The impact of the unemployment rate on infant mortality rate in the US, 2000-2016: the system Generalized Method of Moments approach. Open Access

Shin, Sangsoo (Spring 2019)

Permanent URL:


Introduction: Infant mortality rate has been known to be an indicator of quality of healthcare in a nation and a proxy for overall population health. Infant mortality rate in the US has decreased over time but remains higher than infant mortality rate in developed countries. The unemployment rate has been considered as risk factor for mortality rate. But the effect of unemployment rate on infant mortality rate in the US has not been assessed. Our study tried to identify immediate association between the unemployment rate and Infant mortality rate in the US over recent time periods.

Methods: Infant mortality rates at state-level were extracted from the linked data of death and birth certificate at Center for Disease Control and Prevention. Linked data of 446,426 deaths with 69,158,041 births are collapsed to total 867 observations of 51 states in the US from 2000 to 2016 to create infant mortality rate by year and state. Annual unemployment rate at state-level was collected from Bureau of Labor Statistics. As covariates, annual state-level data of urbanization rate, the percentage of bachelor’s degree or higher and percentage of Black and African American were obtained from American Community Survey or Surveillance, Epidemiology, and End Results Program. The system Generalized method of moments analysis was used to evaluate the effect of the unemployment rate with dataset merged by year and state.

Results: The overall infant mortality rate in the United States decreased from 6.89 per 1,000 in 2000 to 5.87 in 2016, while the unemployment rate showed fluctuated trajectory. In unadjusted model, One percent point increase in the unemployment rate was associated with 0.019 increase in infant mortality rate (95% CI: -0.030, 0.068). After adjusting for other covariates, the strength of the association was halved, but increases in the unemployment rate continued to be associated with a small reduction in the IMR (0.101, 95% CI: -0.118, -0.085).

Conclusion: We found counterintuitive evidence that short-term economic downturns, measured as increases in the unemployment rate, were associated with increases in the IMR. The reasons for this paradox should be investigated in future analyses.

Table of Contents

Introduction 1

  Background 1

Research Aims 4           

Methods 5

Study Design 5

Study Variables 5

Study Methods 7


Results 9

Descriptive Analysis 9

Unadjusted Analysis 11

Two-fixed model 12

The system Generalized Method at Moments 12


Discussion 13

Findings 13

Limitation and Strength 16

Conclusion and Public Health Implication 17


List of Reference 19


Supplement 1. Trends of each variable over time 24

Supplement 2. DAG in the model 27

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
Subfield / Discipline
  • English
Research Field
Committee Chair / Thesis Advisor
Last modified

Primary PDF

Supplemental Files