April 2018 – J. W. Mason

Obstfeld, Globalist

This Maurice Obstfeld op-ed in the FT is a perfect distillation of the orthodox position on trade. Obstfeld is the guardian of free-trade orthodoxy ex officio as head of research at the IMF; he’s also done the circuit of top US departments and is co-author of one of the most widely used undergraduate textbooks. [1]

The op-ed is, of course, against tariffs. This isn’t news. I was at a session on trade policy at the American Economic Association last year, where the chair introduced the panel by saying, “Obviously, if you are in this room then you are for free trade, as much as we can get,” which is a pretty fair description of the range of opinion among credentialed economists. But it’s still striking how many of the tenets of faith Obstfeld manages to hit in 250 words.

Start with the explanation for why trade deficits are not necessarily bad:

“For example, they can help countries finance productive long-term investments that ultimately raise national income and wealth.”

This is the classic argument for international integration. Poor countries, by accepting capital inflows (the ambiguity between capital as money and capital as means of production is essential to the argument) can finance more investment and thereby achieve faster income growth than they would be able to on the basis of domestic savings alone. Obstsfeld’s “for example” is misleading here — as I’ve pointed out before, the option of running trade deficits is the entire benefit of free flows of portfolio investment in the orthodox theory.

Analytically, the op-ed’s key passage is:

A country’s overall trade balance is a macroeconomic phenomenon that mirrors whether it spends less than its income or more. In contrast, the structure of bilateral trade reflects the international division of labor – based on each country’s comparative advantage.

Here we have three key planks of orthodox trade economics. First, the airtight seal between “macro” and “micro” analysis, which protects us from discovering that these are two incompatible approaches based on mutually contradictory assumptions. Second, the anti-Keynesian macro component, with income fixed and savings as a constraint. Third, the bland invocation of the “international division of labor,” as if this were an anodyne technical fact and not a hierarchical, unequal relationship between the rich and poor worlds.

The macroeconomic part of Obstfeld’s argument is that trade restrictions “would not alter the fact that the US spends more than it earns — the source of the overall US deficit.” It is, of course, true as a matter of accounting that the current account deficit is equal to the government deficit plus the difference between private investment and private saving. Writing “the source” implies that this is a causal relationship and not just an accounting one — that how much the US “earns” is independent of the trade position. But there’s a problem — additional US exports constitute additional income for US businesses and households. An increase in US exports (or fall in imports) would, all else equal, increase savings by an exactly equal amount. So it’s not obvious how savings can be a constraint on the trade balance.

The argument that trade policy cannot change the overall deficit because national saving is fixed, is simply a transposition of the “Treasury view” of the 1930s that public investment could not increase output or employment since it would draw on the fixed supply of national saving and would crowd out an equal amount of private investment. It’s wrong for the same reason: Exports, like investment, create their own saving. It’s straightforward to show how interventions like tariffs or devaluations can generate some mix of higher output and a move toward trade surplus, while all the accounting identities are satisfied. This was a standard feature of older textbooks, and of many more recent ones in the form of the IS-LM-BP model, even if it’s not there in the more recent Obstfeld-Krugman books.

Obstfeld himself seems to have some misgivings about this argument, since he adds the caveat “for a country at full employment, like the US.” He also warns that trade restrictions “could derail the world economy’s current expansion,” which is obviously inconsistent with the idea that saving and investment are determined prior to trade balances.

It’s also striking that while Obstfeld acknowledges that “trade balances can of course be excessive” (the “of course” here functioning as a dismissal) there is no hint anywhere else in the op-ed about what the dangers of excessive deficits might be or how they could arise.

On the micro side, Obstfeld simply repeats variations on the same formula several times: trade restrictions “can badly distort the international division of labor.”

This sounds fine: division of labor is good, distorting is bad. (Distort is one of the many keywords that allows economic theory to appear to make contact with observable reality by confusing a technical meaning with an everyday one.) But what this formula actually means is: The countries that are rich, should remain rich. The countries that are poor, should remain poor. The countries that specialize in higher education and software and pharmaceuticals should retain their monopolies, the countries that specialize in plantation agriculture and sweatshop clothing should keep on doing that. “Comparative advantage” means that the hewers of wood and drawers of water are destined to remain such. Everybody should stay in their lane.

The “international division of labor” is a gesture at models that start from the premise that countries’ productive potential is fixed, given by nature or god. It is directly opposed to the idea of economic development, which starts from the premise that productive potentials are contingent and path-dependent, and that the whole goal of policy is to change — “distort”, if you will — the international division of labor.

