U.S. flag

An official website of the United States government

Skip Header


A Random-Effects Approach to Attrition Bias in the SIPP Health Insurance Data

Written by:
Working Paper Number SEHSD-WP1991-11 or SIPP-WP-144

Abstract

Using black male health insurance data from the 1984 SIPP Panel as an example, this paper describes a test for the presence of attrition bias and a consistent estimator for the level of health insurance coverage in the presence of attrition bias. The test and estimator jointly model the attrition and health insurance coverage processes using random effects panel probit models in which the random effects are allowed to be correlated. The empirical results suggest that for black males, ignoring attrition bias leads to a positive time trend for health insurance coverage, when the true, corrected for attrition bias, time trend is negative.

Page Last Revised - October 8, 2021
Is this page helpful?
Thumbs Up Image Yes Thumbs Down Image No
NO THANKS
255 characters maximum 255 characters maximum reached
Thank you for your feedback.
Comments or suggestions?

Top

Back to Header