Venue: The Fuqua School of Business, Duke University, 1 Towerview Drive, Durham, NC 27708-0120

 

Presentation

The Impact of Seguro Popular on Catastrophic Health Expenditures in Mexico: A Comparison of Data Sources and Methodologies

Authors: Aarón Salinas (Mexican National Institute of Public Health); Sergio Sesma-Vasquez (Carso Institute for Health); Sandra Sosa-Rubí (Mexican National Institute of Public Health); Omar Galárraga (Mexican National Institute of Public Health)

Presenter: Omar Galárraga (Mexican National Institute of Public Health (INSP))

Discussant: Claudio Lucarelli (Cornell University)

Session: Paying for Health Care

Room: Classroom G

When: Monday 1 p.m. - 2:30 p.m.

Seguro Popular is the largest health insurance scheme for the otherwise uninsured population in Mexico. An example of federal-level and state collaboration, the program started as a pilot in 2001 and expanded afterwards throughout the country. One of its main goals is to improve the financial protection of the uninsured population against excessive health expenditures. Seguro Popular covers a variety of basic preventive and curative procedures, as well as medicines, and hospital care for the poorest segment of the Mexican population. Families not covered in the traditional Social Security health system are eligible to enroll in Seguro Popular. A randomized evaluation of the impact of the program was conducted in 2005.

This paper estimates the impact of Seguro Popular on catastrophic health expenditures as well as out-of-pocket expenditures from three different sources: 1. Encuesta de Impacto del Seguro Popular [SP Impact Evaluation Survey]; 2. Encuesta de Gasto e Ingreso de los Hogares (ENIGH 2006) [Expenditures and Income Survey]; and 3. Encuesta Nacional de Salud y Nutrición (ENSANut 2005-2006) [National Health and Nutrition Survey].

For the first (experimental) data source, we used an intention to treat (ITT) as well as the complier average causal effect (CACE) estimation. For the second and third (non-experimental) data sources, given that they are observational, cross sectional studies, we used linear and non-linear methods using several robust instrumental variables (IV) to correct for the self-selection bias in the Seguro Popular enrollment equation.

We find that Seguro Popular has reduced the probability of households with catastrophic health expenditures using the SP Impact Evaluation Survey and National Health and Nutrition Survey; but we do not find such effect using the Expenditures and Income Survey. On the other hand, we find that the probability of out-of-pocket expenditures is reduced when we use ENIGH and ENSANut as sources, but not when we use the experimental data.

The paper discusses the differences in data collection methods among the three surveys, as well as the differences in the methodologies employed for the estimation of the average treatment effect. While the Expenditures and Income Survey is stronger in terms of more accurately measuring income and expenditures given is longer field-time and questionnaire extent, the National Health Survey is more comprehensive in terms of the health-related data that can be used as covariates in the estimation of the outcome equation.

The paper provides further evidence of the impact of Seguro Popular on financial protection outcomes in Mexico, and the usefulness of more inclusive insurance coverage to prevent excessive health and out-of-pocket expenditures. The weak associations of the outcomes of interest with the Seguro Popular household affiliation in some of the surveys but not in others confirm the need to use multiple data sources and alternative methodologies to test a causal effect.