (ia 2) la the n for ion b : 6 lit 09 e e e r -ta 62 m w tur p e 9loy e e (n a , 20 e ee 1 g xs t c 4) ee or .o p t m self for ov -e one r e to p— .o rttot eda l num health be ar nd of the em lik erg ee lih nc oo y d d e of pa crhoosin tment gv is the its m HRA ad e p la by n . the Some research combines enrollees in HRA and HSA-eligible health plans into one group for study purposes. However, conscious health care decisions (Herzlinger 2002). However, for employers that preferred to offer a CDHP as a choice does c htt D ont p e:// pr re ibu vsb s the idce t io en tpubli o nta e economic security and employee benefits rw .o cor . rg k/wp er’s- H con SAt, et nt hose /uplo c3% a od nt s/ ribut 2014 ion /0s a 7/ 3% H re D H exc P-lude whitd e -fr p. -om a 0.p 6e rt_f heina wl.p ork defr 2% ’s taxable inc2% ome. Employ -0e .1r Tenure 0.0004 *** 0.0004 *** 9 employer T Mc his Dest vit u, de y provide t e axa l. (m 201 in the e4) s w counterfactual for 32% buil hett h on er of thef e w rior ng a k c the ond hestudy a uc ltth e sa d — b what vyin Lgs avwould a e, ce cou t ahave n l. t( (HSA) 2011) happened - ae nd ligible exa in the m hine eald absence th r pla isk sele n fof or f ctthis ion two reefinancial , alo yn egside ars af o tetrh er 40% 11 The Employee Benefit Research Institute (EBRI) was founded in 1978. Its mission is to e fa Hnro ow milll e H y e v e e ( e as 12 r l t, h a. tnd S 8 p he e rh y v e e ic r d a c e id lth e snt U fin s t ie b lid z ria v atic s) tha ioe. ns t u highe tilization r-incp om atte e rns em p sho loy uld ee s w be e ex r Fi e a mine nanc more d ial lik sep e Inc a lyra tte o ely n choose tb iv ec eas u ’ Eff s the e H H e RRA As ct s . a n I on n d a H R S sep isk As a c r re S ae t ae l te e s c d tudy, tiff ion ere nt in This study was conducted through the EBRI Center for Research on Health Benefits Innovation (EBRI CRHBI), with c aont mong r ibut seve ions m ral he usa t lb th p e the lan o sap m tions e dolla , the r a re m w ount as ao r fe ta he r of sa“m ad ev p ee rse rcesele ntag ce tion” of t he me d ae nin duc g ttible hat for CD H aP ll e s w mould ploy em es. ainl yRe atgta ra rd cle t ss Household Size 0.02 *** 36% h Pa eraelt nt he pla , Stne opt pheio n T., ns w Rog ith a er Fe pre ldma miucFigure n, m ont , a arind but ltee Jon 5 rs e to, t B m . oploy e Chr ncie st oe ur ia a e ngn son. " erollm , and t E em n o e tp c loy nhha oi enc e c eeC t s; hoi he a de c ned T if of v his e r lC opm e a onsu spon naely ntm si de of s d er s rs di -id D oun rinot vfd e e fe n He r b fin mpl d ya oy h st lte e h I reong a be lt nh su ne rfaitnc e incentive? the adoption o The f a analysis n HSA-employs a eligible hea pre lth p -post, lan. treatment-control study design where The authors also examined risk selection i individuals nto HRA are plamatched ns over tat ime baseline, . Using 30% Inpatient Hospital Days (number Fow Parele nts, Jinne e, Feldma t Br n igg and s, El Ch iza ris bte ia tns h A. K on (ind, 2004 Bb a)r b found ara L. tha Bra t u Hn, RA a nd enrJohn ollees ha Bertd k o. a low "Eae rly r aEvxep re ar gie e nc bur e d W eit n o h E f m illp ne loy sse teh 34% C an P hoic Pe O of incethe ntive funding s relating support to the of use the of following health care organizations: American Express, Aon services. HRAs aH reS he A- aElig lth ca ib rel e re H imb ea urs lt Hewitt, h eme Pla ntn o Blue r En sprol e Cross nd lme ing Blue a nt cc oun Shield ts, not savings yC oung harEBRI explo ls oa n nd Com he orablt id hy ity r es the breadth of emplo w Ind or ex kers (Gauthie -0.0 r 0a 5nd ***Clanc -0y .0 y 20 0ee ben 404 ***). efits and related issues. of whether an employer contributes to an HSA, an individual enrolled in an HSA-eligible health plan is eligible to open Figure 6 Who we are 23% programs and sound public policy through objective research and education. EBRI is the only per 10 Difference-in-differences 0) 0.Model 04 Estimates 0.04 of 0.0 evide 0.nc 4 e that sugg 0.e 4sts the posit -0.7ive risk and st daatt au in a the fs. T rom effect Mult h 16 e da ipl laof ta ra g n, f the eor t Mult em financial incentive h piprodu iloy s st erus d dy c ct ur Se ome ing tting." t he fris om t 20 H calculated e04 aw lt? h Se o la 2007 rge rv using ip c e e ers Re m iod ploy a , sea difference-i the errs a crh e sea 39 nd r(cc 4 (P n he o-differences regression framework. This vre s a ar t rtn h Ia I el) y y )ze : e10 a drs 9 ris 2 1k -0 1 sele 1 11 1 t 2 (2 co tion for 2 00 01 4a 4, ) .tspa he n tw no in g t yeahre s Methods The Institute seeks to advance the public’s, YearC =onsu 2013mer-Directed Health Plans 0 .a 0nd 1 *** Satisfa0 c.t0io 2n Wit *** h Enrollment." Health Services Research 39 (4) (Part II)): a ac nd coun POS ts. e Tnr he olle EBRI ina eb s in t ilistudie ty to he ro y sll e the worl a orv e br ef un orus e d e tof health and retirement ben he d a H mo RA un w tsa fro s in m tro od ne u cye ed a.r to the efit ne s xt, — issue or to ta sk e such a unuses d 40 amo 1(k)s, unts IRAs, retire to a new job,ment Association, Boeing, Healthways, IBM, JP Morgan Chase, Kaiser Permanente, Mercer, Pfizer and PhRMA. and contribute to an HSA on a tax-preferred basis. For 2017, a wor 28% ker with individual coverage can make an annual private, nonprofit, nonpartisan, Washington, DC-based organization committed exclusively to Difference-in-differences Figure 4 a Model lso pr Estimates ovides some confirmation of the existing Emergency Department Visits 26% CDHP Entry, Employee-only Coverage 2013 intervention when one of the two employers eliminated employee prem sele ium cs f tion rou or the tine HS ly A r -e ep ligible orted h to ea alt ffh e ct approach is further p Yost ear-= C2D 0H 14 P adoptionsupported (three yea by rs fo four r 0 .11 0years of 4 of *** the data 16 0. 