Just posted this to my blog. Obviously Krugman is wrong, just wanted to see what people here thought about my rational as to why.
Paul Krugman just posted an argument on his blog in a post titled “Who’s Unemployed?” that the problem with unemployment isn’t structural at all because increases in unemployment were uniform across all sectors from 2007 to 2010.
In a perfect example of how numbers can tell lies, he hides the nominal numbers in ratios. His data doesn’t make the argument that we don’t have too many construction workers who need to find jobs in other industries, rather it makes the argument that the degree to which things worsened for construction workers between 2007 and 2010 was no worse than in other industries. It doesn’t tell you anything about the comparative employment prospects for construction workers vs. other industries, just how those prospects changed from 2007 to 2010.
One way to compare the employment prospects for construction workers to those in other industries might be to compare their unemployment rates. Why didn’t Krugman reference those numbers rather than the change in those prospects? Because they tell a much different story. The unemployment rate for construction workers in 2010 was 24.7%. The national average was 10.6% and the next closest industry of those Krugman compares was leisure and hospitality with a 14.2% unemployment rate. Education and health services had unemployment of 5.5%. Of course, as Krugman would tell you, nothing in those numbers suggests that we have too many construction workers and not enough nurses because in 2007 unemployment in education and health services was onlyl 3.0% compared to 7.4% in construction.
Note: I can’t be sure because he didn’t link to where he got his date from, but I believe Krugman got his 2010 numbers from here, rather than the most up to date number which I linked to above
I don’t think Krugman is arguing that the problem isn’t about shifting workers to new industries. I wrote a reply of my own to an older post of his, although it’s a bit open-ended,
I agree with Paul Krugman. Well, I agree as far as arguing that today’s unemployment problem is not one of structural impediments, or problems of training construction workers into nurses. I also agree with him (I think) that today’s unemployment problem is one of general productivity. I have made the same argument before (where I, ironically, disagree with Krugman’s, DeLong’s, and Thoma’s belief that present cyclical unemployment can become structural), “[w]e therefore conclude that all long-term unemployment is the result of wage disequilibrium or insufficient productivity in the capital-goods sector.”
The fact is that labor is one of the most non-specific capital goods the market enjoys. Of course, there are barriers between an average construction worker and an average doctor that need to be overcome by years of training, but the transition in the labor force occurring today isn’t one between low skill and high skill employment, it’s still one of low skill to low skill — or, jobs that are relatively easier to train for (and jobs which employers are usually willing to train as a part of the employment contract). The problem is with the lack of available jobs of this type, not with a problem in the transfer of skills.
How to overcome the economic stagnation which has fed long-term unemployment is another topic altogether, but by agreeing on the fact that present unemployment is mostly cyclical, at least we can focus on the important subject — how to stimulate job creation (and whether artificial stimulation is even beneficial). Then, maybe economists will begin to realize that the actual debate is academic and theoretical — how can you prescribe policy solutions if the theory they are based upon is highly contested?
As an aside, I’m wondering why was unemployment so high for construction workers in the first place?
It seems to me that an important number has been left out of all the analyses. How many people were working in each sector?
Think about it. Let’s say 100 people are nurses, and a million are construction workers. “Everyone else” comprises 100 million people. Unemployment for nurses is 50%, for c. workers it 10 percent, and for everyone else it is 1 millionth of a percent. So we start off with 50 unemployed nurses, 100K unemployed c. workers, and one unemployed everyone else. Then everything doubles. There are 100 unemployed nurses, 200K unemployed c. workers, and two other guys.
According to Krugman and Jonathan, the unemployment problem is across the board, since it doubled for everyone. According to the OP and Esuric, the nurses are crushing the economy, with their 100% unemployment rate, way above everyone else. But obviously it’s really the construction workers that are the problem.
Conclusion: Percents don’t count, it’s the raw numbers that have meaning. So until we have those, I don’t see how we can draw any conclusion.
