What a strange day yesterday was – I post a quick link from Krugman suggesting that sectoral readjustment can’t explain the breadth of the crisis, and suggest that this is a recurrent concern that Austrians need to address seriously. For raising the issue – the not uncommon concern with ABCT – I get accused of not knowing what I’m talking about, lying, posting in bad faith, failing to write an extended critique of the Austrian school, etc. I’m told “we’ve answered this” when what is really meant is “we’ve answered this to our own satisfaction”. Anyway, I did not have time to do such a post justice yesterday – I hope I manage it here.
The Thesis: Readjustment from worker mismatch across sectors does not explain the severity of the downturn
Jonathan tries to respond to this here. He offers two main justifications:
1. “the lengthening of the structure of production relative to one good can have the consequence of lengthening the structure of production relative to a myriad of other goods. In this fashion, the structure of production tends to expand over multiple sectors of the economy—a basic way of saying this is, different sectors become interdependent… [ad hominems]… For all intents and purposes, Austrian theory predicts unemployment across a wide variety of sectors, and this ultimately affects both lower-order and higher-order industries.”
I’m not sure exactly how this suggests a broad-based downturn. Jonathan is not careful to distinguish here between vertical and horizontal production (in the traditional sense, not in the x, y, z coordinate plane that Jonathan mentions in his post). The whole problem with relying on readjustment of a sectoral distortion to explain the downturn is that it only explains unemployment in certain firms – namely, higher order production in more roundabout industries. Jonathan ably explains why we can expect to see a lengthening of the production process across multiple industries. I would go a step further than him and suggest that Austrian theory gives an excellent case for expecting a lengthening of the production process across all industries. But this doesn’t do anything to explain unemployment in lower order production (Jonathan’s use of “for all intents and purposes, Austrian theory predicts” rather than simply “Austrian theory predicts” indicates he may even be aware this doesn’t explain it). It still only explains unemployment in the higher order firms. Instead of explaining how a lengthening of production processes across all industries leads to unemployment of both high and lower order production, he spends his time telling me that I don’t understand any of these fundamental points.
So Jonathan comes back to a lengthening of the production process. Fine. But when the malinvestment is revealed, what happens to the lower order production? He doesn't address the question, but I'll try. Well we know there’s no input bottleneck because (1.) there has been a lot more higher order production than necessary so there should be ample inputs, and (2.) if there is any demand for those inputs, then provision of those inputs wouldn’t be malinvestment in the first place and would have remained profitable even after other malinvestments were revealed. So it can’t be a bottleneck issue that allows unemployment in higher order goods sectors to spread to lower order goods. It’s also not a problem of a lack of labor or capital. The whole point of ABCT is that labor and capital that was used in higher order goods needs to be moved to lower order goods, and while this adjustment process takes time there’s no reason to think that there would be a fall in lower order employment in the interim, while the adjustment is taking place. The increase may not be immediate, but there’s no reason to think there is a fall, as Krugman points out in the data. Jonathan, too busy insulting my understanding of ABCT, failed to provide any reason to believe that a lengthening of the production process across multiple sectors would cause unemployment in both higher and lower order goods. I provided two possible reasons to believe that here, and then provided a refutation to each. Come on man – I’m doing your job for you. If you have any other reasons, though, I’m happy to hear them.
2. “If a box of nails [previously employed in house construction] is suddenly pitted towards another industry altogether (let us say, the construction of a lumber factory), then the stage it is at suddenly changes, as well. While the lumber industry, in this case, is relevant to the housing industry, these kinds of switches can occur in industries which may not be directly related (as suggested above). This, as also mentioned above, causes the structure of production to intensify across multiple sectors. As a consequence of this interdependence (for lack of a better word), it only makes sense that a fall in productivity would affect several different sectors of an economy, not just the housing sector.”
The second argument Jonathan provides is based on the switching of capital goods between industries. This one isn’t quite as clearly sketched out, but I think the point is that if the marginal productivity of the capital good (the nails) falls, because industries are interrelated production in all sectors will fall. Capital goods are substitutable and switchable across industries. But Jonathan fails to even consider any magnitude for the elasticity of substitution between sectors, and implicitly assumes such an elasticity is extremely high. Otherwise productivity shocks couldn’t propagate across sectors. So that’s one implicit assumption of his that he chooses not to share with his readers. It’s a bad assumption I think (and an odd one for an Austrian to make – Austrians usually emphasize the heterogeneity and specificity of capital and labor, something which of course I concur strongly with them on), but I’ll grant the bad assumption for the sake of argument. That still leaves the question of how a fall in the productivity of a higher order good, even if it can propagate horizontally, leads to unemployment in lower orders of production in the context of a readjusting capital structure. Again, the adjustment may not be instantaneous, but it still should not produce balanced unemployment. You could assume an infinite elasticity of substitution between capital goods and it won’t change the fact that lower order production will face expansionary and inflationary pressures, not contractionary pressures.
