NAFTA and Mortality

There's a recent economics working paper called "Trading Goods For Lives: NAFTA’s Mortality Impacts And Implications" that is getting some attention. The NY Times did a piece on it under the headline "‘A Lot of Life Years Lost’: How NAFTA Shortened American Life Spans."

That's quite a provocative title for an academic paper! Let's look at the abstract to see what the paper is actually about:

We estimate the mortality effects of local labor market exposure to NAFTA and compare them to the mortality effects of other U.S. local labor market contractions. Areas more exposed to NAFTA experienced sustained increases in mortality over the subsequent 15 years. Mortality increases occurred for all broad demographic groups, but were especially pronounced among working-age men. Extending the analysis to other economic contractions, we show that the health consequences depend critically on which sectors bear the losses: declines in manufacturing employment increase local area mortality, while declines in non-manufacturing employment reduce it.

I have a couple thoughts in reaction.

First, the title refers to "NAFTA’s Mortality Impacts And Implications" and the abstract talks about "exposure to NAFTA," but as the paper itself clarifies, the focus here is narrower: "exposure to increased Mexican import competition from NAFTA." It's worth noting, though, that NAFTA did more than just allow increased Mexican imports. For example, just focusing on the U.S. side, it led to increased exports of various U.S. goods and services; it lowered prices for U.S. consumers; it helped make some U.S. industries more competitive with the rest of the world; and it provided stronger protections for intellectual property. What were the economic impacts of each of these, and the consequential impact on the mortality of specific regions? I don't know! But that seems worth considering as part of the bigger picture of "NAFTA's impact" on the economy and on mortality, otherwise the assessment leaves a lot out.

The NYT piece brings this point up too, noting that one of the paper's authors "cautioned that NAFTA might have had other benefits that his paper didn’t capture, including in innovation or in helping North America as a whole remain competitive against other parts of the world." But without these points being made in the paper itself, the paper is likely to be mischaracterized in the public debate.

And while we are on the subject of the broader impact of NAFTA, if there is a concern with increased imports from Mexico, shouldn't we think about the impact on people in Mexico too? If there were manufacturing losses in the U.S., weren't there manufacturing gains in Mexico? What was the economic and mortality impact there? Again, if the paper is presented as being about "NAFTA" rather than just "the impact of Mexican imports into the U.S.," these other considerations seem relevant.

Second, the paper only looks at the impact of Mexican imports on regions of the U.S., but of course there is another NAFTA party whose industries suddenly had to compete with Mexican imports. How did Canada fare in terms of mortality rates in its affected manufacturing regions? If Canada didn't see the same mortality impact, it may be that its social policies did a better job helping manufacturing workers adjust to import competition. That could give us a better sense of the bigger picture of causation here, and could tell us that it wasn't so much the imports to blame as it was the government's social policies to deal with the (widely anticipated) import surge. On this point, the NYT piece notes that one of the paper's authors "pointed to other countries, like Denmark, that had set up systems to help support and retrain workers who had lost their jobs." That idea seems like something that could be tested here by looking at the situation in Canada.

Finally, let's go a little deeper on the way the title frames the issue: "Trading Goods For Lives: NAFTA’s Mortality Impacts And Implications." Imagine a similar study where someone looks at the reduction of U.S. defense spending after the Cold War. They find that the reduction in purchases of weapons has the biggest economic impact on the specific geographic regions where the weapons are manufactured, and that there is a similar mortality effect on these regions as the NAFTA paper's authors found with increased imports from Mexico. (I'm not saying this effect definitely exists, but it's easy to imagine that it could). Then someone takes this study and writes a paper entitled something like: "Trading the Cold War For Lives: Peace's Mortality Impacts and Implications." Yes, statistically speaking, there could be something real here, but it seems like it misses the bigger picture. I would say there is something similar going on with this NAFTA paper. My sense is the NAFTA paper's authors might say the paper does not weigh in on the bigger picture, but the title makes it sound like it does, and regardless of how well done the statistical analysis is, many people are going to be misled by the title.