According to Krugman and Jonathan, the unemployment problem is across the board, since it doubled for everyone.
That’s not what I’m arguing, and I don’t think that’s what Krugman is arguing either. The argument is that, yes people of some type of work are disproportionally unemployed, showing a shift in employment patterns. The focus, however, shouldn’t be on the structural changes, because this doesn’t present the problem. The problem is in the lack of productivity, and the lack of new jobs in other sectors of employment.
TY for the clarification. I agree that lack of productivity and lack of new jobs in other sectors is the problem.
I do think that you are givng Krugman more credit than he deserves. But since he is not explicit about what he thinks is the problem, not much to be said.
Structural unemployment is closely tied to the concept of malinvestment. The Austrian position is that there is no clear dividing line between cyclical and structural unemployment; they are interrelated. Krugman, on the other hand, is trying, and has tried in the past, to dismiss the concept of structural unemployment altogether. This is because the existence of structural unemployment, during recessions, is quite problematic for the Keynesian under-consumptionist framework, where recessions are caused by inadequate aggregate demand. In other words, they argue that recessions are characterized by general gluts, rather than relative disproportionalities, and therefore try to marginalize/dismiss all empirical evidence which suggests otherwise.
So I don’t agree with this assertion. The existence of profound structural unemployment is quite relevant, especially for an Austrian.
However, when you refer to structural unemployment you are suggesting that the cause of the unemployment is non-transferrable skills and sudden changes in the structure of the economy. The latter is correct, but the reason those people are not being employed is due to a lack of productivity, moreso than structural challenges.
Btw, Krugman doesn’t deny the existence of structural unemployment. That’s outrageous. It was Krugman, DeLong, and Thoma, after all, who argued that if we didn’t cure our cyclical unemployment now, then we would be stuck with a structural unemployment problem.
This tells us absolutely nothing. All recessions are characterized by a lack of productivity, i.e., disinvestment, but the point is that this lack of productivity is highly uneven; it varies from one sector to the next (the structural changes/corrections cause relative disproportionalities in productivity). In other words, the problem is not that there is a general lack of productivity, though this does exist to some degree, but rather that productivity has collapsed in some industries relative to others (residential real-estate, manufacturing, financial services, construction, etc).
I didn’t mean to suggest that Krugman entirely dismisses the concept of structural unemployment, though it may have came off that way; only that he downplays the existence of structural unemployment during recessions, especially this recession, because it’s highly problematic for the Keynesian under-consumptionist framework.
Krugamn denying/downplaying the existence of structural unemployment during this recession:
It actually tells us a lot. The problem isn’t one of non-transferrable skills, which is what structural unemployment is defined by in your general macro textbook. There is unemployment as a result of malinvestment in certain sectors of the economy, sure, but the problem isn’t of non-transferrable skills, the problem is one of a general lack of productivity. If we were to regain our pre-2007 productivity, or at least operate in an economy with a growing number of jobs, then non-transferrability of skills would be irrelevant, because employers would be willing to train their employees. More generally, it should be considered that most jobs lost and most jobs gained would have be characterized by employees with “special” skills, but “special” skills of a general low quality (relatively easy to acquire). We’re not talking about a booming medicinal field, or a rise in productivity in law firms; we’re talking about a general, economy-wide rise in productivity in usually industrial tasks, where training ranges within a few weeks to a couple of months, and where in booming economic times employers are willing to take on employees, even if they are not experts in their field.
Garisson and Hayek are absolutely right, as I suggested in a former post of mine in this thread. However, structural unemployment refers to a specific unemployment problem, that I don’t think is ever an issue. The problem isn’t expressely changes in the structure of the economy, it’s the speed at which these changes are occurring at. New jobs are not being created quickly enough, for whatever reason that may be (a reason we distinguish as intervention, malinvestment, et cetera).
