Marital Disruption and the Risk of Losing Health Insurance Coverage (Text Version)

Slide presentation from the AHRQ 2010 conference.

On September 28, 2010, James Kirby made this presentation at the 2010 Annual Conference. Select to access the PowerPoint® presentation (350 KB). Free PowerPoint® Viewer (Plugin Software Help).


Slide 1

Marital Disruption and the Risk of Losing Health Insurance Coverage

Marital Disruption and the Risk of Losing Health Insurance Coverage

James Kirby
AHRQ

Slide 2

Research Questions

Research Questions

  • To what extent does marital disruption increase the risk of losing health insurance?
  • Does the association diminish over time?

Slide 3

Motivation

Motivation

  • Many individuals get their health insurance coverage through their spouse.
  • Many marriages end in divorce.
  • Prior studies on coverage do not examine change.
  • Understanding how events affect coverage loss is important to evaluate changes in health care policy.

Slide 4

Theoretical Background

Theoretical Background

  • Being married provides:
    • Opportunity for coverage.
    • Incentives for coverage.
  • People plan their lives around their marital status:
    • Career, fertility, geographic location.
  • Decisions have staying power.
  • Two types of effects:
    • "Stable" disadvantage of being unmarried.
    • "Transitional" disadvantage of becoming unmarried.

Slide 5

A Model of Change

A Model of Change

Image: Graph depicts the risk of losing insurance over time.

Notes: In this graph, time is along the X axis and the risk of probability of loosing coverage is along the Y axis. The three lines represent, respectively, individuals who are married continuously over the observation period (red), unmarried continuously (yellow), and those who experience the dissolution of their marriage during the study period (green).  The difference between the red and yellow lines represents the "stable" effect of marital disruption, or the fact that being married provides some protection from loosing health insurance coverage.  The difference between the green and yellow lines at its widest is a measure of the "transitional" effect of marital disruption.  Note that those who experience marital disruption are not only more likely to loose health insurance than those who stayed married, but also more likely to loose coverage than those who were not married to begin with. Finally, as time passes, individuals adjust to their new marital status and, consequently, the risk of loosing health insurance coverage diminishes.

Slide 6

Analytic Approach

Analytic Approach

  • Discrete-time, Proportional Hazards Model
  • General Linear Model with complimentary log-log link (instead of probit or logit)
  • Unit of Analysis: Person-Month
  • Results interpreted with Hazard Ratios

Slide 7

Data Sources

Data Sources

  • Medical Expenditure Panel Survey (MEPS):
    • Analysis is restricted to 18+.
  • 2 years of data collected at 5 interview rounds.
  • Three most recent panels are pooled.
  • Unit of Analysis: Person-month.

Slide 8

Key Variables

Key Variables

  • Change in Insurance Coverage by Month
  • Change in Marital Status:
    • Married in ALL prior rounds
    • Unmarried in ALL prior rounds
    • Divorced/separated/widowed in ANY prior round
    • Unmarried → Married in ANY prior round
  • Interaction between change in marital status and time since change

Slide 9

Controls for Multivariate Analysis

Controls for Multivariate Analysis

  • Demographics:
    • Race, Ethnicity, Age, Sex
  • SES:
    • Income, Education
  • Attitudes about health insurance & risk:
    • Scale of four measures (a=0.78)
  • Health Status:
    • Subjective mental and general health
    • Disability (ADLs and IADLs)
    • Number of serious chronic conditions
  • Children in household

Slide 10

Hazard Ratios for Coverage Loss by Marital Status

Hazard Ratios for Coverage Loss by Marital Status

Marital Status/TimeHazard Ratio
Marital Status & Change
Married → Married (reference) 1
Married → unmarried2.20* 
Unmarried → Unmarried 1.15* 
Unmarried → Married 1.96* 
Time Interactions
Married → unmarried x time0.97* 
Unmarried → Married x time0.96* 

Slide 11

Monthly hazard of losing health insurance coverage by marital status

Monthly hazard of losing health insurance coverage by marital status

Image: A line graph shows the monthly hazard of losing health insurance coverage by marital status. Hazard for those married continuously or unmarried continuously remains ~.01 and decreases slightly over a 24-month period. Hazard for those married → unmarried (month 3) rises sharply from ~.01 to over .02 at month 4, then decreases to ~.01 over the next 20 months.

Slide 12

Monthly hazard of losing health insurance coverage by marital status

Monthly hazard of losing health insurance coverage by marital status

Image: A line graph shows the monthly hazard of losing health insurance coverage by marital status. Hazard for those married continuously or unmarried continuously remains ~.01 and decreases slightly over a 24-month period. Hazard for those married → unmarried (month 3) rises sharply from ~.01 to over .02 at month 4, then decreases to ~.01 over the next 20 months. Hazard for those unmarried → married (month 3) also rises sharply from ~.01 to over .02 at month 4, then decreases again over the next 20 months.

Slide 13

Monthly hazard of losing health insurance coverage by marital status

Monthly hazard of gaining health insurance coverage by marital status

Image: A line graph shows the monthly hazard of gaining health insurance coverage by marital status. Hazard for those married continuously begins at ~.04 and decreases slightly over a 24-month period to ~.03. Hazard for those unmarried continuously begins at ~.035 and decreases to under .03 over the 24-month period. Hazard for those married → unmarried (month 3) remains at ~.04 and decreases slightly over the 24-month period. Hazard for those unmarried → married (month 3) rises sharply from ~.035 to over .05 at month 4, then decreases again over the next 20 months to under .02.

Slide 14

Summary

Summary

  • Marital Disruption is positively associated with the risk of losing coverage, but so is marriage.
  • The associations decline over time:
    • For every month that passes, hazard decreases by about 3%.
  • Getting married is also positively associated with gaining insurance coverage. There is no such association for marital disruption.

Slide 15

Conclusions

Conclusions

  • Marital transitions should be recognized as a risk factor for losing health insurance.
  • Eligibility is not the sole causal mechanism:
    • Upheaval associated with marital change is a possible explanation for lapses in insurance coverage.

Slide 16

Hazard Rate Models: Hazard Ratios

Hazard Rate Models: Hazard Ratios

Marital StatusHazard Ratio
LossGain
Married → Married (reference) 1.001.00
Married → Unmarried2.20* 0.95 
Unmarried → Unmarried 1.15* 0.85* 
Unmarried → Married 1.96* 1.63* 
Married → Unmarried x time0.97* 1.01* 
Unmarried → Married x time0.96* 0.96* 

Notes: Here are the results from the hazard rate model. Results suggest that the risk of losing health insurance coverage is greatly elevated immediately following marital disruption. Specifically, the hazard of losing health insurance coverage for those who experience marital disruption is 2.20 times greater than that for those who stay married. Just as important, the hazard of losing coverage for those who experience marital disruption is 1.91 times greater than that for those who were not married to begin with (2.20/1.15). Results also suggest that the association between marital disruption and the risk of losing health insurance diminishes with time. Specifically, for each month following the marital disruption, the elevated hazard of losing insurance coverage decreases by 3% (HR=0.97).

While this is as expected, there is something quite surprising. Not only does marital disruption raise the hazard of losing coverage, so does getting married, and by about the same proportion. What could be happening here? It is possible that marriage brings with it other major transitions. For example, married couples frequently move, perhaps change jobs as a result, purchase a home, start planning for kids, among other things. It is possible that he resources and time required by these activities means that maintaining insurance coverage takes drops a notch on people's priority lists, at least temporarily.

In separate analysis, getting married is also positively associated with gaining insurance coverage. It associated both with gaining insurance if you have none, or losing insurance if you are covered.

Current as of December 2010
Internet Citation: Marital Disruption and the Risk of Losing Health Insurance Coverage (Text Version). December 2010. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/news/events/conference/2010/kirby/index.html