MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Healthcare Spending Inequality:

Evidence from Hungarian Administrative Data

Anikó Bíró, Dániel Prinz

MTA KRTK KTI

1 April 2019

MTA KRTK KTI

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Introduction

• New evidence on the distribution of healthcare spending in HungarySample: full-time workers

• Universal health insurance, but

• Health varies with income• Geographic inequalities in the availability of healthcare services

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Data

• Administrative data, covering years 2003-2011 on a random 50%sample of the 2003 population aged 5-74

• Health expenditures are observed on the annual level

• Location is observed in 2003

• Sample restrictions: individuals aged 18-55, earning at least theminimum wage each month and not receiving any public transfers

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Methods

• Maps of healthcare spending for 3 types of spending (outpatient,inpatient, prescription drugs)

• Heterogeneity by income:• Relate the current year's healthcare expenditures to the previous

year's income• Calculate income percentiles (deciles)• Income of those who spent zero days on sick-leave the previous year• Adjustment for age and gender

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Outpatient expenditures

(9958,11465](8596,9958](8159,8596](7574,8159][6610,7574]

77% di�erence between highest and lowest spending counties

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Inpatient expenditures

(8406,8773](7921,8406](7663,7921](7357,7663][7069,7357]

27% di�erence between highest and lowest spending counties

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Prescription drug expenditures

(12821,14265](12439,12821](12178,12439](11653,12178][10353,11653]

37% di�erence between highest and lowest spending counties

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Outpatient expenditures

7,00

08,

000

9,00

010

,000

11,0

00O

utpa

tient

Spe

ndin

g (H

UF)

0 20 40 60 80 100Percentile of Wage

34% di�erence between the 90th and 10th percentile

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Inpatient expenditures

5,00

06,

000

7,00

08,

000

9,00

010

,000

Inpa

tient

Spe

ndin

g (H

UF)

0 20 40 60 80 100Percentile of Wage

33% di�erence between the 90th and 10th percentile

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Prescription drug expenditures

10,0

0012

,000

14,0

0016

,000

Rx

Spen

ding

by

Soci

al In

sura

nce

(HU

F)

0 20 40 60 80 100Percentile of Wage

27% di�erence between the 90th and 10th percentile

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Outpatient expenditures

6,00

08,

000

10,0

0012

,000

0 2 4 6 8 10decile

Budapest Győr-Moson-SopronSzabolcs-Szatmár-Bereg

If we eliminated cross-county variation, variance in spending would bereduced by 71%

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Inpatient expenditures

4,00

06,

000

8,00

010

,000

12,0

00

0 2 4 6 8 10decile

Budapest Győr-Moson-SopronSzabolcs-Szatmár-Bereg

If we eliminated cross-county variation, variance in spending would bereduced by 47%

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Prescription drug expenditures

10,0

0012

,000

14,0

0016

,000

0 2 4 6 8 10decile

Budapest Győr-Moson-SopronSzabolcs-Szatmár-Bereg

If we eliminated cross-county variation, variance in spending would bereduced by 43%

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data

MTA KRTK KTI

MethodsGeographic heterogeneity

Income heterogeneityIncome heterogeneity by county

Discussion

Next steps: explanations

• Health factors? � Next wave of linked administrative data

• Access to care?• Number of hospital beds, physicians, pharmacies?• Distance to healthcare facilities?• Information, connections, informal payments?

Anikó Bíró, Dániel PrinzHealthcare Spending Inequality: Evidence from Hungarian Administrative Data