A Statistical Profit Puzzle

I’m going to warn you right up front. This post is being written on the Saturday evening of Labor Day weekend. I’ve had a tough day (don’t ask).  I’m writing about a topic that no one is going to care about, but is in fact very important.

Let’s start: The Bureau of Economic Analysis publishes quarterly statistics on “corporate profits” and “corporate profits of domestic industries.”  Corporate profits of domestic industries are used to calculate Gross Domestic Income, which is should be equal  to Gross Domestic Product.   (To see a little discussion of GDI, see this James Hamilton article and here).

Here’s the question: Suppose a U.S. company discovers that the cost of importing a good or service is less than the cost of buying it from a domestic producer (or making the good or service itself).  The executives of the company decide to boost profits by offshoring production of that good or service. 

Should that increase in profits be treated as an increase in “corporate profits of domestic industries”? And by extension, should that count as an increase in GDI? And by extension, GDP?

Discussion beneath the fold.

Suppose you believe the answer is yes, that an increase in profits because of offshoring should be counted as an increase in domestic profits. The problem is this: You can imagine offshoring manufacturing, and keeping R&D, marketing, finance, and strategic planning in this country.  Now suppose it becomes cheaper to do R&D in China, so you move R&D overseas and increase profits again. Now it becomes cheaper to do marketing and finance in Paris and London, so you move those operations overseas as well, at each step increasing corporate profits.  Finally you are left with one executive and one janitor in the New York City office. 

Should that hollowing out all count as domestic profits?

This is actually a big deal.


  1. Yes, it should count as domestic if the people receiving the profits are domestic. However, in today’s global markets, it’s impossible to say where those profits actually go with a dividend payout. All this goes to show that GDP was a dumb concept to begin with and it has lost what little meaning it had because of globalization. Good, as such a mass aggregation of all “domestic product” always obscured far more than it illuminated to begin with, better to ditch the GDP concept.

  2. I don’t know the definitions and due to other plans leave it with drive-by commenting, but there is a distinction between GNP and GDP. Maybe what you are talking about is in the GNP category, and GDP is only strictly domestic activities? (Aside from all other shenanigans with hedonic adjustments, transfer pricing, etc.)

  3. Mike Mandel says:

    The BEA switched from GNP to GDP as its feature measure in 1991, in part because GNP runs into the same sort of problems that I describe. I’m pointing out that GDP actually has similar problems to GNP.

  4. If no one else here is going to do it, I will. It is clear that GDP, insofar as it does not track whether dividend payments and retained earnings stay in the U.S. or go overseas, is a flawed measure of economic output. Is it possible to just estimate wage payments, since the vast majority of Americans derive all of their income from wages, and use that as the proxy measure for output? I guess that entrepreneurs are important as well, but surely it would be simple enough to just distinguish between s-corps and c-corps, and just assume that c-corp profits shouldn’t be counted because they are not truly entrepreneurial.

    • Fixating on wages doesn’t capture the gains in profitability due to automation. I think off-shoring (of non-commodities, especially) is interesting because it speaks to competitiveness: the ability to generate new products and perhaps new jobs (or at least better-paying ones). But increased profits due to increased automation may show increased competitiveness, while sufficiently increased off-shoring may indicate transformation from a productive corporation to simply a brand.

  5. It is a big deal. If you’re measuring profit, well it should show up. But if you’re concerned about measuring actual domestic production, that’s a different bucket, which seems to be the insight here. I don’t know how you measure that bucket: it seems to be kind of a fuzzy concept. I don’t believe that means you should just give up, but shoehorning new measurements into old figures seems a bit dangerous: this could use some thought, if we’re serious about it (and we probably should be).

    • What is so fuzzy about it? Domestic production is what is produced in domestic entities.

      The only fuzziness is domestic entities not operating inside the US all the time (e.g. a one-person shop with the proprietor traveling around and doing their work here and there), or where work that has happened outside the US (e.g. in a software MNC) is misattributed to US sources. To the extent that there are regulations that inputs and domestic activities must be separately declared that would be fraudulent if deliberate, whether it can be practically enforced or not. One can imagine US companies claiming R&D tax credits for R&D work that was actually done offshore (not that I would know any examples).

      • It’s fuzzy because good attribution is difficult: that’s part of the reason why there are so many accountants in the world. A US import firm that creates 20 importing jobs should surely contribute something to GDP, right? How about a global marketing firm with a fluid presence in the US: the income contribution to the US is easy to measure, but the revenue is not. You could imagine a myriad of such examples. My current project is being performed in several sites in at least 3 countries, and that’s only in the current stage of its lifecycle. Accurate contribution to US GDP would be hellish to compute.

      • In your case and probably many other cases, attribution is not so much difficult as laborious, and surely in a hellish way. Back in the day I had to fill out detailed work report sheets, breaking the hours down into different projects etc. It was not pleasant but manageable. Many service organizations that work by billing hours (internally or externally) are doing that too. Attributing actual creation of value is essentially impossible, attribution based on work hours is probably the next best thing. After that, estimates based on headcounts. The latter should be interesting in organizations with a large “international presence” (offshorers).

    • On the issue of transfer prices, some US software companies have been playing the game that their US-made IP titles (more general name for e.g. software products but covering more nonphysical “IP” stuff) are distributed through foreign entities that are charging back prices for their “service” and retain profits in ways that the IRS has tried to get classified as not arms-length and illegal a few years ago (without much success as far as I know, under the “business friendly” administration). But in such cases that stuff shouldn’t show up as GDP, unless it can be booked differently for tax purposes vs for GDP statistics. That would be highly questionable, but I wouldn’t be surprised at all with all the BS we got lately.

      • Those things would be quite interesting to know, but expensive and irritating to gather… and not well fitted in the umbrella of GDP, which I thought was the insight behind Mike’s post. But yeah, billable hours, headcount, or headcount-dollars are all technically doable if we want to make businesses measure them.

      • Whether “we” would like the outcome from such more detailed indicators is another question …

      • cm,

        Whether “we” would like the outcome from such more detailed indicators is another question …”

        I guess the idea is it’s better to know….

  6. Elise Rapoza says:

    Hi! I’ve been following your work pretty closely since I discovered your articles on the innovation shortfall and the GDP illusion. I’m very glad to see someone identifying misleading statistics in the world of economics. I believe that the globalized economy and enviromental degradation demand a new statistical framework and system of data collection. Keep up the hood worl

  7. Hi All,

    I came across this article this morning which puts the case against increasing protectionism. Thought it might be food for thought.



  1. […] Michael Mandel muses on the oddity of why hollowing out contributes to domestic corporate profits: “Suppose a U.S. company discovers that the cost of importing a good or service is less than the cost of buying it from a domestic producer (or making the good or service itself).  The executives of the company decide to boost profits by offshoring production of that good or service.  Should that increase in profits be treated as an increase in “corporate profits of domestic industries”? “ […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s


%d bloggers like this: