The importance of a "necesssary, but not sufficient" macroeconomics

In the first chapter of his book Industrial Fluctuations (1926), A.C. Pigou wrote:

"It is common for popular writers to speak of discovering the causes of industrial fluctuations, but it is easy to see, without glancing at deeper philosophic difficulties, that there are here, even on the plane of ordinary scientific terminology, serious ambiguities and a great looseness of thought".

I think this "eureka!" of discovery and the looseness of thought go hand in hand in this case. Too often, people hit on an interesting or plausible theory of economic fluctuations and assume that fidelity to the initial romance of the "eureka" moment requires that it be the only relevant theory. This is a very odd position to take. The economy is a highly complex system. Many decisions contribute to output figures, the fluctuation of which is only the simplest expression of cyclical activity in the economy. The relationship between the supply of money, prices, output, wages, employment (should we be talking about extensive or intensive changes to employment?), etc. are even more complex. And this is only discussing matters at the most aggregated level. We can talk about differential behavior of consumption and investment. Or, we can talk about different kinds of investment. Austrians would focus on the structure of capital itself, from the early stages of the production process to the later stages. More traditional economists might put more emphasis on something like inventories, which are known to behave cyclically and have a predictable impact on factor demand (when inventories are high, there demand for factors of production is lower). All of these things contribute to the business cycle, so a theory accounting for just one of them shouldn't be expected to be sufficient.

Aggregate Demand and Current Unemployment

One of the things that inspired this post was an exchange between Tyler Cowen and Arnold Kling on the importance of zero marginal product workers and recalculation. Cowen specifically cites some characteristics of the unemployed population:

"In general, which hypotheses predict lots more short-term unemployment among the less educated, but among the long-term unemployed, a disproportionately high degree of older, more educated people? This stylized fact seems to point toward search and recalculation ideas, with some zero marginal products tossed in. Do aggregate demand theories yield that same data-matching prediction? I don't see it, at least not without being paired with a theory of concomitant real shocks."

Part of Cowen's problem is that he's trying to point to very specific facts about the business cycle and evaluate theories on the basis of that fact alone, with no consideration at all of what the theories were designed to explain. Keynesian notions of aggregate demand weren't even primarily designed to explain business cycles, much less the composition of the unemployed. They were designed to explain output levels and (somewhat less convincingly) employment levels (although as I've mentioned recently - I think the key to a better explanation of employment levels is continuing along the Keynesian liquidity preference path and talk more about corporate liquidity preference - the whole problem is Keynes didn't take Keynesianism far enough).

So does it bother me that this alone doesn't offer an explanation for more older, educated workers among the long-term unemployed and higher unemployment rates for younger, less educated workers? Of course not. Keynes doesn't go into detail about who is in the pool of unemployed workers, but how exactly does Cowen think Keynesians think this is going to happen? We still believe in downward sloping demand curves, after all! Lower marginal product workers are going to be let go first, and this may very well include older workers with very specific human capital that is no longer useful. If we think about the rehiring process, we'd expect shorter unemployment for younger, lower skilled, but more generally skilled workers, and longer unemployment for workers with very specific human capital (and probably higher reservation wages and fewer outside opportunities like school and starting a family).

So you don't even need that much more than a downward sloping demand curve and a sense of firm-specific human capital to get these insights, even from an aggregate demand perspective - and an aggregate demand perspective isn't even designed to get into these issues! Arnold Kling makes much the same point in responding to Cowen, arguing (among other good points) that:

"When a manufacturer has excess inventory, production workers have effectively ZMP [zero marginal product]. But for the economy as a whole, that is consistent with an aggregate demand story."

So there are really two points: first, demand-driven theories can explain quite a bit when they're combined with a little microeconomic common sense. Nobody has claim to the downward sloping demand curve. But second, a point that I would like to add to Kling's, is that there's nothing in the "recalculation story" of Cowen and Kling, or the "Garret Jones economy" that Kling mentions that contradicts demand-driven models that I know of! We have the luxury of buying into all theories that make sense and are not contradictory - we don't have to be picky. Pointing to evidence supporting any single theory hardly debunks the others.

