Multinationals and National Income Accounting

One of the joys of blogging is that one occasionally gets interesting emails from AB readers such as this one:

When a US based manufacturer (Caterpillar jumps to mind) makes products in Brazil and sells them in Europe, would that count as an export? What about a Toyota Camry manufactured in Ohio? Is that an import?

The first example is similar to something that the folks at Street Authority had in mind:

GNP: Gross National Product measures the total amount of goods and services that a country’s citizens produce regardless of where they produce them. As a result, GNP includes such items as corporate profits that multinational firms earn in overseas markets. For example, if an American firm operates a plant in Brazil, then the profits that the firm earns would contribute to U.S. GNP.
GDP: By contrast, GDP measures the total amount of goods and services that are produced within a country’s geographic borders. Therefore, for GDP purposes, an American company with a plant in Brazil will actually contribute to Brazilian GDP.

The value-added produced when the Brazilian subsidiary of Caterpillar can be thought of being composed of Brazilian wages and the profits for the U.S. shareholders of Caterpillar. Both are counted as Brazilian GDP with this transaction being seen as a Brazilian export to Europe. The profits generated, however, represent U.S. net income from abroad and would be part of our GNP and not Brazil’s GNP.

In the second example, let’s imagine that Toyota’s U.S. subsidiary sells the Camry to some U.S. retail distributor for $20,000 with the cost of components being $15,000 and value-added created in the Ohio plant being recorded as $5000. All of that value-added is counted as U.S. GDP with this being a U.S. produced car, which is purchased by a U.S. consumer. Of course, we could ask where those components were produced. Of the $5000 in value-added, let’s assume that $4000 is wages with $1000 being the profits for Toyota’s shareholders. In our example, Japanese net foreign income from abroad is $1000, which is deducted from U.S. GNP.

The following graph shows the ratio of U.S. GDP to U.S. GNP over the past 35 years. Note that GNP has slightly exceeded GDP as the U.S. has consistently had positive net foreign income from abroad. The rise in the GDP/GNP ratio in the early 1980’s likely reflects the fact that our current account deficits virtually wiped out the net creditor position of the U.S. with the U.S. now in a position of having negative net foreign assets. So why hasn’t the GDP/GNP ratio surprised unity – since one would expect the U.S. to have negative net foreign income from abroad?

I guess the accounting answer is that the profits from U.S. investments abroad are higher than the profits from foreign investments in the U.S. Of course, the allocation of accounting profits for a multinational enterprise depends on its intercompany pricing policies. While the IRS strives to enforce arm’s length pricing, one can identify situations where foreign parent corporations source very little profits in their U.S. subsidiaries, while U.S. parent corporations can source substantial portions of their profits in foreign low-tax jurisdictions.

I can think of two excellent examples of where transfer pricing manipulation distorts the calculation of GDP. One comes from Russia and the types of transfer pricing issues that are part of the Yukos Oil scandal as discussed by the World Bank with the Introduction noting:

Part II addresses a puzzle in Russia’s national accounts. According to official data, the oil and gas sector in Russia comprises less than 9 percent of GDP, while exports from this sector alone amount to 20 percent of GDP. At the same time, the production of services exceeds the production of goods by a wide margin, while the official share of non-market services is very small. Other data also raise questions – for example, the trade sector in the official accounts is huge and profitable, with about one third of GDP and half of all profits generated by trade. We argue that these puzzling observations can be explained by transfer pricing. Many large Russian companies use trading companies to market their output. Using transfer pricing, a firms’ production subsidiary sells output cheaply to the same firm’s trading subsidiary, which then sells it to customers at market prices. Hence, most of the value added accrues to the trading company. Tax can be avoided if the trading subsidiary is able to pay a lower effective tax rate than the production subsidiary would have without the “transfer” of value added. Since Russia’s national accounts are not adjusted for these schemes, transfer pricing has the effect of greatly exaggerating value-added in the service sectors, especially in trade, and underestimating it in industry, especially industries that make heavy use of transfer pricing, such as oil and gas. Correcting the trade margins, using international comparisons, results in oil and gas almost tripling in size, industry again becoming the largest sector, and market services losing some of their weight in GDP. The results are published in part II of this report, including a conversion table that compares the shares of various sectors in GDP before and after the recalculation.

But I’m only warming up for a longer post on the Irish economic miracle – so my second example of transfer pricing manipulation and GDP accounting must wait until later this week.