Wealth, Income, and the Affordability of Health Insurance

1
Wealth, Income, and the Affordability of Health
Insurance

• Didem Bernard, Ph.D.
• Jessica Banthin, Ph.D.
• and
• William Encinosa, Ph.D.

2
The Affordability Puzzle(adults 21-64)

• Many unafforders---the poor who appear to be
  unable to afford insurance--- do indeed purchase
  insurance
• 20.6 of those below the poverty line had
  private insurance in 2006 (3.8 million)
• Many afforders are uninsured
• 20.7 of those above the poverty line were
  uninsured in 2006 (32.5 million)

3
Research Questions

• What is the difference in wealth between insured
  and uninsured families?
• How much better can we predict demand for
  insurance using asset and wealth data?

4
Data

• Medical Expenditure Panel Survey (MEPS)
• Nationally-representative sample of households
• Full-year files for 2002 2003
• Detailed information on health insurance status,
  employment, health insurance offers, private and
  public coverage, health risk, and income
• Data collected during 5 rounds of interviews over
  2½ years (asset data in Round 5), covering a two
  year reference period

5
Study Sample and Key Variables

• Sample Nonelderly families w/o public insurance
  
• (age 21-64) (N23,951)
• Dependent variable Indicator equal to 1 if
  there is at least one person with private
  insurance in the family (insurance is
  point-in-time estimate measured as of the end of
  year)
• Family health insurance eligibility units
• Key variables family-level income
  family-level net worth

6
Regression Samples

• Employer coverage market N16,172
• Someone in family has an ESI offer
• Individual market N 7,779
• No one in family has an ESI offer

7
Asset and Debt Data in MEPS

• Ownership, value and amount owed for 10 types of
  assets and debt
• Home
• Second residence
• Other real estate
• Business/farm
• Vehicles
• Recreational vehicles
• CDs, stocks, bonds, mutual funds
• IRAs, Keogh plans, 401K accounts
• Checking, savings money market accounts
• Any other savings or assets (jewelry, annuity,
  trust or estates, collections for investment
  purposes)
• Debt (credit card balances, medical debt, loans
  from relatives, etc.)

8
Measures of Financial Assets Net Worth

• Financial assets CDs stocks bonds mutual
  funds IRAs Keogh plans 401K accounts
  checking accounts savings accounts money
  market accounts
• Net worth financial assets home second
  residence other real estate business/farm
  vehicles other savings or assets - debt

9
Two Empirical Models of Enrollment in Private
Insurance

• Standard Income Model (OLS)
• HIa by controls
• Wealth Model (OLS)
• HIa b1y b2 wealth controls
• where y income

10
Control Variables

• Preferences
• Health insurance is not worth the money it
  costs.
• Im more likely to take risks than the average
  person.
• Im healthy enough that I really dont need
  health insurance.
• Health risks (1) poor physical or mental health,
  (2) chronic conditions.
• Age, sex, race, education, occupation, married,
  family size, region.

11
Employer Coverage Market Individual Market

• 69 of nonelderly families had access to
  employer-sponsored coverage
• (67 million)
• 31 of nonelderly families were potentially in
  the individual market
• (30 million)

12
Differences between the employer coverage market
the individual market

• 96.1 of families with ESI offers vs. 10.6 of
  families without ESI offers have private
  insurance
• 15.2 in the employer coverage market vs. 62.2
  in the individual market are poor or low income

13
Median wealth holdings among nonelderly
familiesEmployer coverage market
14
Median wealth holdings among nonelderly
familiesIndividual market
15
Differences in asset holdings by income and
insurance status

• Median net worth of privately insured families
  was 23.2 times that of the uninsured
• Among families w/ access to employer coverage,
  median net worth of privately insured families
  was 15.4 times that of the uninsured
• Among families in the individual market, median
  net worth of privately insured families was 34.6
  times that of the uninsured

16
Percentage of nonelderly families that own
assets / have positive net worthEmployer
coverage market
17
Percentage of nonelderly families that own
assets / have positive net worthEmployer
coverage market
18
Differences in asset ownership by income and
insurance status(ALL NONELDERLY POPULATION)

• Among families with private insurance, 83.0
  owned financial assets vs. 48.0 among the
  uninsured
• Among families with private insurance, 89.8 had
  positive net worth vs. 68.4 among the uninsured

19
Estimated effects of wealth on private insurance
enrollmentEmployer coverage market
20
Estimated effects of wealth on private insurance
enrollmentIndividual market
21
Actual and predicted private insurance enrollment
ratesEmployer coverage market
22
Actual and predicted private insurance enrollment
ratesIndividual market
23
The role of wealth in private insurance
enrollment simulation results

• The standard income model performs relatively
  well for the employer coverage market.
• In the individual market, the wealth model
  performs significantly better.
• The standard model overestimates enrollment for
  those with low wealth and underestimates
  enrollment for those with high wealth.
• Standard model estimates are misleading for two
  subpopulations low income and high wealth, high
  income and low wealth.

24
Discussion

• The difference in wealth b/w the insured and the
  uninsured is not fully revealed by income
  comparisons
• Median income of the privately insured was 2.9
  times the median income of the uninsured
• Median net wealth of the privately insured was
  23.2 times the median net wealth of the uninsured
• This discrepancy is even larger among families in
  the individual market
• Median income of the privately insured was 2.3
  times the median income of the uninsured
• Median net wealth of the privately insured was
  34.6 times the median net wealth of the uninsured