Economist Caballero on Macroeconomics: “Fantasyland”

The surprisingly reader-friendly Journal of Economic Perspectives has a symposium on Macroeconomics after the Financial Crisis. I’m working my way through the papers now, starting with a fascinating piece by Caballero on the “core” and “periphery” of contemporary macroeconomics. For Caballero, the periphery of modern macro includes useful analysis of bubbles, liquidity traps, financial disintermediation and the like, and restricts itself to specific problems under semi-realistic assumptions, and thus has provided much of the key explanatory power of economics in the current crisis. On the other hand, Caballero argues that the core includes both New Keynesian and Real Business Cycle theories that emphasize hyperrational agents solving incredibly complicated Dynamic Stochastic General Equilibrium calculations with increasingly complex “frictions” built in to add realism. These models attempt to solve for all variables at once, so to speak. Caballero (2010: 90) argues that the core may suffer from a “pretense of knowledge” problem brought about by its approach of adding more and more complicated deviations into what begins as a simple, straightforward model:

However, I think this incremental strategy may well have overshot its peak and may lead us to a minimum rather than a maximum in terms of capturing realistic macroeconomic phenomena. We are digging ourselves, one step at a time, deeper and deeper into a Fantasyland, with economic agents who can solve richer and richer stochastic general equilibrium problems containing all sorts of frictions. Because the “progress” is gradual, we do not seem to notice as we accept what are increasingly absurd behavioral conventions and stretch the intelligence and information of underlying economic agents to levels that render them unrecognizable.

The beauty of the simplest barebones real business cycle model is, in fact, in its simplicity. It is a coherent description of equilibrium in a frictionless world, where it is reasonable to expect that humans can deal with its simplicity. I would rather stop there (perhaps with space for adding one nominal rigidity) and simply acknowledge that it is a benchmark, not a shell or a steppingstone for everything we study in macroeconomics, which is unfortunately the way the core treats it today.

When a high-profile MIT economist in an AEA-sponsored journal can accuse the core of his subfield of descending “deeper and deeper into a Fantasyland”, I wonder what economic sociology is to do. I mean, we’re not going to come up with better, more authoritative critiques of the unreality of the field than that! So perhaps we should be looking for other tasks – for example, the work of scholars like MacKenzie and Callon who focus on the effects of economic theorizing (independent of its “truth”), or the work of historians of economics like Marion Fourcade who ask, how did economics get to be the way it is in various times and places? Just bashing economics for its unrealistic assumptions seems so… unnecessary? at this point from our discipline. Or at least, if we do so, we must be careful to address the internal critiques along the same lines – from authors like Caballero, as well as the whole behavioral economics movement. What is economic sociology offering that goes beyond these critiques? Etc.

As a fun aside, and research note for myself, Caballero also notes that modern macro models assume that agents possess complicated understandings of the current workings of the economy, and that this knowledge justifies the generalization from everyday microeconomic rationality, yet:

Agents could be fully rational with respect to their local environments and everyday activities, but they are most probably nearly clueless with respect to the statistics about which current macroeconomic models expect them to have full information and rational information. (91)

From this argument we might then conclude that the compilation and distribution of macroeconomic statistical knowledge itself affects the possibilities for rational action. Not that I have any interest in such topics or anything…

Robert Shiller Just Made My Day

Robert Shiller is a prominent economist, associated with the Case-Shiller index for real-estate prices that foresaw and measured the extent of the housing bubble. He also wrote a semi-popular behavioral economics account of the crisis with Akerlof, Animal Spirits. Unrelated to both of those efforts (more or less), in a recent NYTimes piece, Shiller proposes something awesome: a market for “trills”. Alas, Shiller is likely not referencing Star Trek. Rather, Shiller is proposing that countries start financing their debts and deficits with the sales of a new kind of bond-ish-thing, a trill, which is short for a trillionth of GDP. Here’s his explanation:

Let me explain: Each trill would represent one-trillionth of the country’s G.D.P. And each would pay in perpetuity, and in domestic currency, a quarterly dividend equal to a trillionth of the nation’s quarterly nominal G.D.P.

