[PPT] - Health Risk and Insurance Over the Lifecycle Juergen Jung Chung PowerPoint Presentation

SLIDE 1

Health Risk and Insurance Over the Lifecycle

Juergen Jung Chung Tran Towson University, Maryland Australian National University CBE - Conference May 2015

Jung and Tran (TU and ANU) Health Risk 2015 1 / 41

SLIDE 2

Disclaimer

This project was supported by the Agency for Healthcare Research and Quality (AHRQ, Grant No.: R03HS019796), the Australian Research Council (ARC, Grant No.: CE110001029), CBE-Towson University Summer Research Support, and funds from the Centers for Medicare & Medicaid Services, Office of the Actuary (CMS/OACT). The content is solely the responsibility of the authors and does not represent the official views of the funding institutions.

Jung and Tran (TU and ANU) Health Risk 2015 2 / 41

SLIDE 3

Introduction

U.S. Medical vs. Non-Medical Consumption

20 30 40 50 60 70 80 90 2000 4000 6000 8000 10000 12000 14000 16000 Age $ Medical vs. Non−Medical Consumption over the Life Cycle Non−Medical from F−K(2007) Medical from J−T(2010) Total

Figure 1: MEPS 1996-2007

Jung and Tran (TU and ANU) Health Risk 2015 3 / 41

SLIDE 4

Introduction

Healthcare Financing in OECD Economies

Figure 2: OECD (2004)

Jung and Tran (TU and ANU) Health Risk 2015 4 / 41

SLIDE 5

Introduction

U.S. Health Spending

Social Security Medicare and Medicaid Other Federal Non-interest Spending

1962 1972 1982 1992 2002 2012 2022 2032 2042 2052 2062 2072 2082

5 10 15 20 25 30 35

Percentage of GDP

Figure 3: CBO (2010)

Jung and Tran (TU and ANU) Health Risk 2015 5 / 41

SLIDE 6

Introduction

Our Research Program

1 Develop macroeconomic models with micro-foundations of health 2 Analyze economic aspects of health-related behavior 3 Study implications of healthcare policies

The distributional effects: health inequality and wealth inequality The macroeconomic aggregates and welfare.

Jung and Tran (TU and ANU) Health Risk 2015 6 / 41

SLIDE 7

Introduction

This Paper

1 Develop a stochastic dynamic general equilibrium overlapping

generations model with

1

endogenous health expenditures and insurance choice

2

and a realistic structure of health insurance systems

that accounts for the patterns of

1

health expenditures and insurance over the life cycle

2

the distribution of income and health expenditures observed in the data

2 Goal: quantify the effects of social health insurance on 1

macroeconomic aggregates and

2

welfare.

Jung and Tran (TU and ANU) Health Risk 2015 7 / 41

SLIDE 8

Introduction

Results Preview

Lifecycle health risk induces demand for health insurance Private health insurance: very limited

Competitive markets fail to insure lifecycle health risk Introduction of market regulations improves

insurance coverage (up to 70%) small welfare gains(2.3%)

Public (social) health insurance: important

The European-style heath insurance system

Aggregate output loss (9%) but large welfare gains (5.5%)

The American-style insurance system: Mix of public and private insurance

Aggregate output loss (7.5%) and welfare gain (between 3 and 4%)

Jung and Tran (TU and ANU) Health Risk 2015 8 / 41

SLIDE 9

Introduction

Related Literature

1 Micro-health economics

Grossman (1972a,1972b), Grossman (2000) Pauly(1974), Rothschild and Stiglitz (1976) Besley (1989), Selden (1993), Blomqvist and Johansson (1997)

2 Quantitative macroeconomics/public finance

Ayagari (1994), Imrohoroglu et al (1995), Hugget (1996)

