Association for Public Policy Analysis & Management

Black people are two to three times more likely to drown than white people. This is an important health disparity, because drowning is a leading cause of accidental death among African American males 5-14 years of age. Despite overall declines in drowning rates for both blacks and whites, the disparity in drowning rates has actually increased over time. In this project, we examine the impact of the labor market for life guards, policies requiring lifeguards in private pools, and policies requiring fencing around pools on drowning rates and drowning disparities.

We are compiling a dataset that includes information on county-level drowning rates (by race, sex, and age), the number and wages of life guards of different races by PUMA, local laws regarding the requirement of life guards to be present at private pools (e.g., hotel pools), and local building requirements for fencing around pools. We currently have data on drowning rates from 1968 to up to 1988 for all counties in the U.S., and Florida counties from 1970 to present. We also have data on life guards by race within each PUMA 1971 to 2015. We have hired two RAs who will be collecting data on local laws and policies regarding life guards and pool fencing requirements.

We have two main objectives in this project. First, we will assess the role of discrimination in the life guard labor market on drowning disparities. Recent research by Myers and Lai (forthcoming) shows that at every level of lifeguarding, the black drowning rate is higher when lifeguards are white than when lifeguards are black. That is, black life guards are better at saving black swimmers from drowning than white lifeguards. Using cross-sectional data for multiple years, we will estimate log-earnings equations and employment for black and white lifeguards and perform Oaxaca decompositions to measure discrimination in labor markets for lifeguards. We will then compute the portion of the racial gap in drowning rates that can be attributed to racial discrimination in labor markets for lifeguards and simulate how many black drownings could have been averted had there been no unexplained racial disparities in lifeguard labor markets.

Second, we will assess the role of specific laws and policies on drowning rates and drowning disparities. Some cities or counties require that private facilities (e.g., a hotel) provide life guards at their pool, and these requirements have become more common over time. Similarly, there is a great deal of geographic and temporal variation in fencing requirements for residential pools. We will assess if these requirements are effective at reducing overall drowning rates and if they reduce or increase disparities in drowning rates. To do this, we will first use within-county temporal variation to assess if there is a reduction in drowning rates after a county implements or strengthens a law. Because policies about life guards or fencing may be caused by a recent drowning, we also compare drowning rates in neighboring counties who have different laws.