0e 9— m *** p two loye years before rs). While evand idenc after e 25% of the risk HSA-eligible selection emhealth plan premium erged, employer 20% income adequacy, consumer-driven benefits, Social Security, tax treatment of both retirement and health Understanding what drives enrollment in HSA-eligible health plans is important to plan sponsors, since enrollment in 1141-1158 (2004). public policy research and education on economic security and employee benefit issues. c In a read tedsit the ion e to quiv pre ale sent nt o ing f a v“u arsie a-bitle -or m -lo ea sns e-i t” atinc ba eseline ntive. , In mul adtd iv iti ao ria n,t e b em ca eu thod se Hs w RAs e raer ee a m lwa ploy yse fu d nd to eob d b ta yin e emst ploy ime ars te,s of work th ee rs 23% HSA contribution of $3, the media’ 400, while s and policymaker a worker with family s’ knowledge coverage can contribute as much as $6,750. Individuals who of CDHP Entry, Family Coverage (number per 100) 22% 0.1 0.1 literature tha0t. 2 HSA-eligible 0 he .5alth plans fa 0.c 7e positive -1 2 ri.sk 8 . "Evaluation of the Effect of a Consumer-Driven Health Plan on Medical Care Expenditures and Utilization." Chronic Conditions HSA-eligible health plan enrollment was pla elimination, Tollen, cont nr.ibut Ross ions which is important a benefits, w nd erP eoor use 20% ( cost d20 to 04 rmanage e ) d w uc for as t e ensuring it he .ment only , worker a one that of underlying the nd fiv employe e stsecular udiers t attitude o trends use p s, rwere ior policy reform p -yecomparable ar claims dr aopo across the ta asals, s a prand p etwo employers. dictoe r n of sion plan assets EBRI’s membership includes a cross-section of pension funds; businesses; trade associations; tIhe nde se pep nla dens nt V isa re ia xpe blected to grow. According Model to Me 1rcer’s su Model rvey 2 of emp Model loyer3 s, 25 percent of employers with 10?499 ma effe yc tte of nd tto he v e ielim w in the at ion o accoun f e ts m apslo “fre yee e ” pmo rem ne ium y, s. Sp which ec re ifinfo ically rc, es a ad iff spe ernednc ing e— -in r- ad the iffe r rtha enc n es s (DiD aving— ) r m ee sea ntarli cth d y. In esicg on w ntraass t, have re Praim che aryd C a ag re e P 55 hy sa ic nd iana V re is itnot s yet enrolle 1.6d in Medica1 re .5 may make a 2n a .2 dditional $1, 5.2000 catch-up 5.7contribution -7.5. Table o 17% f Contents selection. Specifically, older workers were less likely to enroll Independent Variable Model 1 Model 2 Model 3 HyH pe era te ltnh Se sionrvices Research 39 (4 (P -0.0a 1rt ** II)): 11 -0.89 01-12 ***10 (2004b). and fundi and under ng. There is wi standing of employee benefits and laborde uni spread ons; heare lth c cog are n pr ition ovide that if employee be rs and insurers; governm nee fits nt or data exist, EBRI kno ganizations; and servicew fis rm it. s. Fr Main onsEffects tin, Paul. "Can 'Consumerism' Slow the Rate of Health Benefit Cost Increases? mod " e ErB aRI ted I ss byue e lim Brie inf no.24 ating th 7e ( p Er m ep mloy ium ee . Finally, choic 10%e. I it t is als important o examine to d point the expe out rie that ncewhile that H some umana published Inc., hastudies combined d with its internal r HRA- ollouand t of tHSA-eligible he HRA-like health plan product. The employees and 61 percent of employers with 500 or more employees offered an HSA-eligible health plan or HRA in Specialist Physician Visits 1.6 1.6 3.8 4.9 5.1 -3.8 e Hm SAs ploy are ed s a w vit ing h o s radcina coun ry ts le,a fu stnd squa ed b ry e s r em ep gloy res esi eon s an m de /o thod r emsp tloy o e exa rs,m bine ut a tre he fu im llyp p ao crta t of ble th a end fin o aw nc ne iadl inc by e the ntiv ind e on ividH ua Sl, Aa -nd Income and gains on HSA funds are tax-free. Dyslipidemia Main Effects 0.002 0.01 *** in the HSA-eligible health plan than younger employees. Here are the key insights: Fina Tre ll a yt, mJor entd G arn oua pnd x 2C 0ot 13ter (2016) examined0 .d 2a 2ta *** from th0e .2 int 3 *** roduction o 0.25 f b *** oth an HRA and HSA-eligible health plan in Introduction .......................................................................................................................................................... 3 While there was weak evidence that enrollees study found Benefit into Re tha one sea t the rgroup ch reI ns did tfor itnot utanalytical e , aJuly ppe a 20 r tpurposes, 02 o .b)e . ht a tdpiff :/ the e /ww renc current w e.e in r bri.or is Issue k gp/pd roBrief fil f/ebs b rie does not. e fstp wd ef/ e0 n e 70m These 2ib.p ploydeftwo e s choosin very different g the HR plan types A and 2016 (P Fi re gs ucrre ip 1) tion . B Dy rutheir importance to our nation’ g 20 F19 ills , (3 34 0- dp ae yrcent of employers with 10? s economy 499 employ. ees and 72 percent of employers with 500 or e un ligib usele d a he mo alt un h p ts la ca n e n b nr e ollm rollee dnt o v(e Ar ng tor ith ste a ne nd xt Pis yeca hrk a end 20 e 08 ve)n . The taken Dfro iD m deo sine gn r joe blie tos on the n the ext. ass Furthe umpt rmo ion th re, a un t tlie km e p Hor RA alfu nds, Diabetes Treatment G-r0 o.u 0p 0 4 x 2013 -0.02 *** 0.25 *** 0.26 *** 0.24 *** . "The Impact of Health Status and Price on Plan Selection in a Mutliple-Choice Health Benefit Program Including Treatment Group x 2014 EBRI’s w 0or .2k 1 a ***dvances 0.k Mor 2now 3 *** elov edg ee r, a lend u 0ss .2 4hende *** altrhy sta nwdior ngk eof rs a end mplwoy oreek be ers w nefitith le s and t ss hehe ailrt hy EBRI delivers a steady stream of invaluable research and analysis. 2006 from a single employer with about 20,000 employees. Less healthy households were reportedly more likely to adjusted) 13.1 12.6 2.4 prior 31 us .8 ers of hea 3lt 5h .0care ser-v8ic .5es were should those 0% wbe ith mo examined re tradseparately itional covebecause rage based they on present dissimilar demographic datincentives a alone, but relating it appe to arthe ed tuse hat of thohealth care se in the HRA services, were Data ..................................................................................................................................................................... 5 m Dis or treibu em tions ploy e fre om s sa aid n HSA the Try e a cw ta m e n b e rn et e v G e tra r oy k u e plik n a xe 2ly 0 t 1 a t4o nyof tife mre su . They ch a he are a lt gh p ene 0.la 2r2 a n. lly *** Onl treya t 9 p e 0d .2 e 2 ars t c *** ent ax a of ble e0 m .inc 2p 0loy om ***eer, s w but it h 50 they0 o are r more Asthma/Chronic Obstructive H cha SA nge amo s in he unts caaln th p bela un e sednr fo ol r lm noennt -he w ae lth re etx he pe sa nsm ese ( a ble btew ite o en n th a ta exa Em ble p loy bas eir s )A , w (st hic udy g h furthe roup) r coantrib nd Eute mp sloy to ethe r B s(a cv om e inc pae rn istiv on e. ? Health insurance premi im um pors ta nc ar ee t o t a he majo nation’ r d srive econr om of y a pl ma ong n pol choi icy cm ea . kAfte ers, tr e he n lim ew in s am tin edg e ia, an md ploy the epeublic. It InterH acRA tion a E nd ffec HtSA s W Op ith C tion hars." lsonUniv Come or rb siid ty ity of Ind Minne ex sota Working Paper (2008). EBRI publications include in-depth coverage of key issues and trends; summaries of research families—as measured by the Charlson Comorbidity Index— Fronstin, Pa 20ul, a 10 nd Anne 2011 Elmlinge 20 r." 12Consum 720 er 13 Engagem 20e 14 nt in Hea 20 lt15 h Care: Fi 20 ndi 16ngs fVe rom ry li kt ehe ly to 20 Ve 16 ry E likB ely R tIo/Gr Ve ee rynw likea lyl td o & choose Sour tc he e: Em HSA ploy , ee b u Bt en w efe it r Re es a ea lsrc o h le Insss tit ut lik e e esly tim tat o es choose based o n the adm H inis RA tra.t ive enrollment and claims data. 3 more likely to enroll in the HSA-eligible which in turn may healthier than those be erelated le Intc etring actiotto n o E renrollment. fe fm eca tsin in tr With Ca h d arit ls io ona n C l co omovre br idaitg ye In w de he x n based on prior claims and prior use of health care Pulmonary Disease -0.01 ** -0.01 ** workers offered an HSA-eligible hedoe alth p s thi lasn o byr c aonduc n HRA ting as a tnd he p only ubli sphilang n o pp oltiion in cy rese20 arc 1h, 6 a . nalysis, and special reports on excluded from an individual’s taxable income if they are used to pay for qualified medical expenses as defined under Con g Me roup Cth s hod a erq )l s ue b s o e n ................................................................ ntl for x y 2e ,0 a 1 t he 3wo int rkerr od ha us c ta io muc n of th h mo e re fin a co n-mp c 0ia .0l in e 5lli*** ng ce nt reia ................................ vs eo. n Fino gut rto e 3 s ppelot nds t Hhe SA ............................................................... H fuSA- ndse (ligib relale tiv e he to alt H h p RA la fu n e ndnr s)oll , bm ut ernt athe rart et oov e 6 r What we do offer in 2017 offer in 2018 offer in 2019 premiums f finding or all c s and overa policy ge tie develo rs, HSp Aments; -eligible timel heay lt h factsheet plan enrollm s on hot topics; ent increase regula d from r up 4 pe dates rcent t on legi o 25 slat peive r- and Notes: Comparison group subjects may be used multiple times as controw lse (i. re. e, m leea ss ns lik are e ly fre qu tha enn cy t whe eigh irt ed he ).althier counterparts to enter the Associates Consumer Engagement in Health Care Survey." EBRI Issue Brief, no. 433 (Employee Benefit Research EBRI’s mission is to contribute to, to Depression Charlson x 2e 0 0m 1 .0 3 pl 02 oyee bene 0f.i0 t i 0s 1sues; holding e-duc 0.0a 2ti*** onal briefings for EBRI members, congressional and health plan as a result of the services. Charlson x 2014 -0.05 *** Qua Inter n, naH l Re ., Pv. eH nue alfo C n Sund ode (Ia RrC a)r aSe jac n, . 21 A. 3( Fong, d). A B n ind . Bur ivna idu nd, al n J. ee Cd . L not uthi, L. D be cov. eSa red unde by a rs, C n HS . A A-. eB ligib eckle , T. E. Fe health p asby lan , to and witW hd. rA aw . regulatory developments; comprehensive reference resources on benefit programs and workforce t re im taein for th etm hefo tw r th o e e m fup ture loye . rs including all enrollees (i.e., prior to implementing propensity score matching). It is clear that cent among individuals with employee-only coverage and from 3 percent to 28 percent among individuals Results ................................................................................................................................................................. 6 Source: Figure 6 in http://www.mercer.com/newsroom/national-survey-of-em H pSA loye-re -sligib ponsole re dhe -hea allt thh p -planla s-n. 2016 I.h ndivi tml duals with specific chronic HealI th ns Steit ru vitce e, sMa Utily iz a 20 tio1 n7) C.h arlson x 2f 0ed 14 eral agency staff, and the news m -0ed .02 ia;*** and sponsoring public opinion surveys on employee W Inhe tera n th ction e E Pfa fe tie cts nt W P itrho C te hc ro tio nic n a Cond nditA ioff no srdable Care Act (ACA) passed in 2010, it wa es e limx ina petcion o ted tf p o a rc ec m eiu lem ras, for te the t he inc rm ea ost se issue encour s; and age major survey , and to enhance the development s of public attitudes. mone Gy ha frli. " omC tod heing HSA Algo . Ar ny ithm HSA s for am D oe unt fins u ing nus Com edor ab tidit the ie e s in I nd of CD the -9- C ye M ara nd can b ICD e -r10 olle Add m ovinis er in t trathe ive aD ca coun ta." tMe and dicus al C eda r in e 42 the pre-period trends (i.e., slopes) in HSA-eligible health plan enrollment were very similar, and this evidence bolsters Data with family coverage. benefit issues. EBRI’s Education and Research Fund (EBRI-ERF) performs the charitable, References Interaction Effects With Chronic Conditions Inpatient Hospital Days 0.0001 -0.0002 * conditions were also less likely to enroll, with one exception— H GS re Fi Ae -na ne elin g , c ib e ia le t l I a he l. nc a (e 2006) lth nt iv ple a s’ ns coE m sfho fe bin c uld te s on d a lsu so H r v b SA e ey -e E inf xa ligib m orine m lea d H t ion w s ee ap ltah P ra ith tla e Figure ly cn E l afro im nr m s d ollm 3 HaR te A an -fr re t ola m te a d m heaanuf lth a pc la tur nsing beccaom usep a an n ye m thp aloy t ae drd tha ed tt w off o ers 6 Hypertension x 2013 -0.07 *** in employer adoption of HSA-eligible health plans. Health insurance premiums were expected to increase because of EBRI meetings present and explore issues with thought leade pars rt, from all se the findings w ctors. ere to the contrary. futur(e 11 y)e: a11 rs. 3A 0ls -9 (2 o, e00 ve5) n if . an employ educ era tha iona s c l,o ant nd s ribut ciee ntdif itc o f tunc he tiH ons SA of , it t he is I indi nstv itidua ute. E lly B R ow I-E ne RF d ia snd a tax ful -ly ex epm or pt t or able ga ni if t zahe tion confidence in the use of the DiD approach to yield unbiased results. Emergency Department VH is yip tsertension -x 0 .2 00 113 *** -0.004 *** -0.04 *** Fr Thi ons s st tin, P udy m aul, a akend s us of sound employee benefit pr M. e of Chr he isa top lth insu her Roe ranb cuc e e knr . "ollm Brand en-tN a and m ogr ec a la ams and nd ims d Gene atar ic fr om Pre sc a rlipt arg ion D e marnuf ug a Use ctu rA eft r e he r A ad dq op ua tion o rteref a d in t Fuhe ll- 4 Hypertension x 2014 Enrollment in Consum -0.indiv 07er *** -idua directed ls with fa Health mily c ov Plan, erage with a person diagnosed with Angrist, J. D., and J. S. Pischke. Mostly Harmless Econometrics: An Empiricist's Companion. Princeton University Press the HRA fo op rme tion r m s t igo ht ita s lsaor rb ae y o of ffeprila ng n o op nly tions the in 20 high-04 de. duc Datib tal e on he2,10 alth 4 p plan lan while enr olle neith ese r wfu ernd e us inge d no : r 1,24 facili 3 HR tatinA g e wnr ork olle er es and 861 EBRI regularly provides congressional testimony, and briefs policymakers, member organizations, various provisions in the ACA. These premium increases along with the excise ta In sum x on hm igh ar-y c, ost the he fin ala th p ncia lal inc ns (t ehe ntiv so e - whic Fina h in nctia ul I rn r ncaeis ne tiv s t ehe s’ E he ffeaclt th r s on iskRis s upp in t k or he Se te le ot d by che tion in c r ont pla rin H bs. Ris ut SA ions -Ek l a igib n sele d g lerc aH tnt ion is c esa . lth Prla itin En cal in rollm eme pn loy t ment-based health plans 8 wor Prk im e ?a r rle y H C a ea va ree ls hi t P hi hys or e sr ic-it a he han- n V rH isjyob itp a sev .r te en ra sio gn -e 0 x . 