I think Krugman agrees, by the way. He writes in a post you link to, “Maxine Udall responds to my writing about structural unemployment and the absence thereof by arguing that the US economy probably does have some major structural problems. She’s right, of course…” For Krugman, the source of the issue is different than that of an Austrian, but in general the message is the same. For Krugman, the problem was a steep fall in aggregate demand, seen mostly in the housing sector, but really seen throughout the economy. The solution to re-employing these “idle resources” (including labor) is increasing aggregate demand. Whether this be in the housing sector, the book publishing sector, the medical sector, et cetera, I don’t think Krugman cares. In his macroeconomic vision, the solution is in recovering that income gap created between 2007 and 2008 (within the Hicksian IS/LM framework, which is evident in another post you link to). He realizes, to an extent, that the recovery in aggregate demand might occur somewhere other than the housing sector, and thus require a movement of construction workers to other sectors of the economy. He, like me, just doesn’t see the notion of “non-transferrable skills” (which is an integral part of the idea of structural unemployment) as the issue; he sees an insufficient aggregate demand as the issue.
I think the Austrian macroeconomic vision generally parallels that of Krugman, even though the specific details are very, very different. We don’t recognize a need for a “willy-nilly” rise in aggregate demand; we realize that the economy is structured and enjoys a price mechanism which guides spending along the lines most demanded by consumers. We also realize that socialized investment leads to a squandering of resources, not to the production of even more resources. However, we also should recognize that the unemployment problem is one of inadequate investment, caused by whatever reasons, and that a rise in employment either requires a fall in wages or a rise in productivity (or, more likely than not, a mixture of both).
I suspect that short term unemployment is from the reasons Esuric is emphasising, and long term persistent unemployment is from what Jonathan is emphasising.
I don’t think that’s the case. The cause of unemployment was malinvestment, recession, and the secondary effects of recession; the unemployed are characterized by belonging to certain sectors of the economy. The method by which to eliminate unemployment is through a mixture of a fall in wages and a rise in productivity. The key is, though, that the problem is not one of non-transferrable skills, it’s one of productivity.
wage rates and productivity are not two separate entities for which you can have a mixture of the two. A rise in productivity means a rise in wages, and a fall in productivity means a fall in wages. They are two sides of the same coin. So what you are effectively saying is that you need a mixture of a fall in wages and a rise in wages, which doesn’t make too much sense. Productivity determines wages, not employment (or unemployment).
So More accurately, what needs to happened is that wages need to adjust to productivity whatever it may currently be, i.e., wages need to fall in our particular case.
Only that the “capital stock” is shrinking, or not expanding at the “normal rate.” But this doesn’t tell us anything either, because it entirely ignores the concept of malinvestment, i.e., the fact that capital is heterogeneous and must therefore be organized. In other words, an expanding “capital stock” (increased investment) isn’t inherently optimal/desirable.
The data and Austrian business cycle theory suggest otherwise.
Returning to the pre-2007 capital structure would not solve our problems; it was the cause of our problems in the first place, which is why we have this recession (correction). So I don’t understand the argument you’re trying to make here. The pre-2007 capital structure was plagued by profound malinvestment due to monetary injections which arbitrarily altered relative prices and therefore the allocation of the original factors of production (land and labor) and capital towards ultimately untenable/unsustainable productions. The correction process, therefore, attempts to realign the structure of production, in-line with actual economic fundamentals/preferences. This means that labor must be redeployed from some sectors towards others, and it is why unemployment rates are more profound in some sectors (construction, real-estate, financial services, manufacturing, etc) relative to others; because those are the sectors where the accumulation of malinvestments was most dramatic.
In short, returning to the pre-2007 capital structure would not eliminate our problems, it would only temporarily conceal them. Additionally, structural unemployment could result from autonomous shifts in consumer demand or from technological innovations. But again, I ask you to refer to Garrison’s argument,
This is because, unlike the Monetarists, the Austrians do not adhere to the strict nominal/real distinction. Monetary changes affect both nominal and real variables.
