Sie sind auf Seite 1von 33

Done by

Chan ?unxln
Llleen Soh ?lllng
Lln !lehul
Cng 1ahong
lm of Lhe paper
W 1o provlde evldence on Lhe effecL of Medlcald
ellglblllLy expanslons on prlvaLe healLh
lnsurance coverage
JhaL ls Medlcald
W Medlcald ls an uS publlc healLh lnsurance
program for Lhe poor
W Covers Chlldren and regnanL women
Cvervlew
lnLroducLlon
CrowdouL based on lndlvldual ellglblllLy
Lmployer lnduced crowdouL
1oLal crowd ouL esLlmaLes conslderlng famlly
declslons
CrowdouL esLlmaLes lncludlng condlLlonal
coverage
Concluslon and crlLlque
1radeoff beLween more generous
lnsurance and all oLher goods
Plgh valuaLlon
Low valuaLlon
Low valuaLlon
1radeoff beLween more generous
lnsurance and all oLher goods
Plgh valuaLlon
Low valuaLlon
Low valuaLlon
Jeaker red|ct|on
Cn average lndlvlduals made
ellglble for medlcald w||| reduce
Lhelr prlvaLe coverage
CompllcaLlons
,ost r|vate hea|th |nsurance |s
prov|ded through emp|oyment
- CurrenL lnsurance pollcy need noL
lndlcaLe lnsurance pollcy
- Jorkers may noL recelve Lhe
savlngs from forgolng employer
provlded coverage
- Lmployers may encourage
workers Lo drop coverage/
employees dropplng coverage
Lhemselves
nsurance dec|s|ons are made at a
fam||y |eve|
- Lack of dlsLlncLlon among
dependenLs
key po|nt 1hese comp||cat|ons may |ncrease or decrease the crowdout
aLa
CurrenL opulaLlon Surveys (CS) for
1988 Lo 1993
Drawback 1
oes noL
lndlcaLe
wheLher a
woman ls
pregnanL
Drawback 2
oes noL have
lnformaLlon
abouL asseLs
Drawback 3
CuesLlons abouL
lnsurance
coverage may
be a mlxLure of
lnformaLlon
from pasL year
and currenL
Drawback 4
May noL
accuraLely
dlsLlngulsh
Medlcald
coverage of Lhelr
famlly members
and Lhemselves
Jhy ls Lhere a worry for crowdlngouL?
n lncrease of
92
n lncrease of
238
Jhy ls Lhere a worry for crowdlngouL?
Lllglble ln 1987 |ess
||ke|y Lo have prlvaLe
coverage
Made ellglble beLween
1987 1992 more
||ke|y Lo have prlvaLe
coverage
of Lhe nonelderly
populaLlon have prlvaLe
coverage
Jhy ls Lhere a worry for crowdlngouL?
uggest|ve but not def|n|te |e other factors (eg recess|on |n 1990s) m|ght
cause dec||ne |n pr|vate coverage
lrsL esLlmaLes of crowd ouL
egress|on ,ode| 1
lndlcaLor of lnsurance
coverage elLher
Medlcald rlvaLe or
unlnsured
varlable measurlng Lhe
ellglblllLy of lndlvldual l
for Medlcald
seL of
demographlc
conLrols
1wo dummy
varlables ln SLaLe
and ?ear
lrsL esLlmaLes of crowd ouL
egress|on ,ode| 1
Measures Lhe marglnal
Lakeup raLe for pub||c
|nsurance/ measures Lhe
crowdouL effecL for
other forms of
|nsurance coverage
Measures Lhe
demographlcs effecLs on
coverage
lrsL esLlmaLes of crowd ouL
LsLlmaLes of
Chlldren may be
blas
Lg lnfanLs are generally
more llkely Lo be ellglble
SoluLlon
lnclude age dummy
varlables for chlldren
Lndogenous
facLors
Lg CosLs of prlvaLe
lnsurance reflecLed ln
lower wages (and hence
hlgher lmpuLed ellglblllLy)
SoluLlon
SelecL a naLlon random
sample of 300 chlldren
and 3000 women of chlld
bearlng age each year
geyear and stateyear |nteract|ons show s|m||ar resu|ts
lrsL esLlmaLes of crowd ouL
- 0074/0233
- 31
educt|on ln prlvaLe
coverage due Lo lncrease
Medlcald coverage
- 1 0119/0233
- 49
ercenL of |ncrease |n
,ed|ca|d noL assoclaLed
wlLh a reduct|on |n the
un|nsured
lrsL esLlmaLes of crowd ouL
- Cver 100 crowdouL
- Lesspreclse
8oLh
esLlmaLes
- JhlLes households wlLh a male
head households wlLh fewer
people are more ||ke|y Lo have
prlvaLe coverage
ln all
cases
1hree ,echan|sms Lhrough whlch employerprovlded
coverage could fa|| as Medlcald ellglblllLy |ncreases
1 Lmployers
may reduce Lhe
generoslLy of
Lhelr lnsurance
as Medlcald
expands
2 Jorkers may
choose Lo
decllne coverage
for Lhemselves
and Lhelr famlly
3 Jorkers may
choose Lo
decllne coverage
for Lhelr
dependenLs
Lmployer lnduced crowd ouL
Jeaker pred|ct|on ssum|ng that a|| fam|||es va|ue |nsurance equa||y
and have the opt|on of equ|va|ent pr|vate coverage
Lmployer lnduced crowd ouL
PealLh
lnsurance
unlLs"
robablllLy of
a worker
havlng
employer
coverage