0 e 2 0mp 0 114 **loye -0e .0s 0 0a 1re enticed by $0 prem -0ium .04 *** s for HSA-eligible health plans. Dyslipidemia Replacemx ent 2013 , Consumer-Directed Health Plan With a -0 .0 H3ea *** lth Savings Account." EBRI Notes, vol. 35., no. 3 Midwest (Employer A), but with emp by loy Emplo ees resid ying er, tCov hrouerage ghout tT he y pe, Unit201 ed St 1?2014 ates. For several years, this firm offered a dysl ipidemia where the effect was opposite in direction. (2008). and the media on employer benefits. c Po Pntrib O enr ution ollese s to . The an Hst SA. udy found In many t H ha SA t -w elihil gib ele self he-arlth ep or pltae ns d ,he wo ark lth s erst a atre us le w fta o s no n the t p irr e od wic n tito ve find of p ala n n ch HSA oic pro e, vie dnr erolle and e s i opn en an sound public policy through objective called “Cadillac Tax”) were expected to cause more employers to offer HSA-eligible health plans. The outlook for HSA- StataCorp. Stata Statistical Software: Release 14. College Station, TX: StataCorpd LrP e w (2 ne 01w 5) .indiv iduals and families into b Se eS c pp a ae us rc ait e ae 35 l is it ly % t inf P for hy lue se ic m n iac p ne lo V s w y iseite ho s D -y only sis lip e id nr a end m olle ia fa x 0d .m 2 0 in a 0 0il 1 1y 3 cp ola ve n a ra 0g nd, .0 e0, 1 four the ***r em for od ee , ls the w e crost e e st of im a a -he 0 te .0 a d 0lt 1 for h p tla he n. d For epe e nd xaem nt p le va , rif a iabll he le, HaSA lthy -e ligible ConclusiOff on e ................................................................ ring coverage with no payroll deduction a................................ ttracted individual en ................................ rollees who were ma............................ rginally healthier 9 Dyslipidemia x 2014 -0.03 *** choic (e E m of ploy hea ele th p Bela ne ns fit t ha Retsea inc rlude ch Ins d t pit ru efe ter , rMa ed rp crh 2014 ovider aor ).g anizations (PPOs), health maintenance organizations (HMOs), EBRI issues press releases on newsworthy developments, and is among the most widely quoted Similarly, individuals and families who used health services in a HcRA co- un ba t.sed plans were healthier tha EB n P RI Issu PO ee nr B olle riefess w is ahe ser n e ialxa wm ithini inng -dep ptrh e iorv c alla ua im tion of s dat e am . The ployest e be udy found nefit issue tsha and t t HRA rends , Dyslipidemia x 2014 0.0002 e Hligib R PrA es - le c E rlig he iptia o ilt n b h p l D e ru H la g e ns Fia lls lr t e (h 3m 0 -P a dins lans ay st ra ong nd, H sinc eael te h mR ploy eim erb s urs aree eme xpent cte Arra d to c n ont geinu me ent to ts he g ( r H a HpR SA pAs le -e w ) ligib ith e leff he ora ts t lth p o m laa n nw ag ho e tw he er e employees choose one plan and all sick employees choose a second plan, then the difference in plan costs will be large health plan enrollment. The first of the four models derived the main effect of the financial incentive, and the other Diabetes x 2013 research and education. -0.10 *** than those who would have enrolled without this financial incentive in place, therefore adverse selection sources on employee benefits by all media. A Ap us ptein, P ndix. — CB . a "c Ak n I grou ntrnd oduc on tio Cn to P DHPs ................................ ropensity Score Metho................................ ds for Reducing the E................................ ffects of Confounding31% ......................... in Observational 11 exclusive provider organizations (Eas POs) wel, l as and crH itiSA cal- e an ligib alysle es o he f e alt m h p plola yee n o bp en tion efits. B polie ci ges inn an ing d pin 201 roposal3, th s. EBeR e I No mptloy es e isr slo erw ial e red the year prior to the elimination of premiums for the HSA- Tollen, Laura A., MurrayD N iab . e R te oss s x, 2a 0nd 13 Stephen Poor. "Risk Segmentation Rela -0t.e 0d 7 t*** o the Offering of a Consumer- enrolle aO d ejs w uur stee dr ) e healthier and had highe -0.001r *** educ-a 0t.ion le 0003 v *** els than PPO enrollees. The health status effects were based on rising cost of providing health benefits. Furthermore, public policy changes may also drive future growth. Both the Diabetes x 2014 CHECK OUT EBRI’S WEBSITE! -0.11 *** on average healthier than those who t ahr nd ee the mse ode la ls r geexa dm iffine ered nc w ehe s itn cos her tthe w ill enr lik ollm ely e pnt rom rep sponse t differ e to nc the es in e incem ntpiv loy e e de iff c eont redr ibut by: ions the , Cw ha hic rlson C h willo cm aus orb eidit evy e n Index, Beca us e.w e"a m H s n p eloy aot m lth C ers a it hra iga ev e Sp t e E be m e d a nding e pln oys a e oeff -o n e n at ring lft y ic ce iopa r v HRA eA rt ad e gop d. e- (tr bta eing a stm ed ea np t Fu g la ro nll us p - Re )lonpgla ec r e th m ae n nthe t, H yigh ha-vDe e bdeuc en tible offe ri Hng ea lt Hh P SA-la en ligW iblit eh ha e aH lth eap ltla h Sa ns, v mo ing st s HRA-related h ealtEBRI h plans direct have s membe been in e rs xis and tenc other constit e longer thauen n HSA cie -s e ligib to the informatio le health plans n . they need an Unlike the HSA d -unde eligibrtakes le healtnew h providing current information on a variety of employee benefit topics. EBRIef is a weekly Studies." Multivariate Behavioral Research 46 (3): 399-424 (2011a). the employee premium contribution for the HSA-eligible health plan to $0—for all coverage tiers—to encourage Test Group Diabetes x 2-0 01 .0 403 -0.03 *** -0.09 *** HSA-E 30 ligib % le Health Plans and Health Savings Accounts e(ligib HSAle s) health plan were generally less likely than those who 11 Directed Health Plan: A Case Study of Humana Inc." Health Services Re 28% search 39 (4 (Part II)): 1167-1188 (2004). the A s fin thm ding a/CO tP ha D tx p 2ha 013 rmacy costs were 56 percent lower -a0m .0ong 5 ** higher-deductible HRA enrollees than PPO enrollees in American Health Care Act, passed by the U.S. House of Representatives in April 2017 would , a ha nd vet he ent Beertetd e rw Citahout re the further risk selection. This process can lead to the well-known phenomenon of the “death spiral,” where a health plan five chronic disearesea se indic rch ator on a s (h n yongoi pre ou rtnd enup of ng sion, ba E sis. dB yR slIipid reseem arc iah a , dnd i iabe ns te igs, A hts, a sth s m wae/C ll a OP s upda D, atned s on s deprur ess veion) ys, s, tudi and es,fiv lie ti gation, o pf lathe n, A t ce he ca ount rly em re : psA loy e a Fi rc evrh e c-o a Yn n ea CDH cru St stP o ud sm fo yize ." cus E the B ed RI o k e n Iss yH ue e Rle A m - Bbre a ie nt sf, no. es of d pla tns 38 he.8 Unfo H (RA Em - rtun p re loy laate te ee d ly B he , e ma ne altn fit h p y Re sla tud n sea iin esrc co lh I uding nlyn no st it t te he ut tha e d , e July t dthe uc 20 tyible a1 r3) e, e . xamining publications Constant A Fam sth ilm y c ao/C veO raP ge D ( 0 x tr . 