3 Macro-health economics:

Exogeneous health expenditure shocks: Kotlikoff (1988), Leven (1985), Palumbo (1999), Attanasio, Kitao and Violante (2008), Jeske and Kitao (2009), Pashchenko and Porapakkarm (2010), Janicki (2011) Endogenous health expenditures and insurance: Suen (2006), Feng (2009), and Jung and Tran (2008, 2010, 2013)

Jung and Tran (TU and ANU) Health Risk 2015 9 / 41

SLIDE 10

The Model

MODEL

Jung and Tran (TU and ANU) Health Risk 2015 10 / 41

SLIDE 11

The Model

The Model: Bewley (1986) and Grossman (1972)

Overlapping generations model with

heterogeneous agents

lifespan: age 20 to 90 idiosyncratic shocks: labor productivity and health shocks

health capital accumulation

health as consumption and investment goods endogenous health spending endogenous health insurance choice

Maket structure: goods, capital, labor markets, and incomplete financial markets Government-run health insurance systems Dynamic stochastic general equilibrium

Jung and Tran (TU and ANU) Health Risk 2015 11 / 41

SLIDE 12

The Model

The Model: Preferences and technology

Preferences: u (cj, lj, hj) Health capital: hj = h

mj, hj−1, δh, ǫh

j

Human capital (“labor”):

ej = e

ϑ, hj, ǫl

j

Health and labor income shocks:

Pr

ǫh

j+1|ǫh j

∈ Πh

j and Pr

ǫl

j+1|ǫl j

∈ Πl

j

Jung and Tran (TU and ANU) Health Risk 2015 12 / 41

SLIDE 13

The Model

The Model: Health Insurance Arrangements

Private health insurance Public (social) health insurance Health insurance status: inj =

    

if no insurance 1 if private insurance 2 if public insurance

Jung and Tran (TU and ANU) Health Risk 2015 13 / 41

SLIDE 14

The Model

The Model: Out-of-pocket Health Spending

Agent’s out-of-pocket health expenditures depend on insurance state

(mj) =

  

pinj

m × mj,

if inj = 0 ρinj

pinj

m × mj

,

if inj > 0

Jung and Tran (TU and ANU) Health Risk 2015 14 / 41

SLIDE 15

The Model

The Model: Technology and Firms

Final goods C production sector for price pC = 1: max

{K, L} {F (K, L) − qK − wL}

Medical services M production sector for price pm: max

{Km, Lm} {pmFm (Km, Lm) − qKm − wLm}

pm is a base price for medical services Price paid by households depends on insurance state: pinj

j

=

1 + νinj

pm νinj is an insurance state dependent markup factor Profits are redistributed to all surviving agents

Jung and Tran (TU and ANU) Health Risk 2015 15 / 41

SLIDE 16

The Model

The Model: Household Problem - Timing

+ a: asset

+ h: health capital + in: insurance

Shocks:

+ h: health + l: productivity

Choices:

+ c: consumption + l: leisure + m: medical services + a’: savings + in’: insurance + a’: asset + h’: health capital + in’: insurance

x(j) = {j, a, h, in,

h, l }

x(j+1) = {j+1,a’, h’, in’, h’, l’} choice = {c, l, m, a’, in’}

Shocks: + h’: health + l’: productivity

Jung and Tran (TU and ANU)

Health Risk 2015 16 / 41

SLIDE 17

The Model

Insurance Sector

1 + ωin

J

j=1

µj 1[inj(xj)=1]

1 − ρin

mmj (xj)

dΛ (xj)

= R

J

j=1

µj 1[inj(xj)=1]premin dΛ (xj) ,

Jung and Tran (TU and ANU) Health Risk 2015 17 / 41

SLIDE 18

The Model

Government Budget

G + T SI + T Med =

J

j=1

µj τ Cc (xj) + taxj (xj)

dΛ (xj) ,

where T SI =

J

j=1

µj

tSI

j (xj) dΛ (xj) and

T Med =

J

j=1

µj 1 − ρMed pMed

m

mj (xj) dΛ (xj) −

J

j=1

µj

premMed (xj) dΛ (xj) .

