Multinational Sales v. GDP: Drum v. Drezner

I almost never disagree with Kevin Drum but Daniel Drezner has a point:

One of the more invidious comparisons analysts like to make is to compare the size of something with a country’s gross domestic product. An old warhorse of political economy/anti-corporate types, for example, is to say that the sales of multinational corporations exceeds many countries GDP. This is true but irrelevant — GDP measures the value-added that an economy generates per year, so the proper and correct comparison is between a firm’s profits and GDP. When using that metric, corporations suddenly don’t look so big.

Kevin replies:

Hold on a second. This isn’t true, is it? Leaving aside trade deficits, GDP is basically consumption plus investment. If I buy a Ford Taurus for $20,000, that adds $20,000 to GDP even though Ford makes a net profit per car of about $3 these days. (In a good quarter, that is.) I don’t know if sales revenue is precisely comparable to GDP, but it’s pretty close. Right?

But what if Ford imported most of the components from Mexico? For a dramatic example of what Daniel is talking about, considering the sales of Mitsui, Mitsubishi, Itochu, Sumitomo, and Marubeni – which as the five largest Sogo Shoshas (Japanese trading companies). Their combined sales sum to 12 percent of Japan’s GDP but the gross profits of these companies are often less than 5 percent of their sales as reported by “turnover”. In other words, these companies buy and resell products internationally with very little value-added. Then again – US GAAP under EITF 99-19 might require that these companies report as revenues what is reported as gross profits under Japanese GAAP. So what do we mean by sales in the first place?

Dan, however, is not quite right with profits capturing value-added if he means operating profits. One would have to add back the value of domestic labor. Of course, gross profits only captures value-added when cost of goods sold is exclusively cost of purchasing intermediate goods. But GAAP does not exactly do this the way economists might like.