1 ea 22 0 tme 1 *** 3nt group)0.08 *** -0.04 *** enrollment. There was otherwise no change in the set of plans offered. To study the effect of this financial incentive, Asthma/COPD x 2014 -0.d 0id 5 ** not use health care to enroll in the HSA-eligible health the year before the HRA was adopted. Similarly, pharmacy costs were 42 percent lower among lower-deductible HRA EBRI has earned widespr legislation and reead r gulationegar affed as cting employee benefit plans. The EBRI Databook on ReH conc RA-ili Ea ligib tion Ac le Hte a unv lth eP ile lad ns in a n the d H U. ea S. lth Re Sena im teb in June urseme nt 20 17 Arr, aw nge ould me rnt ais s ( e H the RA lim s) it on HSA contributions, lower the excis 11 e EBRI maintains and analyzes the most comprehensive databa incse entof ive401(k) in plac -type e. programs in the e mve ea nt su ua rells of y bep crom iore he s so c alth se ostr ly v ic tha es ut t it ili isza no tion (inp longera w tie or ntth o hoff spit ering. alizaItf e ions m, pe loy me ee rg s a enc rey v d ee ry p asensit rtment iv ev itsi o ts, p diffe rr im enc are ys in care “CDH Sour c P e: s Em ., ” "pl w Op oithou yee tim Ba en t l C s efpita e R lipe ces ify ea ing A r rc sW th hw idth Im ns he a tit /C the ut s for O e P es r Dth t im x P e r at 2op 0 CDH es 1e 4 ba ns sP ed i tiy s o -n a Sn cor HRA e Ma -ba tc se hin d p gl aW n,he an n E HS st Aim -ea li-g t0iing .b 0l4 e D *** he iffaeltrh enc pla en, s in Me or boa th. ns Als and o, CDH DiffePr e sn tud ceie s s copayments, health care services subject to or exempt from the deductible, and the comprehensiveness of coverage. EBRI’s website is easy to use and packed with useful information! Look for Whit Dep er, eH ss . io "n A x H2 e0t1 e3 roskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity." this analysis uses data fr Eo mm plo 20 yee11 -onl yt c o ov20 era1 g4, w e (comhic parh p ison g ro rov - u0 ide p.)00 s t 04wo years of data pre- and post-intervention. To be included Employee Benefits is a sta pla tisn tic —aa l r ge afin w erenc iteh o wor nek on e contrm apl dic oy te ion, e beindi nefiv t pr idua oglrs w amsit a h fa nd w mor ilyk 27% enrollees than PPworl O enr d. olle Its co es in t mputer he yesimul ar befor ation e the analy HRA s e w sa on So s adopcial ted. Secu The rity study reform uncov and retireme ered no age nt or income a income eff de equa cts, cy administrative enrollment and claims data. tax o n d . is "Qua tribu lit tions y of H foera non lth C -q aua re lif Aie fte dr eA xpe dop ns ting es, c a rF eull ate -Re a p gla racceem p ee nrtiod , H igh for -op Dee dnin uctg ible an HSA Healt, h P and lan Wi make th a it e H ae sia elt r h S fora vings premiums, then this sele Dc etpio ren ca ssion b n x e 20 p 1r 3oblematic in terms of offering more th0a .0 n o 1 ne type of health plan. On the other A phy ppsi endix cian vi —P sirtior s, spe Litecria atliur ste p ................................ hysician visits, and pr................................ escription drug fills). A ................................................................ ll models used robust standard errors and a.... ll 11 in Pr 25 op %ortions in Ob an or seganization that “tells it like it is rvational Studies." Pharmaceutical Statist,” ics 10 (2): 150-161 (2011b). A ma lso, y ha in c veon only tra s bte e to n H ex Sa Aminin s, HRgA a s a he re a lth offe plraen d, wow ith ne a hi d, gh fun de dd euc d, tib de le si ; gin ne od the and r wc oont rdsr , olle a hd ea b lth y e pm lap nloy tha etr s w . aMor s neeithe overr , Ht She A- plan Depression x 2014 -0.02 force-related issues. Econometrica 48 (4): 817-838 (1980). in t Notes: he these special features: *** st p<0.01, udy, e ** m p<0.05, ploye *p<0.10 es had to be full-time, ages 18?64, continuously eligible throughout the entire four-year study coverage who utilized specialist physician visits. It is worth are uni Fque. amily coverage (comparison group) Plan sponsors are concerned with but did find that there was less favorable selection into the more popular lower-deductible HRA plan. spouses to make catch-D up epc re osnt sir oibu n xt 2 ions 014. In addition, a draft Executive Orde-r 0 .w 00 ould 5 direct the Commissioner of the Account: A Five-Year Study." EBRI Issue Brief, no. 404 (Employee Benefit Research Institute, September 2014b). ha ana nd, lyses if ew meprloy e ce ond es a uc re te not d us ving ery St sensit ata/MP ive 14 to .2 diff (S et ra etnc aC eor s in p. p 20 re15 m) ium (W s, t hithe e 19 n it 80 is ). more likely that a firm can offer a Interaction Effects With Health Services Utilization eligible nor associated with an HRA. Not knowing the type of CDHP under study makes it difficult to draw conclusions from sponsors dictate the amount of money contributed to and available in the H25% RA, the extent to which unused amounts References .......................................................................................................................................................... 13 period (2011?2014)based on the f , and not enrolledacts in the . A Hs the Bylaws state: SA-eligible health plan in 2011. After applying these criteria, the study noting that the magnitudes of all of these health status effects on HSA-eligible health plan entry were relatively small. Interaction Effects With Health Services Utilization positive risk selection into HSA-eligible B Int aInpatient rerryna , C l Re olle Hospital v ee n L., nue Ma Se Days rrvki c x R. e2013 (C IR uS) lle n, to D ae llow ron ser Galus vice ha s , aMa nd rb tin D enefit . Sla s re dla e, te a0 d nd . 0t0 o Sus 04 the a n H. B manag us em ch. ent "W of ho chr Choose onic ds A ise aC ses onstu o m be er - varie•t y of EBRI’s entire library of re plans. The upshot is that the search publications starts at responsiveness of employees to the m change ain W s in pe re b page. Click on miums is an importEBRI ant Contact EBRI Publications, (202) 659-0670; fax publication orders to (202) 775-6312. p ca rior n bw eo rrk olle . d over into subsequent years, and the portability of the HRA upon job separation. E c Bohor ndno arry EBRI make , tt e e ctons s a ................................................................ l. is(t20 ed08 of s information freel ) 11 exa ,71 m Inp ine 9 w atie dor n t t he k He or ss w e pxpe itailt D h e rie ay ync m s av x e p l2a oy ailable to al 0t1 e A 3e lc -o only a ................................ a fc te ov r e itlr .a int ge rod , auc nd e37 d a ,12 n HR ............................................................ 9 w Ait in 20 h fam 04 0ily .0 0 t o c 0o 0ov 2 vee rr a g 17 e. ,0 In a 00 ll a emn pa loy lyses, ees. the 16 5 Fronstin, Paul, Martin J. Sepulveda, and M. Christopher Roebuck. "Consumer-Dirhe ectaeltdh p He la an lts, p h Pla arn tic s Re ulad rly uc w e h The en th Long e - 20% Inpatient Hospital Days “In all its activities x 2014 , the Institute shall 0.001 exclude Dire dc tfr eom d H e the alt h P ded la uc n? tible " He .a lth Af Sufa bsicrrsip 27 tio n(s6) to : E 16 B7 R1- I I16 ssu79 e B (ri 20 efs 08 ar)e . included as part of EBRI membership, or as part of a Results determina Issue Briefs nt of the stab and ility EBRI Notes of a firm’s hea for our in-depth and lth plan offerings. nonpartisan periodicals. W 8 hile O therd resu e EBRI lt rs/ s pr assume esent Ine pd at s iin Figur en a publi t Hospe itc a 4 a l service re Darye s int x 2e 0r1e 4 spo sting nsi ability nd we to remake largely its finding expected s , co und mpletely er 0s.0 ta 0nding 02 acce whe ssitble her at the www.ebri.org financial sa Using mple a is com strb ain tified atio n o by f e typ m ep of loyce ov e e fil ra eg s a e: nd em cp laloy ime s d e-a only ta, it a nd found famtilha y c t ov HRA era g en e. rolle es were more likely than PPO enrollees SE em ee Cha rgenrl cs y oD n, e p ea t rta m l. e( n1987) t Visits , x De 2y 0o 13 , Cherkin and Ciol (1992), and Quan, -e0 t .0 a1 l. (2005). Term Use Of Outpatient Physic$19 ian Vi 9 ann situa s A l s nd ubsPcrre ipt sc io ripti n to on EBRI Drugs." Note sH ae nd altE h BA Rff I I a psir la ss uns e32 B a r( rie 6) ef soffe : . C 11 hr26 an edge - 11 a of long 34 A (si ddr 20 de e 13 sot sa : he )E . BRI, r Finally, there is no statutory requirement that an HRA be paired with a high-deductible health plan. HRAs can be paired Financial Incentiv function strictly in an objective and es’ Effects on HSA-Eligible Health Plan Enrollment — so that all deci Emergenc sy io D ns epa that rtme relate to emp nt Visits x 2013 loyee benefits, whether made in Cong -0.01 ***ress or board rooms or 6 incentive to enroll in the HSA-eligible health plan moderated positive risk selection into that plan wa 15% s also a key Emergency Department Visits x 20141100 13th St. NW, Suite 878, Washi-ng 0.t0on, 04 DC, 20005-4051, (202) 659-0670; fax number, to be younger, of higher-wage, and white. They also found that prior health care spending and the presence of a Sinc Brot- eG 20 old 01 be , rtg he , Za re rha ekv C e ., be Ae m n num itabh C erh ous s andrtaudie , Bes e njaxa mm in R. ininH g atnde he d l, ea te nd rmJona inant ts a han T. nd op eK ff tol e ions c st ts of a. dG . iv "C W e Dn a ha HPt s. fix Doe e The ds a pool ini De tia of dl s uct tudi iblee s Do? There are examples of adverse selection and death spirals when employers have offered a choice of health plans. For 9 • Visit EBRI’s blog. Subscriptions with any type of health plan, although it is common for them to be paired with high-deductible health plans. When an families’ ho Em m ee rg s, enare cy Dbased epartme on the highe nt Visits x 2014st quality, most dependable informatio -0.01 *** n. EBRI’s Web site posts See Austin (2011) and Austin (2011). (202) 775-6312; e-mail: subscriptions@ebri.org Membership Information: Inquiries Figur To Pe rim staim re y 15 as C t % a e r e the Ph yim sicpia an ctV of isits thx e2 fin 013 ancial incentive (i.e., health insuran -0 c.e 0 0p 4re ** mium elimination) on HSA-eligible health objective of this study. Figure 5 contains the regression results that address this research question for individuals with As int r od . "uc Me ed dic parteion U vioustliy liz , a Fi tg ion ure a nd 3 pA re dsent heres t nche e in a ove H raell HS alth Sa A-ev ligib ings leA he cca ount lth p -E la ligib n enr le ollm Plan." ent A rm ateers for ican J the our tnw ao l o ef M mpa loy nae gre s d chronic condition afunbiased manner and not as an advocate fected choice of health plan. Those without a chronic condition were more likely to choose the HRA tende The d to Im foc paus ct on of C bo roa st-d Sh qa ure ing stions on H su ec ah a lth s Ca wrho e Pe ric nr eolls s, Qua in a n tC itD ieH s, a P, how nd Sp ee nr nding ollee s d Dyiff naem r ifr cs." om N non BER Wor -enrolle king es, r Pia sk pe r No. employees and their families, as the instance, Buchmueller and Feldstein (1997) and Buchmueller (1998) looked at the effect of moving to a fixed Primary Care Physician Visits x 2013 0.01 *** all research findings, publications, and news alerts. EBRI also extends its education and public service HRA is paired with a high-deductible reg p arla din, ng E a w BR or I km ere m ebe ssresn hi tp a ially nd ha /ors fir cont stri-bu dolla tions r t cov o E eB ra RgIe -E up RF s to hou a llim d be it di (i.e re., ctetd heto e E mBRI ployer Primary Care Physician Visits x 2014 -0.005 ** p 10lan enrollment, we used the experiences of another firm as a comparison group. Briefly, this other large, national Fi e ov m geur p rloy e the 1, P e ee-nt only erirce e nt c sov ta udy p g ee r aof ge er E . iod m Fip g . lur oy In Em ee r 6 p s Offe prlo esent ye rirng A s t C (he s onsu tudy g sam me e r r oup for -dir )te ,hose ce tnr edollm w Hit ea h fa elt nt h P m in t la ilyn, he cov bH ye S Fi rAar -g m ee ligib . Size For le, b he 20 rea 10 vlt ith p – y20 , only la 1n inc 6, Wit the re h P a m sed arin e o jfr ec ff otm e ions c t s Care 19 (12): e400-e407 (2013b). tha Th n ets hose e prew ve itntiv h oe ne s.e rvices are in addition to those that the Patient Protection and Affordable Care Act (ACA) mandates be • EBRI’s reliable health and retirement surveys are just a click away through the topic boxes at selec21632 tion, a ( nd Natthe ion im al B pa ur c Pe tr ia m of u o arC yD f E CH ac P re onomi s on Physov ic c ia e Re n ra V sea ll use isitr sc x h, of 20 20 1 serv 4 15)ic . e ht s a tpnd ://w spe ww nding .nbe. r.or Mor g/pa H e SA re p- c ee e rligib nt s/w2163 st 0le .udie 0 1he *** 2.p a s lt hh p d av f e la e n d xa 12 rm aw ine s in mor d moree contribution a role t the to improving Ameri University of Ca Plif reo s cr id an nia ens’ t (H UC finan arry ). Co B cial en for aw kn ea y owle 19 at94 the dge , a UC bov thro set i e addr ugh ts ecsont sits a , (20 ribut w 2)ard 6 io 59 n e --win 06 q70; ua ning e l to th -mp aubli il: ec c ona co st se w a of rvice y@ the ebr cam he i.ora glpaign t h plan or opponent of any position.” 