Jung and Tran (TU and ANU) Health Risk 2015 18 / 41

SLIDE 19

The Model

Pensions and Bequests

Pensions:

J

j=J1+1

µj

tSoc

j

(xj) dΛ (xj) =

J1

j=1

µj

τ Soc × (ej (xj) × lj (xj) × w) dΛ (xj)

Accidental Bequests:

J1

j=1

µj

tBeq

j

(xj) dΛ (xj) =

J

j=1
˜

µjaj (xj) dΛ (xj)

Jung and Tran (TU and ANU) Health Risk 2015 19 / 41

SLIDE 20

The Model

A Competitive Equilibrium

Given the transition probability matrices and the exogeneous government policies, a competitive equilibrium is a collection of sequences of distributions of household decisions, aggregate capital stocks of physical and human capital, and market prices such that Agents solve the consumer problem The F.O.Cs of firms hold The budget constraints of insurances companies hold All markets clear All government programs and the general budget clear The distribution is stationary

Jung and Tran (TU and ANU) Health Risk 2015 20 / 41

SLIDE 21

Calibration

CALIBRATION

Jung and Tran (TU and ANU) Health Risk 2015 21 / 41

SLIDE 22

Calibration

Parameterization and Calibration

Goal: to match U.S. data pre-ACA (before 2010) Data sources:

MEPS: labor supply, health shocks, health expenditures, coinsurance rates PSID: initial asset distribution CENSUS: demographic profiles Previous studies: income process, labor shocks, aggregates

Jung and Tran (TU and ANU) Health Risk 2015 22 / 41

SLIDE 23

Calibration

The U.S. Health Insurance System

Mixed system:

Private health insurance for working population

Individual based health insurance (IHI) Group based health insurance (GHI)

Public health insurance

Medicare for retirees Medicaid for the poor: 2/3 is retirees

Key Facts:

Low coverage: 47 million uninsured in 2010 (≈ 15%) High cost: 16% of GDP on health in 2010 and close to 20% by 2015

Jung and Tran (TU and ANU) Health Risk 2015 23 / 41

SLIDE 24

Calibration

Moment Matching: Health Expenditures

✶✶ ❆♣♣❡♥❞✐① ❊✿ ❋✐❣✉r❡s

❋✐❣✉r❡ ✶✿ ▼♦♠❡♥t ♠❛t❝❤✐♥❣ ✉s✐♥❣ ▼❊P❙ ✷✵✵✵✲✷✵✵✾✿ ❤❡❛❧t❤ s♣❡♥❞✐♥❣ ❛♥❞ ✐♥s✉r❛♥❝❡ t❛❦❡ ✉♣ r❛t❡s ♦✈❡r t❤❡ ❧✐❢❡✲❝②❝❧❡✳ ✻✵

Figure 4: Moment matching: Model vs. Data

Jung and Tran (TU and ANU) Health Risk 2015 24 / 41

SLIDE 25

Calibration

Moment Matching: Insurance Take-up Rates

✶✶ ❆♣♣❡♥❞✐① ❊✿ ❋✐❣✉r❡s

❋✐❣✉r❡ ✶✿ ▼♦♠❡♥t ♠❛t❝❤✐♥❣ ✉s✐♥❣ ▼❊P❙ ✷✵✵✵✲✷✵✵✾✿ ❤❡❛❧t❤ s♣❡♥❞✐♥❣ ❛♥❞ ✐♥s✉r❛♥❝❡ t❛❦❡ ✉♣ r❛t❡s ♦✈❡r t❤❡ ❧✐❢❡✲❝②❝❧❡✳ ✻✵