These phrases might not have leaped out at me so much if I hadn’t just been reading Quinn Slobodian’s Globalists. (It’s a great book — look for my review in the Boston Review of Books sometime soon.) As Slobodian lucidly recounts, the real content of Obstfeld’s pieties was expressed more clearly by Mont Pelerin luminaries like Wilhelm Ropke, who

believed that an economically equal world might simply be impossible, and that developing countries might have to remain underdeveloped as a way of preventing possible ‘over-industrialization and underagriculturalization of the world.’ … the conditions for industrialization in the Third World did not exist. .. ‘The rich countries of today are rich because, along with the necessary prerequisites of modern technology, they have a particular form of economic organization that responds to their spirit.’ … Ropke believed that the ‘lack of punctuality, reliability, inclination to save and create’ … meant that industrialization schemes in the Global South were ‘doomed to fail.’

The position these early neoliberals were arguing against was the “global New Deal” which aimed not to reinforce the global division of labor, but to erase it through a convergence between the poor and rich worlds — in the memorable words of Senator Kenneth Wherry, to “lift Shanghai up and up, ever up, until it is just like Kansas City.” It’s worth emphasizing how diametrically opposed Obstfeld’s 2018 vision of trade is to Wherry’s 1940 one. Comparative advantage and the international division of labor are, for Obstfeld, fixed and god-given. You’d think the fact that Shanghai has in fact risen up well above Kansas City — and more broadly, that China, the greatest economic growth story of our times, has violated every one of his precepts — would give him pause. But it doesn’t seem to.

The neoliberals of the 1940s and 50s took exactly Obstfeld’s line. In Slobodian’s summary, their “critique of mainstream development theories began with the conviction that the industrialization of formerly agricultural areas … distorted the international division of labor and led nations to specialize in branches of production for which their natural endowments were unsuited.” Since so many people in the newly independent South were unhappy with their current position in the division of labor, this led naturally to calls for restrictions on political rights in the former colonies, and of non-whites in South Africa and elsewhere.

Obstfeld would, I’m sure, be appalled at the frank racism of Ropke. But Ropke at least had an explanation for why the benefits of industry and technology should be concentrated in a small part of the globe — the genetic-slash-cultural superiority of white Europeans. What’s Obstfeld’s explanation for why the “undistorted” international division of labor happens to so favor Europe and North America? Is it the climate?

[1] I’ve assigned Obstfeld’s textbook. It’s pretty good, as mainstream texts go.

“The financialization of the nonfinancial corporation”

One common narrative attached to the murky term financialization is that nonfinancial corporations have, in effect, turned themselves into banks or hedge funds — they have replaced investment in means of production with ownership of financial assets. Financial profits, in this story, have increasingly substituted for profits from making and selling stuff. I’m not sure where this idea originates — the epidemiology points toward my own homeland of UMass-Amherst — but it’s become almost accepted wisdom in left economics.

I’ve been skeptical of this story for a while, partly because it conflicts with my own vision of financialization as something done to nonfinancial corporations rather than by them — a point I’ll return to at the end of the post — and partly because I’ve never seen good evidence for it. On the cashflow side, it’s true there is a rise in interest income from the 1960s through the 1980s. But, as discussed in the previous post, this is outweighed by a rise in interest payments; it reflects a general rise in interest rates rather than a reorientation of corporate activity; and has subsequently been reversed. On the balance sheet side, there is indeed a secular rise in “financial” assets, but this is all in what the financial accounts call “unidentified” assets, which I’ve always suspected is mostly goodwill and equity in subsidiaries rather than anything we would normally think of as financial assets.

Now courtesy of Nathan Tankus, here is an excellent paper by Joel Rabinovitch that makes this case much more thoroughly than I’d been able to.

The paper starts by distinguishing two broad stories of financialization: shareholder value orientation and acquisition of financial assets. In the first story, financialization means that corporations are increasingly oriented toward the wishes or interests of shareholders and other financial claimants. The second story is the one we are interested in here. Rabinovitch’s paper doesn’t directly engage with the shareholder-value story, but it implicitly strengthens it by criticizing the financial-assets one.

The targets of the paper include some of my smartest friends. So I’ll be interested to see what they say in response to it.

The critical questions are: Have nonfinancial corporations’ holdings of financial assets really increased, relative to total assets? And, has their financial income risen relative to total income?

The answers in turn depend on two subsidiary issues. On the first question, we need to decide what is represented by the “other unidentified assets” category in the Financial Accounts, which is responsible for essentially all of the apparent rise in financial assets. And on the income side, we need to consistently compare the full set of financial flows to their nonfinancial equivalents. Rabinovitch argues, convincingly in my view, that looking at financial income in isolation is not give a meaningful picture.

