Impacts of Health Insurance on Infant Health and Utilization: An Instrumental Variables Approach

Empirical estimates of the effects of private insurance on health can be confounded by selection 
problems and moral hazard. And the cost of giving birth in New Jersey is among the highest in the 
nation. Yet outcomes in New Jersey are often worse than other states. Understanding the effects of 
greater spending on newborn care and whether outcomes are affected by increased utilization is pivotal 
for keeping costs low for expectant parents and understanding minimum insurance mandates for 
maternal health. This study investigates the causal effects that health insurers have on medical and 
financial outcomes for children by leveraging a Coordination of Benefits rule which assigns the primary 
insurer of a dependent child to the parent whose birthday is first in the calendar year. Using exogenous 
variation from this “birthday rule”, we aim to measure whether health care spending differs across 
insurers, whether outcomes differ, and whether approaches to treatment vary. Typically, selection by 
patients into plans and by plans about which employers they contract with complicates the ability of 
researchers to answer these questions; our quasi-experimental approach allows us to circumvent this 
selection problem and provide novel, credible causal estimates of these effects. We propose 
accomplishing this study by leveraging the unique ability to link the New Jersey Birth Data with Uniform 
Billing (UB) Data through the New Jersey iPHD.