If substantial markets could be established for them, trills would be a major new source of government funding. Trills would be issued with the full faith and credit of the respective governments. That means investors could trust that governments would pay out shares of G.D.P. as promised, or buy back the trills at market prices.

If trills were issued by Canada, for example, they would pay about 1.50 Canadian dollars in dividends this year, one trillionth of the annual cash flow. The value of the security is derived from the dividend, and might be priced very highly in the market — perhaps at around 150 Canadian dollars — given that country’s strong prospects for growth. Trills issued by the United States Treasury would pay about $14 in dividends this year and might fetch $1,400 a trill or more.

I love this idea on so many levels. The most important way in which I love this proposal is the way that it both relies on and enhances the black box-y-ness of G.D.P.. That is, this type of security only makes sense if G.D.P. is an objective fact (in Porter’s mechanical sense, of not subject to manipulation by the whims of any group). But, Shiller assures us, G.D.P. is an objective fact:

Historically, one impediment to such a move was the difficulty in accounting on a national scale: governments didn’t even try to measure G.D.P. until well into the 20th century.

Although G.D.P. numbers still aren’t perfect — they are subject to periodic revisions, for example — the basic problem has been largely solved. So why not issue shares in G.D.P. now?

We already have international equity markets that allow international investments in private firms within countries. But these do not represent the entire economy. Corporate stocks represent implicit claims on after-tax corporate profits, which typically amount to no more than 10 percent of G.D.P. Moreover, after-tax corporate profits are a much more slippery concept than G.D.P., affected as they are by many domestic policies, including taxes, government involvement in labor disputes and even government bailouts — as we now know very well.

So, what Shiller is saying (along with other things about the relationship between central banks, interest rates, inflation, and the growth of the economy as a whole) is that G.D.P. is a more factish-fact than corporate profits, which are more like artefacts of corporate accounting practices (Enron, anyone? See also Espeland and Hirsch 1990, MacKenzie 2009: ch. 6). But why should we believe that? What makes us think that G.D.P. is not just as much an artefact of its construction – by government actors, to be sure, rather than corporate ones, but since when have economists been unwilling to examine the motives, biases, and failings of government actors just as much (if not more than) corporate ones? Not to mention the fact that the data for G.D.P. comes itself from all kinds of other reporting procedures – the I.R.S. being a big one, for example – and is thus a sort of secondary fact. No single survey produces the G.D.P. (as the Census, more or less, produces the overall population of the U.S., say). Rather, G.D.P. is a kind of hodge-podge, thrown together based on what data is available and can easily be acquired. Admittedly, it’s a fairly stable hodge-podge, having been mostly stable in its definition for the past 60 years*, but it’s still a funny sort of measure, a routine solution to an impossible problem (how much material welfare was produced in an entire nation in a given period of time?) that’s only good enough for-all-practical-purposes. The question is, if we expand the practical purposes (from state planning and development aid to actually being part of the technical infrastructure of a market for government securities), will it still be good enough?

And despite Shiller’s assurances that the “basic problems” have been solved, a great many people disagree, arguing that G.D.P. gets hugely important things wrong like not valuing unpaid housework, the environment, or free information goods. The first two have always been important, and the last is seeming more so everyday. How are we to assure investors that a measure invented in the mid-20th century to make sense of the economic systems of the U.S. and U.K. during the Great Depression and World War II will make sense in Brazil in 2050 or China in 2100? What delicious hubris to think that we have solved such a massive problem in a way that will remain stably solved for the foreseeable future!

Anyway, thanks Robert Shiller, for giving me new food for thought.