This is a contradiction. Garrison argues that recessions are characterized by a condition of structural unemployment, as the capital structure is realigned, and as the original factors of production are redirected towards warranted and sustainable productions. Structural unemployment, therefore, is particularly relevant for Garrison; it is the main issue. The crux of his argument is that the strict distinction between structural and cyclical unemployment is somewhat spurious, because they are often interrelated, and caused by the same factors, namely monetary injections and the corresponding and inevitable readjustment process (the recession).
So no, you don’t agree with Hayek and Garrison.
Not at all. In fact, they are almost entirely different. Inadequate ineffective demand, which is what Krugman’s analysis solely focuses on, is, according to the Austrians, merely a potential consequence of a correction/recession. It’s only a secondary problem, if it occurs at all.
Yes but your analysis entirely ignores this.
A major Austrian insight is that over homogenization and aggregation leads to spurious and untenable conclusions. Hayek was correct when he proclaimed that aggregating and homogenizing all forms of investment into “I” “conceals the mechanics of change.” Again, the problem is not a general one; we don’t have too much of everything, and the demand for labor is not contracting proportionately across the board. This is a problem for under-consumptionist Keynesian doctrine, but is entirely accounted for by the Austrian framework.
Additionally, we know the cause of wide-scale involuntary unemployment (it’s not “for whatever reason”). The only way to correct this problem is to allow the structure of production to readjust itself to real economic fundamentals, which means that, at least in the short run, investment must contract (the capital stock must shrink), as malinvestments are liquidated. In other words, the correction will initially mean a fall in general productivity, but will yield long-term sustainable growth in the long-run. So just increasing productivity (investment) will not solve our problems; we must allow the economy to increase investment in some industries, and scale back investment in others.
I other words, your entire analysis is problematic.
wage rates and productivity are not two separate entities for which you can have a mixture of the two. A rise in productivity means a rise in wages, and a fall in productivity means a fall in wages. They are two sides of the same coin. So what you are effectively saying is that you need a mixture of a fall in wages and a rise in wages, which doesn’t make too much sense.
I’m talking about nominal wage rates, not real wage rates.
So More accurately, what needs to happened is that wages need to adjust to productivity whatever it may currently be, i.e., wages need to fall in our particular case.
I don’t see why productivity can’t rise without nominal wage rates falling first. Labor is a factor of productivity, sure, which is why a fall in wage rates would likely be a boost for productivity (since the firm can higher more laborers, up until the firm begins to see a fall in marginal returns per laborer), but if productivity can rise despite stagnating (or, “non-falling” wages, or “insufficiently low wages”) and the firm can then allot greater aggregate nominal demand for labor, then wages don’t necessarily have to fall as low as they would otherwise.
You are really missing my point (and I do agree with Garrison and Hayek), and trust me, I know the Austrian argument. I’m not arguing that there wasn’t malinvestment, or that the structure of the economy in the future will look different from that in 2007. In fact, I said this is the case in previous posts. So, I’m not sure why you continue to rant about these particular points — they are not in dispute. Yes, like I said before, the cause of unemployment and the method by which to re-employ these laborers are different in both sets of frameworks (the Austrian and Krugman’s), but this isn’t what is being discussed. What is being discussed is a more general topic: whether the unemployment problem is structural, or one of non-transferrable skills, or a problem of general productivity. That is, if the problem is only one of people which lack of skills, or the problem is one of lack of jobs. The problem is clearly the latter.
You misunderstand my use of “general productivity”. I’m not arguing that we should see a rise in productivity throughout all sectors of the economy. I’m arguing that the problem is one of productivity, full stop. That is, a lack of productivity, whatever the sector be that should see a rise in productivity, after the re-alignment of the structure of production.
By the way, I never argued that the solution to the unemployment is a rise in investment, regardless of where this investment might occur. I explicitely attributed that to Krugman, but suggested that the Austrians know better. My argument has always been that the problem is a lack of that “long-term sustainable growth” that you refer to in your last post — growth that is a product of investment, wherever that investment might occur.