In recent posts, Pierre Lemieux and Scott Sumner refuted a common but mistaken idea: that because gross domestic product can be calculated by taking national expenditure data (i.e., the “C + I + G” part of the equation) and adding the value of exports while subtracting the value of imports (the “X – M” in said equation), imports reduce GDP.

This mistake should be obvious to people who have weighed themselves at the gym and then subtracted a few pounds to allow for the weight of their shoes and clothes: wearing heavier clothes doesn’t change their body weight. Similarly, subtracting imports that were included in C, I, or G expenditures does not reduce the “weight” of GDP. Still, this mistake is so common that such folks as Trump administration trade adviser Peter Navarro (a Harvard economics Ph.D., no less), political figure Pat Buchanan, countless economics reporters, and others who should know better routinely make it.

One of the reasons this error persists is that it echoes another mistaken but intuitively compelling idea: the imported goods could have been made domestically, and thus would have added to domestic employment. To show that this intuition is false, economists recite the theory of comparative advantage. Unfortunately, trying to explain what has been called the most counterintuitive idea in social science is not a winning debate strategy.

Let’s Talk about Chocolate

So here’s a different strategy: tell the story of the cocoa bean, one of many U.S. imports that increases American GDP and employment.

Cacao trees, which produce the bean, grow only in hot and humid conditions. Most of the world’s cacao comes from right along the equator; practically none comes from the United States. (American Samoa produces something like 0.0002 percent of the world’s crop). Yet the United States is one of the world’s largest chocolate producers, using cocoa beans imported from Ivory Coast, Ghana, Indonesia, Trinidad, and elsewhere.