* Although we did shift from emphasizing GNP to GDP, a consequential shift in an era of increase global trade, as GDP includes the value of goods and services produced in a country but owned by foreigners (say, a factory owned by an American company in Mexico). I have not yet found good sources that discuss how big a difference this shift has made in overall aggregates or in rankings of nations – does anyone know of any?

Archives Are Awesome (and a Brief Complaint about Economic History)

I received approximately the coolest piece of mail ever today: the collection guide for the archive of Simon Kuznets’ papers at Harvard. It’s like a preview of coming/very old attractions. For example, Box 1 contains “Corresp. w/J.M. Keynes 1936; Corresp. w/Prof. Don Patinkin re-Keynes [Reprints]”. I do not know what this means or what the correspondence concerns (though I can make some educated guesses). At some point – summer of 2010? – I am going to go find out.

On an almost unrelated note, this article in the NYTimes about FDR’s economic policy rings a little false to me. Here’s a section:

Roosevelt’s New Deal is often portrayed as an embrace of Keynesian economics, which advocates increased government spending to combat economic downturns and generate jobs.

Yet despite New Deal programs and some aid to the states, total government spending — federal, state and local — as a share of the economy throughout the 1930s remained at just under 20 percent. (Today, total government spending is more than 35 percent, a larger buffer against weakness in the private sector.)

Here’s the thing: Keynes didn’t even publish the General Theory until 1936. What Keynes believed in 1930 was different from what he settled on by 1936 (for example, if I understand correctly, his emphasis on fiscal policy over monetary grows in that time period). So, the New Deal may be portrayed as an embrace of Keynesianism, but it wasn’t, at least not until around 1937. Alan Brinkley documents this story excellently in The End of Reform. In the 1933-1936 period, Roosevelt and his economic team were focused on managing separate crises with separate interventions. Jobs programs were conceptualized as just that – ways to help unemployed people make a living while doing something moderately socially useful. The Keynesian notion of aggregate demand, and its importance in the economy, enters into New Deal thinking relatively late, with the sudden downturn in 1937 following Roosevelt’s attempt to balance the budget. So, Roosevelt’s early actions (1933-1936) were not anything close to Keynesian, in intention anyway.

Add to that the fact that there wasn’t really a firm conception of “the economy”, specifically “the macroeconomy”, until the end of this period and the whole story makes even more so. Early New Deal policy was concerned with individual crises – declining agricultural prices, rising unemployment, the collapse of banks, and rising budget deficits (one of the acts FDR passed in 1933 was the “Economy Act” to reduce government spending. Economy didn’t yet mean what we now mean by it). These problems were tackled, albeit only partially, by a series of somewhat unrelated measures. The point of the WPA, TVA, etc. was not to stimulate aggregate demand but simply give people jobs who would otherwise be at loose ends and unable to make ends meet. Only as the concept of the macroeconomy becomes firmer, measured by official government statistics like the National Income figures produced by Simon Kuznets at the Department of Commerce in 1934, along with better national unemployment data, do Keynesian ways of acting on the economy become possible. A macro-level, above the individual market, becomes available as a site of intervention. Once unemployment is seen not just as a problem of people not being able to make a living, but rather as the source for a decline in aggregate demand which in turn affects private investment, etc. can “Keynesian” or even “macroeconomic” policies be implemented, rather than a host of microeconomic initiatives (interacting with single markets or problems at a time). So, yeah, the early New Deal wasn’t very good at being Keynesian.. but it wasn’t trying to be, nor could it have been.
Every time a historian writes about how pre-1930s economists were trying to understand “the economy”, or about whether or not some policy was “Keynesian” substantially before the General Theory, I get confused and annoyed*. So, to paraphrase Foucault, stop writing the history of the past in terms of the present, and start writing the history of the present!

* And yes, I know the article mentions Keynes visiting FDR to try and give him advice on what to do in 1934. But certainly the first 100 days stuff in 1933 can’t be judged against Keynesianism. No? Am I even coherent this late at night?