Figure 5: Moment matching: Model vs. Data

Jung and Tran (TU and ANU) Health Risk 2015 25 / 41

SLIDE 26

Calibration

Income Distribution

50

100 150 0.0 1.0 2.0 Income in $1,000 frequency in %

Income distribution SS1 with FPL

Data

Model

FPL 133FPL 400FPL MaidFPL Maid133FPL Maid400FPL

Figure 6: Moment matching: Model vs. Data

Jung and Tran (TU and ANU) Health Risk 2015 26 / 41

SLIDE 27

Calibration

Calibration: Matched Moments

Moments Model Data Source

Medical expenses HH income

17.6% 17.07% CMS communication

Workers IHI

6.7% 7.6% MEPS 1999/2009

Workers IHI

62.2% 63.6% MEPS 1999/2009

Workers Medicaid

9.0% 9.2% MEPS 1999/2009

Capital output ratio: K/Y

2.9 2.6 − 3 NIPA

Interest rate: R

4.2% 4% NIPA

Size of Social Security: SocSec/Y

5.9% 5% OMB 2008

Size of Medicare: Medicare/Y

3.1% 2.5 − 3.1% U.S. Department of Health 2007

Payroll tax Social Security: τ Soc

9.4% 10 − 12% IRS

Consumption tax: τ C

5.0% 5.7% Mendoza et al. (1994)

Payroll tax Medicare: τ Med

2.9% 1.5 − 2.9% Social Security Update (2007)

Total tax revenue/Y

21.8% 28.3% Stephenson (1998) and BarroSahasakul (1986)

Medical spending profile

see figure

Medical spending distribution

see figure

Insurance take-up ratios

see figure Total number of moments

Jung and Tran (TU and ANU) Health Risk 2015 27 / 41

SLIDE 28

Analysis

EXPERIMENTS

Jung and Tran (TU and ANU) Health Risk 2015 28 / 41

SLIDE 29

Analysis

Experiments

1 Construct a benchmark economy with no health insurance for

comparison

2 Introduce alternative insurance regimes 1

Private insurance

2

Public health insurance

3

A mix of private and public health inusurance

3 Quantify the macroeconomic and welfare imiplications Jung and Tran (TU and ANU) Health Risk 2015 29 / 41

SLIDE 30

Analysis

Private Insurance

Unregulated market (IHI)

Price discrimination: age and health status prem = prem(j, hj) No government subsidy

Regulated market (GHI)

No price discrimination: community rating Premium payment is tax deductible

Jung and Tran (TU and ANU) Health Risk 2015 30 / 41

SLIDE 31

Analysis

Private Insurance: Aggregates and Welfare

[1] No Ins. [2] Private Health Insurance Only (a) Unregulated (b) Regulated Insured (%) 0.00 8.03 70.62 + IHI (%) 0.00 9.76 0.00 + GHI (%) 0.00 0.00 70.62 Capital (K) 100.00 100.15 99.40 Output (GDP) 100.00 100.36 100.75 Welfare 0.00 −0.26 2.31

Table 1: The Effects of Private Health Insurance.

Jung and Tran (TU and ANU) Health Risk 2015 31 / 41

SLIDE 32

Analysis

Public Insurance

The European/Canadian/Australian health insurance system

Mandatory membership Open enrollment Community rating financed by payroll or consumption tax

Jung and Tran (TU and ANU) Health Risk 2015 32 / 41

SLIDE 33

Analysis

Public Insurance: Aggregate and Welfare

[1] No Ins. [3] Public Health Insurance Only (a) Medicaid for all (b) Medicare for all Insured (%): 0.00 100.00 100.00 + Medicaid (%) 0.00 100.00 00.00 + Medicare (%) 0.00 00.00 100.00

Cons. tax - τC (%)

4.31 23.36 17.02 Capital (Kc) 100.00 86.11 86.71 Output (Yc) 100.00 90.42 91.41 Welfare 0.00 5.59 5.81

Table 2: The Effects of Social Health Insurance

Jung and Tran (TU and ANU) Health Risk 2015 33 / 41

SLIDE 34

Analysis

Mix of Private and Public Insurance

The U.S. health insurance system

Partial coverage Private insurance for workers Public insurance for retirees and the poor

