Saturday, September 27, 2014

Explaining Sluggish Growth in Latin America

The Brookings Institute has a short and gloomy analysis of economic performance in Latin America. They base the gloom on the fact that Latin American economic growth is sluggish right now even though the "external environment" is very favorable, whereas previous dips corresponding with poor external conditions and the economies recovered when those conditions improved.

Here's the chart they base the argument on:

This is really a classic case of confusing correlation and causation. "External environment" is defined very narrowly and indeed arbirtrarily as a few major crises. Those crises fit the argument, but in the absence of testing other possibilities, we can only say there is correlation, not causation.

Most prominently, we know Latin America depends heavily on commodities. We also know China has been buying commodities. Yet in the past few years China's growth is slowing. It is therefore reasonable to hypothesize that the current sluggishness stems from that. It's correlation until we do more work to show causation, but it's at least more precise than their analysis.

Along similar lines, since Latin America exports commodities we need also look at commodity prices. As it turns out, corn and soy are at four year lows. This would take much more work, but we could hypothesize that dropping commodity prices (or dig down and specify what commodities in what countries) are causing weak economic growth.

Their analysis could be right. Maybe. I feel there are plenty of ignored independent variables that likely do a better job of explaining the dependent variable of growth rates. You don't know until you test them, and if you don't test them at all, then you don't have much of an analytical foundation.


Otto 8:57 AM  

"Their analysis could be right."

That's economics, right there.

Greg Weeks 9:51 AM  

No--economics is "MY ANALYSIS IS 100% RIGHT."

  © Blogger templates The Professional Template by 2008

Back to TOP