World Trade Organization membership and changes in noncommunicable disease risk factors: a comparative interrupted time-series analysis, 1980–2013

Krycia Cowling, Elizabeth A. Stuart, Roni Neff, Daniel Barstow Magraw, Jon S. Vernick, Keshia M. Pollack PorterBulletin 2018 · DOI

TL;DR

Joining the World Trade Organization (WTO) appears to be associated with both positive and negative shifts in dietary and consumption patterns linked to noncommunicable diseases (NCDs). Specifically, WTO accession correlates with an immediate increase in fruit and vegetable supply, but also with gradual increases in tobacco and alcohol availability, highlighting a complex interplay between global trade and public health.

What Problem Does This Paper Address?

Noncommunicable diseases (NCDs) are a growing global health crisis, with unhealthy diets, alcohol consumption, and tobacco use being major contributors. While globalization and trade liberalization are suspected drivers of increased availability of these risk factors, robust evidence linking specific trade agreements, like WTO membership, to changes in these consumption patterns has been limited. This study aims to fill that gap by quantitatively assessing how joining the WTO impacts the domestic supply of various commodities associated with NCDs.

Methods

The researchers employed a "natural experiment" design, treating WTO accession as an event that allows for a comparative interrupted time-series analysis. They compared trends in commodity supply between countries that joined the WTO after 1995 (the "exposed" group) and those that remained non-members (the "unexposed" group) from 1980 to 2013.

Key methodological elements include:

  • Propensity Score Weighting: To address potential pre-existing differences between WTO members and non-members, propensity scores were calculated to estimate the probability of WTO membership based on baseline commodity supply and other covariates. These scores were then used to weight the unexposed group, aiming to create a more comparable control group.
  • Interrupted Time-Series Analysis: Separate linear regression models were used for nine commodity categories (tobacco, alcohol, fruits/vegetables, nuts/seeds/legumes, seafood, red meats/animal fats, starches, sugars, and edible oils). The models incorporated WTO membership as a treatment variable and a "treatment*year" interaction term to assess changes in both the level and trend of commodity supply after accession.
  • Covariate Adjustment: Models were adjusted for GDP per capita, urban population percentage, and, for alcohol, the percentage of the Muslim population. For tobacco, ratification of the Framework Convention on Tobacco Control (FCTC) was also included as a covariate.

Datasets

The study utilized data on the domestic supply of commodities from national commodity balance sheets and food balance sheets, sourced from the Food and Agriculture Organization. Other data sources included:

  • World Development Indicators (World Bank) for GDP per capita and urban population.
  • United Nations Population Division for population data.
  • Institute for Health Metrics and Evaluation for GDP per capita.
  • Pew Research Center for the percentage of the Muslim population.
  • United Nations Treaty Collection for FCTC ratification dates.

The analysis covered 21 countries/territories joining the WTO between 1996 and 2008, and 26 non-member countries as of 2011.

Key Insights and Ideas

  • Trade Liberalization's Double-Edged Sword: The most striking insight is that WTO membership appears to simultaneously promote both protective (fruits/vegetables) and harmful (tobacco/alcohol) consumption patterns. This suggests that trade agreements have complex and often conflicting impacts on public health.
  • Immediate vs. Gradual Changes: The study highlights a difference in the timing of supply changes. Fruits and vegetables saw an immediate increase post-accession, while tobacco and alcohol showed more gradual, sustained increases in supply.
  • The Importance of Context: The authors emphasize that while their findings are suggestive, there is significant unexplained country-specific heterogeneity. This implies that the impact of WTO membership is not uniform and is likely influenced by a multitude of local factors.
  • Trade Agreements as Health Determinants: The paper argues that trade agreements should be viewed as significant social determinants of health, underscoring the need for a "Health in All Policies" approach when negotiating and implementing such agreements.

Key Results

  • Fruits and Vegetables: WTO member countries experienced an average immediate increase in the domestic supply of fruits and vegetables of 55 g per person per day compared to non-member countries.
  • Tobacco: The supply of tobacco showed a gradual increasing trend, with a geometric mean increase of 6.2% per year following WTO accession.
  • Alcohol: Similarly, alcohol supply increased gradually, with a geometric mean increase of 3.8% per year after joining the WTO.
  • Other Commodities: No significant changes were detected in the availability of red meats, animal fats, seafood, nuts/seeds/legumes, starches, or edible oils. Results for sugars were inconsistent across different model variations.
  • FCTC Impact: In the tobacco model, ratification of the FCTC was associated with an 18.5% lower geometric mean supply of tobacco.

Limitations and Future Work

The authors acknowledge several limitations:

  • Comparability of Countries: Despite propensity score weighting, inherent differences between countries that joined the WTO and those that did not may still influence trends.
  • Supply as a Proxy for Consumption: The data measures commodity supply, which is an indirect measure of actual consumption and health impact.
  • Data Granularity: The available data did not allow for distinguishing specific types of foods (e.g., high-fat, high-salt, high-sugar) or accounting for illicit trade in tobacco and alcohol.
  • Country-Specific Heterogeneity: The significant unexplained variation across countries suggests that future research should focus on smaller groups of countries or individual country case studies to understand these differences.
  • Income Group Stratification: Stratifying by income group yielded inconclusive results, possibly due to small sample sizes within each group.

Future research directions include:

  • In-depth analyses of smaller country groups and individual countries.
  • Examining the influence of economic development, geography, and climate.
  • Investigating specific provisions within WTO accession agreements and their differential impacts.
  • Analyzing other unhealthy food products (high in fat, salt, and sugar).

Summary updated

Abstract

The results suggest that WTO membership may lead to increases in both harmful and protective factors for noncommunicable disease, but further exploration of country-specific variation is warranted.
BibTeX
@article{cowling2018world,
  author = {Krycia Cowling and Elizabeth A. Stuart and Roni Neff and Daniel Barstow Magraw and Jon S. Vernick and Keshia M. Pollack Porter},
  title = {{World Trade Organization membership and changes in noncommunicable disease risk factors: a comparative interrupted time-series analysis, 1980–2013}},
  journal = {Bulletin},
  year = {2018},
  doi = {10.2471/blt.18.218057},
}
Figure