If Jonathan assumes that all capital goods at all stages of production experience a fall in marginal productivity I can certainly accept that as an explanation for widespread unemployment. That’s Real Business Cycle Theory. That’s New Classicism. That sort of thing could happen. But ultimately that’s an “aggregate supply” explanation that mirrors the “aggregate demand” explanation that he oversimplifies and ridicules. The difference is, there’s a theoretical framework based on liquidity preference underpinning a precipitous fall in demand. There’s no good reason to assume the sort of economy wide technological retrogression that would be needed to put forward an aggregate supply explanation like Jonathan seems to be doing here (by assuming a broad-based reduction in the marginal productivity of capital). And even if you could provide such an explanation (dubious in our current case - and he hasn't proposed one at all), that gives you unemployment but it doesn’t give you downward pressure on prices, so you’d still have that to explain as well (one explanation might be a surplus of goods left over from the boom, of course).
Ultimately, I still see no good reason to believe that any explanation based on sectoral readjustment or on the rebalancing of the structure production after the lengthening of the production process could be a primary explanation of this downturn. Reading Jonathan's post didn't change that, and I quite well understood Jonathan's post. That’s not to say such a process isn’t occurring. I would be surprised not to see that unemployment was worse in higher order production processes than in lower order production processes (and it would be really nice to have empirically-inclined Austrians look into this more rigorously!). But that’s ultimately not an explanation for the broad based unemployment we’re seeing.
Nevertheless, this doesn’t mean that Austrians are worthless. My claim and Krugman's claim was only that readjustment of the structure of production doesn’t explain the sort of unemployment we’re seeing. For some reason, in the criticism of my post, this got twisted into me dismissing the Austrian school as a whole.
Xenophon shared with me a link to a Bob Murphy lecture that I do think offers an explanation for broad based unemployment that Austrians can get behind: unsustainable growth and capital consumption. I didn’t mention this for two reasons:
1. It wasn’t what Krugman or I were talking about – the initial point was that the readjustment of the capital structure and ABCT can’t explain the downturn, not that an entire school of thought can’t explain it, and
2. I never really thought that the idea that inflationary monetary policy can create unsustainable growth belonged to the Austrians or was the "Austrian business cycle theory" – it’s one of those things like supply and demand or market efficiency that I never really think of as belonging to anyone.
So, as far as I’m concerned the initial thesis still stands and Jonathan didn’t really make much of a counter-argument. Nevertheless, it’s worth highlighting this point that Xenophon shares because it’s a good one. And honestly the worry about capital consumption is just icing on the cake. You don’t even need that. If growth was unsustainable across the board because of a loose monetary environment then when that monetary environment tightens (either because of an increase in money demand or a decrease in money supply… in this case we have the former), you’re not only going to have a decline in demand (a la Keynes), but you’re also simply going to have a lot of production that is revealed to be unsustainable.
Fine. That isn’t the Austrian business cycle theory, that’s basic bubble dynamics. I’ll certainly acknowledge that Austrians acknowledge this – I never claimed they didn’t. But it doesn’t change the fact that capital structure arguments offer no explanation for the kind of unemployment we’re seeing.
Another reason why I don’t readily associate this with ABCT is because of how many times Austrians drill into people that “malinvestment is not overinvestment!”.
I spend a lot of my time thinking about economics. It’s an interest, a calling, and a profession. I’ve come to many conclusions about economics, and those conclusions will change over time and have changed over time, but I don’t draw a conclusion lightly or without reason. I have absolutely no qualms about being ambiguous or "on the fence" if I honestly don’t know what to think. And if I think something, it’s because I’ve thought it through. When I blog, sometimes I’ll produce long posts, sometimes it’ll be just to share a quick thought or link. If you disagree with me, it’s not incumbent upon me to satisfy you on my time, and you really need to realize that and get over it. Don’t insult me by assuming I toss around ideas casually, and don’t have the hubris of assuming that because someone doesn’t agree with your argument they don’t understand your argument. Truth in the social sciences is extremely hard to come by. A lot of things remain quite ambiguous and are consistent with a variety of explanations. If you think you have an air-tight theory that someone would have to misunderstand to disagree with, then you probably don’t understand your own theory. Very few theories are that air-tight.
Two other things:
- I'm probably not going to respond much in this post unless I truly feel that a new argument has been offered that's convincing, or that there is a flaw in my own argument (as I did with Xenophon's point, which I appreciated him raising... I especially appreciate it because I haven't thought much of what I've heard from Bob Murphy in the past - probably because I only really have contact with things he's said on 1920-21 and the Great Depression - but the thoughts he shared in the lecture were often good). In all likelihood, I'm not going to post on this again either (unless there's a good counter-argument I haven't thought of that needs a response). I really need to spend a few mornings writing and working.
- I still haven't seen much reaction to Krugman besides on this blog and on Jonathan's. Does anyone else think that's weird? Usually the very term "hangover theory" is like catnip to Austrians. Feel free to direct me to any links I've missed.
Pigou was not a Pigovian
6 hours ago