ConLrol for Lhe
demographlc and famlly
sLrucLure of Lhe Plu
Westr|ct samp|e to workers aged 24 to 64
W1ake |nto account of b|as and Lndogenous factors
Lmployer lnduced crowd ouL
W 1he share of fam||y med|ca| spend|ng that |s e||g|b|e for ,ed|ca|d Coverage
W Lach Plu conslsLs of a famlly head spouse and any chlldren under 19
Wk lndexes slngleyear age groups for chlldren and broader age groups for adulLs
LxpecLed healLh
Spendlng ln a
year for LhaL age
group
LllglblllLy for
Medlcald of
famlly members
ln age group k
no of famlly
members ln LhaL
age group
Lmployer lnduced crowd ouL
W estr|ct to whether coverage |s offered and taken up
WLmployee 8eneflLs SupplemenLs Lo Lhe CS
no effecL of Medlcald
ellglblllLy on Lhe
declslon Lo offer
lnsurance
LsLlmaLes suggesL LhaL all of Lhe reducLlon ln worker coverage ls comlng Lhrough lower
Lakeup raLes not reduced employer offerlng of lnsurance
Lmployer lnduced crowd ouL
W Lffects of ,ed|ca|d L||g|b|||ty on own and dependents' coverage for men w|th
dependents potent|a||y e||g|b|e for ,ed|ca|d
WMarrled male workers aged 2464 who have a spouse of chlldbearlng age or chlld
ln addlLlon Lo Lhe 017 of workers who drop employer coverage anoLher 007 are
dropplng dependenLs coverage and shlfLlng Lo lndlvldual
tronger effects of
Medlcald pollcy on Lhls
group
|zab|e |ncrease ln
lndlvldual coverage for
workers maLched by drop
|n dependents' coverage
1oLal crowd ouL esLlmaLes conslderlng
famlly declslons
Cwn
dollar
CLher
dollar
Plu
dollars
1oLal crowd ouL esLlmaLes conslderlng
famlly declslons for chlldren
Medicaid coverage
- 02 Private coverage
1 (Own
dollars)
Medicaid coverage
- 016 Private coverage
1 (Other
dollars)
1oLal crowd ouL esLlmaLes conslderlng
famlly declslons for women
- 0064 Medlcald coverage
- 032 rlvaLe coverage
1 (Own
dollars)
- 0S2 Medlcald coverage
- 013 rlvaLe coverage
1 (Other
dollars)
1oLal crowd ouL esLlmaLes conslderlng
famlly declslons
W CLher dollars have a sLronger effecL on
Medlcald coverage
8easons 8las
1oLal crowd ouL esLlmaLes conslderlng
famlly declslons
ConverL resulLs lnLo populaLlon flgures Lo esLlmaLe Lhe LoLal crowdouL
- lncrease ln Cwn dollar
ellglblllLy (8792)
Own
Dollar
- lncrease ln oLher
dollar ellglblllLy(8792)
Other
Dollar
1oLal crowd ouL esLlmaLes conslderlng
famlly declslons
1oLal crowd ouL esLlmaLes conslderlng
famlly declslons
Crowd out est|mates for ch||dren 06m]1Sm
40
Crowd out est|mates for women of ch||dbear|ng age
08m]07m
100
Crowd out est|mates for other adu|ts 30
1ota| Crowd out est|mates 77
CrowdouL esLlmaLes lncludlng
condlLlonal coverage
CondlLlonal
Coverage
Chlldren who
reporL
Lhemselves
unlnsured
regnanL women
who reporL
Lhemselves
unlnsured
LsLlmaLed value of condlLlonal coverage
Wshare of hosplLal spendlng by chlldren and pregnanL women
Wllkely be covered by Medlcald
CrowdouL esLlmaLes lncludlng
condlLlonal coverage
lnal crowdouL esLlmaLe 49
33 mllllon
persons wlLh
Medlcald
17 mllllon
persons wlLh
prlvaLe coverage
Concluslon of Model
W Jhlle Lhere ls evldence of crowdlngouL lL mlghL
noL be enLlrely bad should Medlcald expanslons
lmprove healLh ouLcomes
W Some suggesLlons Lo counLer crowdouL
W slldlng scale subsldy
W lmposlng a walLlng perlod beLween when a
person loses prlvaLe lnsurance and when Lhey
become ellglble for Medlcald
W owever crowdout |s |nev|tab|e s|nce ,ed|ca|d
|s offered for free
CrlLlque
W 1rends noL accounLed for ln Lhe regresslon
W lfferenL ways of deflnlng crowdouL yleld
dlfferenL resulLs
W LlmlLaLlon of Lhe daLa
W lfflculLy ln observlng dlrecL relaLlon beLween
changes ln prlvaLe coverage and Medlcald
expanslons
lnnovaLlve aspecLs of Lhe paper
W lnsLrumenLal varlable approach and conLrol
for age wlLh dummles respecLlvely
W 1he auLhors used a proxy Lo measure varlous
values of esLlmaLes
W examlned crosssLaLe dlfferences ln Medlcald
ellglblllLy and publlc and prlvaLe lnsurance
W conLrols for famllysplllovers as Lhey recognlze
LhaL lnsurance ls a famlly declslon

Das könnte Ihnen auch gefallen