Two arrangements

Pre-ACA Post-ACA

Jung and Tran (TU and ANU) Health Risk 2015 34 / 41

SLIDE 35

Analysis

Mix of Private and Public Insurance

[1] No Ins. [4] Public and Private Ins. (a) Pre-ACA (b) After-ACA Insured (%): 0.00 76.69 99.58 + IHI (%) 0.00 6.53 22.46 + GHI (%) 0.00 60.60 63.08 + Medicaid (%) 0.00 9.56 14.04 + Medicare (%) 0.00 17.68 17.68 Capital (Kc) 100.00 85.72 84.86 Output (Yc) 100.00 92.40 90.55 Welfare 0.00 4.06 3.71

Table 3: The effects of mixed public and private health insurance systems

Jung and Tran (TU and ANU) Health Risk 2015 35 / 41

SLIDE 36

Analysis

Consumption Variation

20 30 40 50 60 70 80 90 Age 10 15 20 25 30 35 40 %

Skill group 1 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 5 10 15 20 25 30 35 40 %

Skill group 2 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 5 10 15 20 25 30 35 %

Skill group 3 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 8 10 12 14 16 18 20 22 24 26 %

Skill group 4 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

Coefficient of variation of C by skill group Jung and Tran (TU and ANU) Health Risk 2015 36 / 41

SLIDE 37

Analysis

Out-of-Pocket Health Spending Variation

20 30 40 50 60 70 80 90 Age 40 60 80 100 120 140 160 180 %

Skill group 1 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 40 60 80 100 120 140 160 180 %

Skill group 2 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 40 60 80 100 120 140 160 180 200 %

Skill group 3 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 40 60 80 100 120 140 160 180 200 220 %

Skill group 4 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

Coefficient of variation of OOP expenses by skill group Jung and Tran (TU and ANU) Health Risk 2015 37 / 41

SLIDE 38

Analysis

Health Capital Variation

20 30 40 50 60 70 80 90 Age 0.05 0.10 0.15 0.20 0.25 0.30 %

Skill group 1 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 0.05 0.10 0.15 0.20 0.25 0.30 %

Skill group 2 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 0.05 0.10 0.15 0.20 0.25 0.30 %

Skill group 3 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 0.05 0.10 0.15 0.20 0.25 0.30 %

Skill group 4 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

Coefficient of variation of H capital by skill group Jung and Tran (TU and ANU) Health Risk 2015 38 / 41

SLIDE 39

Analysis

Insurance Take-up Variation

20 30 40 50 60 70 80 90 Age 5 10 15 20 25 30 35 40 %

Skill group 1 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 5 10 15 20 25 30 35 %

Skill group 2 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 5 10 15 20 25 30 35 %

Skill group 3 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 70 80 90 Age 5 10 15 20 25 30 35 40 %

Skill group 4 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

Coefficient of variation of income by skill group Jung and Tran (TU and ANU) Health Risk 2015 39 / 41

SLIDE 40

Analysis

Labor Variation

20 30 40 50 60 Age 0.4 0.5 0.6 0.7 0.8 0.9 1.0 %

Skill group 1 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 Age 0.3 0.4 0.5 0.6 0.7 0.8 %

Skill group 2 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 Age 0.2 0.3 0.4 0.5 0.6 0.7 0.8 %

Skill group 3 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

20 30 40 50 60 Age 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 %

Skill group 4 NoIns 1_U.S. 4_PublicII 5_Private 6_Private_GHI

Coefficient of variation of Labor by skill group Jung and Tran (TU and ANU) Health Risk 2015 40 / 41

SLIDE 41

Conclusion

1 Construct a heterogeneous agents macro-model with health as a

durable good

2 Account for lifecycle patterns of health expenditures and private

insurance take up rates

3 Quantify the macroeconomic and distributional effects of different

health insurance systems

Jung and Tran (TU and ANU) Health Risk 2015 41 / 41