M. Arantxa Colchero · BMJ 2016 · longitudinal household panel study with fixed-effects counterfactual modeling · n=6,253 households (205,112 observations)

Beverage purchases from stores in Mexico under the excise tax on sugar sweetened beverages: observational study

Cited 835 times in the scientific literature.

Level 3 - non-randomized controlled study

Non-randomized longitudinal panel study evaluating a population policy intervention (graded Level 3 by design analogy).

OpenAlex W2225912682 · doi:10.1136/bmj.h6704 · record verified 2026-08-27

What was done

An observational study evaluated beverage purchase trends among an unbalanced panel of 6,253 households (205,112 observations) across 53 Mexican cities with populations exceeding 50,000 from January 2012 to December 2014. A difference-in-difference fixed-effects model adjusted for macroeconomic variables, pre-existing trends, household demographics (age and sex), and socioeconomic status (low, middle, high). Post-tax beverage volumes (mL/capita/day) purchased in 2014 were compared against counterfactual projections derived from pre-tax trends for both taxed and untaxed beverages.

What was found

Compared with the counterfactual in 2014, purchases of taxed beverages decreased by an average of 6% (-12 mL/capita/day), with the decline reaching 12% by December 2014. Reductions occurred across all socioeconomic groups but were largest in low socioeconomic status households, averaging a 9% decline during 2014 and reaching a 17% reduction by December 2014. Purchases of untaxed beverages rose by 4% (36 mL/capita/day) over the counterfactual, driven primarily by increased purchases of bottled plain water.

Why it matters

This evaluation provides real-world national evidence that fiscal policy through excise taxes can reduce consumer purchases of sugar-sweetened beverages, with the most pronounced reductions observed in lower-income households.

Limits

The study is observational and restricted to urban areas (>50,000 residents), omitting rural populations. It tracked household store purchases rather than individual consumption, total caloric intake, or downstream clinical health outcomes, and evaluated only one year of post-tax data.

Cited by