On the face of it, the asset and income pictures look quite different. In the official accounts, financial assets of nonfinancial corporations have increased from 40% of nonfinancial assets to 120% between 1946 and 2015. Financial income, on the other hand, is only 2.5% of total income and shows no long-term increase. This should already make us skeptical that the increase in “financial” assets represents income-generating assets in the usual sense.

Rabinovitch then explores this is detail by combining the financial accounts with the IRS statistics of income (SOI) and the Compustat database. Each of these has strengths and weaknesses — Compustat provides firm-level data, but is limited to large, publicly-traded corporations and consolidates domestic and overseas operations; SOI gives detailed breakdowns of income sources for all forms of legal organization broken down by size, but it doesn’t include any balance-sheet variables, so it can’t be used to answer the asset questions.

iI the financial accounts, the majority of the increase in identified financial assets is FDI stock. As Rabinovitch notes, “it’s dubious to directly consider FDI as a financial asset if we take into account that it implies lasting interest with the intention to exercise control over the enterprise.” The largest part of the overall increase in financial assets, however, is in the residual “other unidentified assets” line of the financial accounts. The fact that there is no increase in income associated with these assets is already a reason to doubt that they are financial assets in the usual sense. Compustat data, while not strictly comparable, suggests that the majority of this is intangibles. The most important intangible is goodwill, which is simply the accounting term of the excess of an acquisition price over the book value of the acquired company. Importantly, goodwill is not depreciated but only written off through impairment. Another large portion is equity in unconsolidated subsidiaries; this accounts for a disproportionate share of the increase thanks to a change in accounting rules that required corporations to begin accounting for it explicitly. Other important intangibles include patents, copyrights, licenses, etc. These are not financial assets; rather they are assets or pseudo-assets acquired, like real investment, in order to carry out a company’s productive activities on an extended scale.

These are all aggregate numbers; perhaps the financialization story holds up better for the biggest firms? Rabinovich discusses this too. Both Compustat and SOI allow us to separate firms by size. As it turns out, the largest firms do have a greater proportion of financial income than the smaller ones. But even for the largest 0.05% of corporations, financial income is still only 3.5% or total income, and net financial income is still negative. As he reasonably concludes, “even for the biggest nonfinancial corporations, financialization must not be understood as mimicking financial corporations.”

What do we make of all this?

First, the view of financialization as nonfinancial businesses acquiring financial assets for income in placer of real investment, is widely held on the left. After my Jacobin interview came out, for example, several people promptly informed me that I was missing this important fact. So if the evidence does not in fact support it, that is worth knowing. Or at least, future statements of the hypothesis will be stronger if they respond to the points made here.

Second, the fact that “financial” assets in fact mostly consist of goodwill, interest in unconsolidated subsidiaries, and foreign investment is interesting in its own right, not just as negative criticism of the financialization story. It a sign of the importance of ownership claims as a means of control over production— both as the substantive content of balance sheet positions and as a core part of corporate activity.

Third, the larger importance of the story is to the question of whether nonfinancial corporations and their managers should be seen mainly as participants in, or victims of, financialization. Conversely, is finance itself a distinct social actor? In a world in which the largest nonfinancial corporations have effectively turned themselves into hedge funds, it would not make much sense to talk about a conflict between productive capital and financial capital, or to imagine them as two distinct sets of people. But in a world like the one described here, or in my previous post, where the main nexus between nonfinancial corporations and finance is payments from the former to the latter, it may indeed make sense to think of them as distinct actors, of conflicts between them, and of intervening politically on one side or the other.

Finally, to me, this paper is a model of how to do empirical work in economics. Through some historical process I’d like to understand better, economists have become obsessed with regression, to the point that in academic economics it’s become synonymous with empirics. Regression analysis starts from the idea that the data we observe is a random draw from some underlying data generating process in which a variable of interest is a function of one or more other variables. The goal of the regression is to recover the parameters of that function by observing independent or exogenous variation in the variables. But for most macroeconomic questions, we are dealing with historical processes where our goal is to understand what actually happened, and where the hypothesis of some underlying data-generating process from which historical data is drawn randomly, is neither realistic nor useful. On the other hand, the economy is not a black box; we always have some idea of the mechanism linking macroeconomic variables. So we don’t need to evaluate our hypotheses by asking how probable the it would be to draw the distribution we observe from some hypothetical random process; we can, and generally should, ask instead whether the historical pattern is consistent with the mechanism. Furthermore, regression analysis is generally focused on the qualitative question of whether variation in one variable can be said to cause variation in a second one; but in historical macroeconomics we are generally interested in how much of the variation in some outcome is due to various causes. So a regression approach, it seems to me, is basically unsuited to the questions addressed here. This paper, it seems to me, is a model of what one should do instead.