Research shows that Mexico’s recent health reforms appear to have considerably reduced catastrophic and out-of-pocket health spending on both inpatient and outpatient medical procedures, especially among the poor. The study, Public policy for the poor? A randomised assessment of the Mexican universal health insurance programme, determined that enrollment in the national health insurance program significantly reduced the proportion of poor households that suffered from catastrophic health  expenditures from 9.9 percent to 6.9 percent—a reduction that translates into major savings for thousands of Mexican families. IHME researchers collaborated with researchers at Harvard University, Princeton University, Washington University, and Mexico’s Secretaria de Salud, Instituto Nacional de Salud Publica, and Conestadistica.

Research findings

Households that were assigned to a control group (i.e., not encouraged to participate in the health program known as Seguro Popular) reported low affiliation with the program (7%), whereas 44% of households that were persuaded to participate in the program reported being affiliated. Enrolling in the program was related more to the wealth of the geographic area than to the wealth of individual families, so people with more assets living in poor areas were as likely to enroll in the program as their poorer neighbors.

Enrolling in Seguro Popular reduced by 23% the proportion of all respondents experiencing catastrophic expenditures. The reduction was 55% among compliers (individuals who adhered to whatever treatment status they were assigned to). Seguro Popular also reduced out-of pocket expenses for inpatient and outpatient medical care, especially for the poorest individuals. Contrary to expectations and previous observational research, the researchers found no effects on medication spending, health outcomes, or utilization.

The majority of enrollees (69%) in the program rated the quality of health services as very good or good, and 97% planned to enroll again in the follow-up period. These findings were similar for all asset clusters.

Analytical approach

Researchers identified geographical cluster regions throughout 13 states in Mexico. Each cluster included a health clinic or hospital and the population in its catchment area. One cluster from paired health clusters was randomly assigned to a treatment group, with families persuaded to enroll in Seguro Popular and states encouraged to improve health facilities and increase medical personnel and drug supplies. Another cluster was assigned to a control group, with no change in enrollment or service delivery and the usual for-pay health care.

Baseline and follow-up household surveys within each cluster were conducted to collect family spending data (including health care expenditures) and individual data such as health risk factors, health self-assessments, and blood pressure. Outcomes were examined based on several subgroups, including asset class (low versus high), sex, and age group.

Research objective

Assessing the effect of Seguro Popular allows policymakers and the public to understand the impact of that investment, which aims to deliver health insurance, health care, medicine, and health facilities to 50 million uninsured Mexicans. By conducting rigorous evaluations into the impact of health reform efforts such as Seguro Popular, IHME aims to help inform the design and implementation of new health policies in other countries.

Recommendations for future work

The study was unable to determine Seguro Popular’s impact on a number of other important outcomes beyond reductions in catastrophic spending on medical procedures, such as the impact on utilization of services or coverage of key interventions. The researchers note that additional analyses based on a longer assessment period are needed to determine other program effects.

The affiliation program could be improved, especially in the enrollment of poor families living in relatively wealthy communities. Expanding the study to include parallel experiments with new groups in additional health clusters in other parts of the country would strengthen generalizations to the nation. Finally, although Seguro Popular is unprecedented in scope, the researchers could only test the effects of the program as a whole rather than each component, which might yield additional areas for improvement.


King G, Gakidou E, Imai K, Lakin J, Moore RT, Nall C, Ravishankar N, Vargas M, Téllez-Rojo MM, Hernández Ávila JE, Hernández Ávila M, Hernández Llamas H. Public Policy for the poor? A randomised assessment of the Mexican universal health insurance programme. The Lancet. 2009 Apr 9; 373:1447–1454.