11% cont Spr eibut cialis ion t Phty o sitche ian H VRA isits ), x a2 ft 0e 1r 3 which a deductible must be satisfied -out 0.00-1 of-pocket with after-tax dollars, and then company offered health insur®a nce plan options like those of Employer A. This choice set and its associated premiums of the financial incentive, and its interactive effects with health status, are reported in these tables. 4 percent Thr to oug 27 h 201 perce9 n t................................ for employee-only cov ................................ erage, and from 2 pe ................................................................ rcent to 31 percent for family coverage. In cont ........... rast 4 , 9% covered in full. Specialist Physician Visits x 2013 0.004 *** targeted the top of the page. quest Cions hoosetoSave about indivand idual he the a co lth e mpanio ngage n m site ent, ww suc w.c h ahoo s mset edic osav ation a e.od rg he rence for individuals with chronic healthy individuals, the other plans are wit Sh th pecie al is la t rP ghe ys sti cm ian em Vib se its rsx hip, 2014 but in 1994 UC reduced the employer -p 0.r0 e0m 1ium contribution to the amount charged by 10% insurance coverage takes over. This contrasts with HSA-eligible health plans, where employers can simply offer the 8% Gauthier, Anne K., and Carolyn M. Clancy. "Consumer-Driven Health Care—Beyond Rhetoric with Research and w The erefir not st st subst udy a to nte ia xa llym aine lte re enr d ollm overe tn he t in a four n -H ye Sa Ar- e st ligib udyl ep e he riod alth p . lan was conducted by Parente, Feldman and Editorial Board: Harry Conaway, publisher; Stephen Blakely, editor. Any views expressed in this publication and those of t8% he authors should not B Huc SAhm -eligib uelle ler , he Tho altm h p as C lan e S . p "nr e Dcoe o iallm lis a st e P nt h Fi y xe s ait cd ia E -n D m V olla pilo sity r s e P xrr 2 e B 0 m 1 (4 iu com m p Ca ont rison ribugtrion Low oup) cont er in Sp ue ending d the ?sl " ight H0 e.a 0 0 lt upw 5 h Af *** afa rdi rtsr a17 je c(t6) or: y 228 sha-r2 e35 d b y Prescription Drug Fills x 2013 -0.002 *** 11 conditions (Fronstin, Sepulveda and Roebuck 2014a), generic drug use (Fronstin a left nd wR itoe h a b uc highe k 20 r 14 av)e , rtahe ge lik lee vlih el o ood f ris of k. Fi the gur leea st 2, c Va ost ria lyb p lela Me n aavns aila ab t lB e ast seline atew ide (20 . 11 Th )is , b p yolic Gry o up, led C to ov ae high rage d Ty eg pree ................................................................ e of plan switching by UC employees. Of those ..... 7 he Th alt be h p e a res la c arre in bed a end t xc o t ele h pe o tions t tfhe fic er to w s, or t the ru kse tc ees ro d mp , m ecem a ide ra bb er for ilsity , o rhim rule othself er . s For poor n ins s ohe rst o a rs nc f t ehlf w ee E , em mp he plto loy he yee B e r rs to en ma op efity e R n a ma esea nd kre c hma fun Instc td ithi ua tng e, n HSA t hce E ontrib B . RI E ution ducast io tha n atn d a R ree search Experience." Health Se Presrcvriic pe tio s Re n Drs ue ga Fric lls h x39 20 1 (4 3) (Part II)): 1049-1054 (2004). -0.001 *** I Cf th hrist e ifin ans aon ncia (20 l in08 ce)n . tThey e ive to e xa nrm oll in inedt he da tH a S frAo-m eligib an e lem he ploy alth erp w lait n r h o ed ve uc r e 13 d 0,0 posit 00iv ceo v rie sk re s de liv lec ets in a ion, tll 50 st he coeff atie cie s. nt In 20 s on 06 the , • Need a number? Check out the EBRI Databook on Employee Benefits. the st (19 udy g 98).r oup prior to the premium being made free to employees. These figures provide clear evidence of a large Pre EBRI is sup scription Drug Fiported b lls x 2014 y organizations from all industries and sectors that appreciate the value of -0.002 *** 5% priceF u sh ndop , op r t ing heir s am tafong fs. N o indiv thingid he ua reils n i w s titoh ba e c C oD nsH tru Pe d ( a Brsot an- a Gtold temb pte trog a , ie d o t a r h l. in 20 de15 r th)e, aqdua optli io tn y o of f ahe ny p alt en h c dina gr le e grie slc ate io iv ne , r de g(uFr laon tionst , or in This leads to increased costs and employees whose premiums did not increase between 1993 and 1994, roughly 5–6 percent switched plans. The 5% To cond const itiona ru l c otn the a c o cntrib ompution aP ris reon s c bryg ip r the tioup on e D m ra utp g E loy Fm ills e p e x loy if 2 0 e d 1ro 4 ne B, tthro his ug anh aly a scis a fe us te eri da a p p larn. op e Furth nsitye rmo scorre e , -m 0e .0 m a0 tp c 1hin loy ***e gr sa p mpary o a co ch to e ntributns e u mo rere 5% Fi So gur ur ce e: Em 3, pl Eonr yee oll Bm ene efnt it R in esea Cons rch Ins um titut ee r- es dtir im eat ct es ed ba H sed ea olt n h P admla inis n, trab tiv ye E en m rop llm loy ene t r an , d Ccov laim er sa ge Type, 2011?2014 .............................. 7 int thee ream ctp ions loye w r offe ith trhe ed m eight easu d re iff s of erent he op alth st tions a, tus in cp luding resent a en d H in Figur SA and es 5 a two H nd RA 6 w -baould sed p bla e ns posit . The ive inv and est sig taa tis totric s c ally on cluded effect of the financial incentive on HSA-eligible health plan enrollment. unbiased, reliable information on emplo interpr5% etative rule, or as legal, accounting, actuarial, or other suchy pee b rofessienefits. onal advice. Vis www it .ewww.e bri.org bri.org/about/join/ for more. and Roebuck 2014b), and the combination of deductible size, presence of an HSA highe or H r RA pre , m and ium ty s, w pe hic of C h in DHP tu ( rH n aviland, movement to the fixed So curc ont e:r Em ibut ploion a yee Benef pprito R aes ch co earch nt Ins ribut titute e es dt im to ata esd bas ver ed se on sele admc inis tion, trativw e enro hich d llment rov and e cone laim sp lan out of the UC Greene, Jessica, Judith H. Hibbard, Anna Dixon, and Martin Tusler. "Which consumers are ready for consumer-directed to tNotes: ha the t the *** Hp<0.01, S tAs wo og f ** r no p<0.05, oup n-hi s w g *p<0.10 hly ere c o co mm pe pns ara ate ble d e on mp aloy veecets o.r of observed characteristics before the elimination of the HSA-eligible 4% Buchm • ueInstantly get e-m ller, Thomas C., ana dil noti Paul J. ?F cations of the latest EB eldstein. "The Effect of PrRI data, ice on Swit surveys, publications, and m ching among Health Plans." Jour eetings nal of 4% si tha gnif t w iche ant n b . Tot he h a est n HR imaA te a s la nd ra gn HSA ely reflect wer e the offe op re pd osit in a e. dAdm ition to o ong those thew r it cho h e ic m ep s— loy se uc eh -only as a c n HMO, overageP , Pindiv O, aidua nd ploint s wit -of h - data. Figure 4, Difference-in-differences Model Estimates of CDHP Entry ............................................................................ 