Health Shocks and Disability Transitions among Near-Elderly Workers

Transcription

1 Healh Shocks and Disabiliy Transiions among Near-Elderly Workers David M. Culer, Ellen Meara, Seh Richards-Shubik The research was suppored by a gran from he U.S. Social Securiy Adminisraion (SSA) as par of he Reiremen Research Consorium (RRC). The findings and conclusions expressed are solely hose of he auhors and do no represen he views of SSA, any agency of he federal governmen, he affiliaions of he auhors, or he Cener for Reiremen Research a Boson College.

2 Moivaion Fiscal oulook need for reform Enormous heerogeneiy in response o a major healh shock among near-elderly workers 12% apply for DI wihin 4 years, 60% coninue FT work 27% of high school drop-ous apply, 21% of blacks How do individuals respond o healh shocks? Why do some ake SSDI, ohers don?

3 Two Broad Theories Healh capial In a perfec world, you only receive DI benefis if healh is oo poor o work Fewer papers emphasize healh: Bound e al. (2010), Meara and Skinner (2011), Culer, Meara, R-S (2011) Labor supply Repl. raes (Parsons 1991) Recessions, demand for low-skill workers (Auor and Duggan 2003, 2006) Healh benefis (A & D) Allowance raes (Burkhauser e al. 2001; Maesas e al. 2011; French and Song 2011)

4 Our Conribuion Focus on dynamic response o well measured, exogenous healh shocks Preliminary analysis How imporan are hese rapid healh declines in ransiion o DI among near-elderly workers? Main analysis How and why he response o healh shocks differs across groups? Draw on healh capial and labor supply heories Sronges evidence is for effec of high earnings

5 Healh & Reiremen Sudy sample: All waves from Age (censored a age 65) Full-ime workers prior o healh shock Have ~14,500 male, ~12,500 female personwave observaions on ~10,500 individuals Use rich daa on healh condiions, funcional limiaions, work, earnings and oher income, healh insurance, household members

6 Defining Healh Shocks Follow Jim Smih (1999) HRS asks abou a series of healh condiions: Has a docor ever old you ha you have? New diagnoses define shocks Major shocks: cancer, lung disease, hear disease, sroke, or psychiaric condiion Minor shocks: hyperension, diabees, or arhriis More objecive han self-repored healh saus or a condiion ha limis abiliy o work, less objecive han physical exam (e.g., NHANES)

7 Healh shocks among full-ime workers (age in year ): New diagnosis beween year and +2 Males Females Major healh shock Cancer Lung disease Hear disease Sroke Psychiaric condiion Minor healh shock Hyperension Diabees Arhriis

8 Preliminary Analysis: Healh Shocks in DI Transiion Prob s Esimae regressions for fuure SSDI (or SSI) applicaion/receip among full-ime workers Jus as a funcion of demographics: Pr( DI FT + k 1 ) = π demog + π Then add healh and economic variables: Pr( DI + k FT ) = β1hshock+ 2 + β Hsock + β econ β hhold + β demog β

9 Timing in models Working (year ) DI saus? (+2) DI saus? (+4) Shock occurs ( : +2)

10 Effec of healh shocks is large: Conrol variables: new diagnosis o +2 Males DI in Females DI in Major healh shock *** *** *** *** [0.0086] [0.0118] [0.0105] [0.0159] Minor healh shock *** ** [0.0039] [0.0066] [0.0039] [0.0075] Mean of dep. var. (DI +k ) Models include age, year, census division, occupaion and indusry dummies; race and Hispanic ehniciy, marial saus, # of hh members; exising and new healh diagnoses, # of ADLs & IADLs; earnings and income quiniles, healh insurance, and healh requiremens for job. SEs in [ ] s.

11 Change in demographic variables when healh & econ facors are added: Conrol variables Educaion < 12 years years 16 + years Males (+4) Basic model Full model ** [0.0098] *** [0.0060] *** [0.0054] * [0.0100] [0.0063] [0.0073] Females (+4) Basic model Full model *** [0.0108] [0.0068] *** [0.0052] *** [0.0110] [0.0067] [0.0060] Black [0.0091] [0.0093] *** [0.0101] *** [0.0098] Hispanic *** [0.0078] *** [.0080] [0.0098] [0.0112] Models include age, year, census division, occupaion and indusry dummies; race and Hispanic ehniciy, marial saus, # of hh members; exising and new healh diagnoses, # of ADLs & IADLs; earnings and income quiniles, healh insurance, and healh requiremens for job. SEs in [ ] s.

12 Main Analysis: Differenial Response o Healh Shocks Healh capial more likely o apply for DI if Low iniial healh sock Bigger healh decline (worse shock) Greaer healh requiremens a available jobs Labor supply applicaion depends on Prices (wages, healh insurance) Non-labor income (spouse, reiree benefis) Preferences for work vs. leisure

13 Regressions for SSDI (or SSI) applicaion/receip afer healh shock We esimae he following regressions, separaely for men and women: Pr( DI + k FT, Hshock + 2) = β1hdiagd, β Hsock 2 + β Hreqs 3 + β prices 4 + β income 5 + β demog 6 Same variables as before, organized in erms of he wo heories Resricing o workers wih healh shocks is like ineracing major shock wih all variables

14 Resuls Fracion applying/receiving afer 4 years: 12.4% males, 13.1% females Healh sock no consisen effecs of exising condiions, bu maybe ADLs (+5 o 10%) Type of shock srokes are relaively severe (+15% vs. hear disease) No clear effecs of healh requiremens a job High earners less likely o apply (-3 o -10% in op 2 quiniles), low earning males more likely Some evidence for high unearned income

15 Wha have we learned? Major healh shocks are srong predicors of ransiion o DI among full-ime workers Healh differences appear o accoun for differenial beween college and high school grads No so for high school drop-ous or race differenial Our economic variables do no srongly predic ransiion o DI among near-elderly workers (bu no exacly a fair comparison, need economic shocks) In erms of differenial response o healh shocks among near-elderly workers Some suppor for price effec and income effec in a sandard labor supply decision Lile consisen evidence on healh capial effecs

16 Wha can we do wih his? Accoun for differenial arrival of healh shocks by educaion when hinking abou ineracion of reiremen and disabiliy policies Raising he reiremen age or limiing disabiliy benefis will have unfavorable equiy implicaions Considering age in eligibiliy decision could help o offse some of his adverse disribuional effec Provide earnings suppor for a-risk workers before hey decide o apply for SSDI e.g., workers wih ADLs To exen ha healh insurance affecs he response o shocks, healh reform may help