Promoting Exports, Preventing Poverty: Toward a Causal Evidence Base

Abstract

Near the climax of the Scientific Revolution, two influential theories in chemistry were put forward within five years of each other. In 1662, Robert Boyle proposed that the volume and pressure of gases are inversely proportional: decreasing the volume of an amount of gas pressurizes it predictably. Next, in 1667, Johann Joaquim Becher posited that all flammable substances contain the element terra pinguis , which he claimed is released during combustion. Georg Stahl later renamed the combustible element phlogiston . Both Boyle's and Becher's theories went on to profoundly affect the science and practice of chemistry for the next century. But only one of the ideas has survived: Boyle's Law remains a foundational principle of modern chemistry. On the other hand, Becher's terra pinguis/ phlogiston was definitively disproved by Antoine-Laurent Lavoisier's closed experiments, reported in 1783, in which he demonstrated that burning requires oxygen and that solids sometimes gained mass when combusting. Yet for more than a century, chemists believed in phlogiston and used the idea to guide their work. As social scientists searching for remedies to global poverty, in some ways we find ourselves in a similar position as chemists in the seventeenth and eighteenth centuries. David Ricardo's Law of Comparative Advantage, one of the most pivotal ideas in social science—the single economic proposition that Nobel laureate Paul Samuelson (⇓) asserted was both nontrivial and nonobvious—suggests that specialization and exchange might provide a powerful antidote to poverty. The mathematical proof is compelling. Most social scientists, indeed most educated people, believe it. But in combating global poverty, is Ricardo's Law like Boyle's Law? Or is it like phlogiston? The only way to know is through compelling empirical support. However, the causal evidence base for trade as an antipoverty strategy is scanty, at least so far. Observational evidence, capable of suggesting …