8 et al. 2011). The most recent studies have examined plan enrollees’ behaviors over longer time periods. M odels include all covariates reported in Figure 4, but are sup pressed for brevity. reverberates with more future adverse A intppendi ernal marx ke — t. P The rior high L-it op etr ion p atur lan ha e d to raise premiums 49 percent in 1996 because of adverse selection. EBRI Issue Brief is registered in the U.S. Patent and Trademark Office. ISSN: 0887 -137X/90 0887 -137X/90 $ .50+.50 health plans?" Journal of Consumer Policy 29 (3): 247-262 (2006). health plan premium in the study group. Specifically, the analysis began by applying the inclusion criteria used for 12 Not su Hera plt r and sem ih singly, e Econom inr nars by clicking on the “Notify Me” or ic ol s le 16 es w (2)it : h e 231 m-p 24 loy 7 (1 ee-99 only 7). coverage were young “RSS” buttons at the t er, less likely to be mop of our hom ale, and generally e page. healthier highe servicre C (ha POS rls)on pla Cnom —to he Notes: rb idit HS*** A y I p<0.01, ande ttrax sc c** te p<0.05, d or ae s w p*p<0.10 ope ula re tle ion o ss lik f r ee ly la tto ive re ly spond healt hy to t w he or k fin erasnc , w iahil l in e ca e nt miv oree bgye ne ent re ous H ring R the A aH ttSA ra- cted Pre-tax dollars can be used to pay for these out-of 3% -pocket expenses if the employee has funds in either a flexible spending 2% selection ultimately resulting in the demise of the riskiest plans. In this study, eliminating premiums for the HSA-eligible E Finga gurg e ing 5, D cons ifferum enc ee rs in t -in-diff heeir r ehe nca elt s Mod h care e l Es has tim long ate s of been vi CDH ew P eEdnt bry y , pE la m n sp ploy onsor ee-os nly a s a Cov w ea ra yg of e ................................ dampening health car...... e 9 Employer0% A, the treated firm. Next, a probit model of group assignment (i.e., member of treated versus comparison M odels include all covariates reported in Figure 4, but are suppressed for brevity. than those with family coverage. Figure 2 reports all variable means at baseline (2011) by group and coverage type for e re ligib lativ le e ly he unh alth p eala lth n. y Mor work ee ov rse. r, The emp stloy udy a ees w lso itfou h hype nd tha rtetn in sion, comd ey w slipid as peosit miaiv , e dly ia b ae ss te os, a ciatnd ed a wst ith h e ma nr /C ollm OPD ent w e in t re he all HSA A ac c consi ount st (e FSA) nt fi o nding r HSA in t . he research literature (discussed in more detail below) is that CDHPs introduce positive risk Similarly, Cutler and Reber (1998) examined changes that took place at Harvard University. In 1995, Harvard moved to 2011 2012 2013 2014 Haviland, Amelia M., Neeraj Sood, Roland McDevitt, and M. Susan Marquis. "How Do Consumer-Directed Health Plans health plan did not reduce adverse risk selection. Charlson, M. E., P. Pompei, K. L. Ales, and C. R. MacKenzie. "A New Method of Classifying Prognostic Comorbidity in costs (Fronstin 2002). Despite the fact that CDHPs were introduced about 15 years ago, there are only about one dozen firm) was estima © 2 ted 0 1a 7s a , E m fun plo cy tion o ee Ben f t efhe it R follow esearcing h Inset stitu of te - bE aseline ducati o(n a 2011 nd R ) v ea se ria arb cle h F s: g und. ende Allr r , ia gh gte s, res tenur erve ed ( . number of the subjects used in the ensuing regression models. It is worth reiterating that—by design—all of these individuals were b Fi and e gtur wH e eRA e 6, D n 3 pla iff ans nd er.e 11 nc e p-ein rc -e dnt iffa eg re enc point es Mod s less el Es likteim ly atto e There’s lo s of enroll i CDn th HP E e ts more! nt HrSA y, -Fa eligib m ily le C he ova elt ra h p gela ................................ n as a result of the fin ................ ancial 10 selection when workers are given a choice of plans. Namely, healthier individuals are more likely to choose the CDHP, a fixed contribution for health benefits. Contributions were set at 85 percent of the least-costly policy for employees th Affect Vulnerable Populations?" Forum for Health E 8conomics and Policy 14 (2): 1-23 (2011). studie Long s thit audin t haa vl Stu e trie dd ie ts: D o unde everlop stam nd ent w ho and choo Valid ses a1100 13 tio su n." ch p Jour St lan reet n s, t al o NW · Suite 878 he f C fir hst ronic ste p D is in b eae se com 40ing (5)a : 37 mor 3-e 83 e nga (19g 8e 7) d. consumer. years with the employer), Charlson Comorbidity Index, as well as dummy variable indicators for the presence of six not enrolled in the HSA-eligible health plan in 2011. As mentioned earlier, the propensity score matching process incentive. And, individuals who had more primary care visits and prescription drug fills during the baseline year were Source: EBRI analysis based on administrative claims data. earning below $45,000 a year, 80 percent for employees earning between $45,000 and $70,000 a year, and 75 percent Washington, DC 20005 Visit EBRI online today: www.ebri.org The seminal work was presented at a conference on consumer-driven health care in September 2003, and those chronic conditions (congestive heart failure, hypertension (high blood pressure), diabetes, dyslipidemia, asthma/chronic yielded study and comparison groups that were very similar in terms of covariate means. (202) 659-0670 www.ebri.org www.choosetosave.org e e A e e e e e e e e e e e e eb b b b b b b b b b b b b b brr r r r r r r r r r r r r r re i. i. i. i. i. i. i. i. i. i. i. i. i. i. i.s o o o o o o o o o o o o o o oe r r r r r r r r r r r r r r rg g g g g g g g g g g g g g g a rIIIIIIIIIIIIIIIc s s s s s s s s s s s s s s sh s s s s s s s s s s s s s s s u u u u u u u u u u u u u u u re e e e e e e e e e e e e e e ep B B B B B B B B B B B B B B B or r r r r r r r r r r r r r rrief ief ief ief ief ief ief ief ief ief ief ief ief ief ief t fr o• • • • • • • • • • • • • • •m Ju Ju Ju Ju Ju Ju Ju Ju Ju Ju Ju Ju Ju Ju Ju th ly ly ly ly ly ly ly ly ly ly ly ly ly ly lye 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 EB ,,,,,,,,,,,,,,, R 2 2 2 2 2 2 2 2 2 2 2 2 2 2 20 0 0 0 0 0 0 0 0 0 0 0 0 0 0 I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 E7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 d u • • • • • • • • • • • • • • • c N N N N N N N N N N N N N N N ato o o o o o o o o o o o o o o ion ............... 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 a5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 nd Research Fund © 2017 Employee Benefit Research Institute 16 3 6 11 12 15 8 10 2 7 14 9 5 13 4