"Necessary, but not sufficient" and the Austrians

Two instances recently made me think about this odd "I have to pick one" approach to macroeconomics, and the Austrian School. First, I commented on this post at Cafe Hayek (yes, I still go occasionally), agreeing with Russ Robert's "recalculation" explanation of the breakdown of Okun's Law, and then also offering my schtick about corporate liquidity preference as well. A commenter responded to me:

"All that being said, let's cut to the chase: if you like the Austrian "Recalculation" storyline [I asserted I did], then being so bullish on stimulus and multipliers and aggregate demand is pointless and nearly contradictory."

It was a line that you could only say if you thought that these theories were mutually exclusive. How can they be? What forces them to be? My answer, of course, was that I think the Austrians have hit on something and when we think about construction workers or something like that the malinvestment story may explain a lot - but when you think about the broader downturn and look at the other contributing factors, it simply doesn't account for why we are teetering on the brink of a second Great Depression.

Another confusing moment I had with the Austrian School came from reading this post by Jonathan at Economic Thought on Hayek on unemployment. Hayek writes:

"It was John Maynard Keynes, a man of great intellect but limited knowledge of economic theory, who ultimately succeeded in rehabilitating a view long the preserve of cranks with whom he openly sympathised. He had attempted by a succession of new theories to justify the same, superficially persuasive, intuitive belief that had been held by many practical men before, but that will not withstand rigorous analysis of the price mechanism: just as there cannot be a uniform price for all kinds of labour, an equality of demand and supply for labour in general cannot be secured by managing aggregate demand. The volume of employment depends on the correspondence of demand and supply in each sector of the economy, and therefore on the wage structure and the distribution of demand between the sectors. The consequence is that over a longer period the Keynesians remedy does not cure unemployment but makes it worse."

We can ignore the boorish sniping at a dead friend that can't defend himself for the time being, and focus on the theoretical critique. Perhaps Jonathan can humor me and explain what exactly Hayek has refuted Keynes on here. What in Keynesian demand management eschews these market specific concerns? And how exactly does Hayek claim that "the Keynesians [sic] remedy" makes unemployment worse? What assumption is required for that claim? Well, as far as I can tell the required assumption is that recalculation and malinvestment problems dominate demand deficiency as an explanation of unemployment. If most of the problem is malinvestment, then I imagine fiscal stimulus could make unemployment worse. But if malinvestment is only a small aggravating factor, or a factor that explains the distribution of the unemployment across sectors rather than the ultimate cause, then there's no obvious reason why fiscal and monetary policy should make unemployment worse. The point is, this is an empirical question. A priorism is great - it can tell you a lot about how a system works. But it's not going to give you insights into how a system is working in a given historical instance. You need evidence to make a statement about that.

So what empirical evidence is there regarding Austrian business cycle theory? I haven't done anything like an exhaustive search, but quite little, and quite uninformative. This isn't surprising, since many Austrians themselves deny the very logic of testing their theories empirically. Two examples I found are here and here. Both are in the Review of Austrian Economics. Both strike me as somewhat inadequate. What they demonstrate is that monetary policy shocks explain business cycles and impact relative prices. This doesn't seem to me to be proof of ABCT specifically, although it is consistent with it. What I would have hoped is that someone would demonstrate the relationship between the monetary policy and the structure of production, and then the adjustment of the structure of production back to an equilibrium state, causing a business cycle. Is there any work out there on that? I suppose I'm saying that I find ABCT plausible, I'm just curious about the amplitude of the cycle. I don't think any Misesian a priorism can provide that.

UPDATE: Comments are closed on this post, because of excessive, unrelated commenting we've been receiving. I'm very interested in looking empirically at changes in the capital structure. I've been thinking potentially of something using BEA input-output tables, which are basically snapshots of the capital structure. It would almost be a reunion of the Kiel School of economics and the Austrian School of economics, two very similar approaches that (I think mostly for historical reasons) ended up as ships passing in the night. Anyway - anyone with substantive thoughts on this knows how to grab my attention - sorry I feel compelled to close this particular post.