Explaining the Rise in U.S. Cesarean Rates: The Role of Labor Market Factors

The U.S. cesarean rate has risen dramatically in the last generation, from 6% of all births in 1970 to 33% in 2011.  This increase remains poorly understood.  The explanatory factors that have been thoroughly studied, including the legal environment, physician training, prenatal care, source of payment, and financial incentives, are not powerful enough to explain a large fraction of this increase.  Other potentially relevant influences, however, have not been closely studied.  Exploiting a detailed, underutilized, national survey of birthing mothers, this project focuses on the role of one understudied set of factors: labor market factors, including family income, maternal employment, and education.

The relevance of these factors is motivated by the standard model of health capital, which is accumulated via investments of money, time, and knowledge.  Women with greater health capital should be less likely to require a cesarean delivery.  Maternal employment lowers the time invested in health capital, while greater family income and education increase the financial and knowledge base that can be used to develop health capital during pregnancy.  These effects should persist even after accounting for the presence of health insurance.

The empirical analysis is executed in two steps.  First, the causal effect of labor market factors on the incidence of cesarean delivery is estimated, using the most extensive survey available for this purpose: the 1988 National Maternal Infant Health Survey.  The linear probability model utilized relates the probability of a cesarean delivery to a large set of demographic, clinical, and labor market variables, including the parents’ education, pre-pregnancy employment status, rate of pay, and health insurance coverage.

Second, the coefficients from the first stage regression are applied to means of these variables for new mothers in each of the years 1968-2011.  These means are calculated using the March Supplement of the Current Population Survey, the nation’s largest survey of labor market activity, supplemented with Census Bureau data for the employment patterns of first time mothers.  Each first-stage coefficient, multiplied by the second-stage difference in that variable’s mean between any two years, yields the estimated contribution of that factor to the change in cesarean rates over that period.

The main findings are as follows.  Labor force attachment has a significant, positive effect on the delivery method of first-time mothers (but not mothers who have previously given birth).  The coefficients on specific labor market variables strongly support the relevance of the health capital model here, and the effect sizes are large.  Compared to a woman who has never worked, a woman who works before, during, and after her pregnancy has an increased cesarean risk, of almost ten percentage points.   

The aggregate change in cesarean rates explained by these labor market factors, however, is modest: only about two percentage points between 1968 and 2011.  But the combined effect of labor market and demographic factors together is about six percentage points, almost one-quarter of the rise in cesarean rates over this period.