Not Your Father's Real Business Cycle
Two helpful critiques of my post on non-hydraulic macro. I will address both of them in more detail below, but let me start by questioning the label “real business cycle model.”
Back before I tuned out of academic macro, Franco Modigliani could mock the real business cycle model as treating a downturn as a “mass outbreak of laziness.” You could get a real downturn from a productivity shock in a one-good, one-type-of-labor economy. Heterogeneity was not part of he story.
Imagine that we had a central planner. The planner decides what goods to produce and how to distribute them according to his own tastes. The planner does not know the optimal way to allocate resources relative to his tastes, so he arrives at an allocation by trial and error. After enough trial and error, he gets the right number of auto workers, doctors, teachers, and so on.
Then a major event comes along that makes his current allocation wrong. Until he is able to grope his way to the new best allocation, some firms have idle workers and some firms have shortages of workers. In this telling, unemployment results from a calculation problem.
Think of the stimulus package. Why isn’t it immediately putting people to work? Because the planners want to put people to work productively, and it takes time to organize projects that do so. They could have hired people 6 months ago to do nothing, but coming up with useful work takes years.
Now, substitute the decentralized market for the planner. We think of the market as this wonderful calculation mechanism, with wages and prices adjusting to get the optimal allocation of resources. But this calculation mechanism, like our planner, works gradually through a process of trial and error. Like our planner, it can be temporarily overwhelmed by the extent of reallocation required.
Think of our current economy as The Great Recalculation. Our planner (the market) is trying to figure out what to do with former mortgage brokers, real estate agents, manufacturing workers, construction workers, and so on. Which of them should wait for their industries to come back? Which ones should switch careers? Which ones need to go back to school? While our planner tries to figure it out, lots of folks remain unemployed, waiting for the planner to give them clear direction.
Perhaps the biggest policy difference between the hydraulic model and the Recalculation model concerns the value of bolstering state budgets. In the hydraulic model, sending money to the states keeps them from tightening their budgets, helps to maintain total demand, and therefore plays a really constructive role in mitigating the downturn.
In the Recalculation model, the challenge is to employ people who have been thrown out of work by the dislocations in the system. Since relatively few state and local government jobs have been lost, but millions of private sector jobs have been lost, throwing money at the states seems unlikely to help with the transition. Karl Smith writes,
If people are retooling I see a huge demand for retraining. Or them accepting very low wages in a new industry but why persistent unemployment. Why doesn’t the labor market clear.
There are many labor markets. But why don’t they all clear?
My answer is that they are groping, and if wage adjustments were rapid,they might grope inefficiently. A wage drop that is sufficient to clear the market in the short run in, say, construction work, might be so drastic that it would exceed the drop that will be required when the recalculation is all finished. The adjustments that result from rapid wage and price movements might be excessive and destabilizing. They could cause people to shift back and forth too much, incurring all sorts of transition costs. Sluggish wage adjustment reduces the transition costs, but it leads to some unemployment.
Or, it could be the case that wages are just plain sticky and that is a bad thing. It could be that wage stickiness is something that impedes recalculation rather than helps it.
Smith goes on,
how do you get from here to the Fed being able to start a recession. The experience of the early eighties seems to clearly show us that the Fed can.
It is possible that the Fed did cause the recession of the 1980’s. It is possible that the Fed raised interest rates, and higher interest rates changed the mix of employment away from housing and consumer durables to other stuff. Maybe the market never even figured out what that “other stuff” was–later in the 1980’s interest rates came down, so housing and consumer durables were able to come back.
It is also possible that the Fed had little to do with it. It is possible that bond-holders got tired of being surprised by low ex post real returns, and so interest rates would have gone up regardless of what the Fed did. My personal opinion is that this is what would have happened had the Fed not explicitly changed course, but under Volcker the Fed got the process started sooner.
Finally, Smith asks,
How do we get the Great Depression from here?
We think of it as another Great Recalculation. Compare the economy in 1950 with that in 1930. Huge differences. Land has been taken out of local farming, and instead used for suburbs, with food shipped in from longer distances. Manufacturing is moving out of big cities. Women are a bigger part of the labor force. Many fewer people are being paid to lift and pound, and instead more are doing clerical work.
The theory of Great Recalculations does not imply that the unemployment is benign or that government should not attempt to offer relief. I could easily imagine fiscal or monetary measures that moderate the downturn, because they happen to stimulate temporary employment in ways that do not impede long-term adjustment. However, in practice, we have not yet had proven instances of fiscal and monetary success. (In any event, it is hard to prove success or failure, because we do not have controlled experiments. Already, we are seeing vehement arguments over whether the measures taken over the past eighteen months have worked, and we cannot settle these arguments empirically.)
The anonymous blogger at Free Exchange adds,
in the mean time, why should good and willing labour go unused? What digging out is being done, that involves millions of workers sitting around doing nothing while they’d like to be working? Are workers doing penance for the time they spent in, say, the housing construction industry? I suspect that most of the working population is wishing they’d done something differently in terms of occupation or education nearly all the time, and yet this typically doesn’t lead to 16% un- and underemployment.
Again, it is possible that fiscal or monetary measures can temporarily put people to work productively. However, I come back to the fact that the fiscal stimulus enacted earlier this year is coming on stream slowly. The market does not have a rapid solution to the Recalculation Problem. But the planner does not have a rapid solution, either.
It is certainly possible that the planner can do things to make the transition less painful. The challenge is to come up with ways to employ the most out-of-demand workers without repressing the signals that tell them how they need to re-tool themselves.
Aug 11 2009 at 9:08am
The best accounting of the stimulus thus far seems to be that it’s gone to keep medicare and unemployment payments online. Since these are in a sense unfunded mandates and states aren’t allowed to issue debt like the treasury, this seems more like a way of stacking one Keynesian turtle on top of another than anything else.
Aug 11 2009 at 10:02am
I’m not certain if RBC theory is really so completely crippled by the introduction of heterogeneous capital or multiple sectors as you seem to be suggesting (though I might be misreading you, or them, or both). I mean, the “duck eggs & chicken eggs” Robinson Crusoe story is pretty simple/standard, isn’t it? A productivity shock in one sector (half your ducks die) leads to a recession in both sectors (you eat more chicken eggs). Can you say more about why heterogeneity is such a fundamental problem with this way of thinking?
Aug 11 2009 at 10:21am
When I took econ 101, we were taught that there were 3 kinds of unemployment, natural, structural and cyclical. Arnold seems to be arguing that structural employment is significant. Smith argues that cyclical employment is significant. Both of these positions seem compatible to me. That is, monetary shocks can cause cyclical unemployment and “recalculation” shocks can cause structural unemployment also. I believe this is what Scott Sumner was pointing out when he said that a steady NGDP growth would allow allocation decisions to be made without price level shocks interfering.
Aug 11 2009 at 11:37am
Purely anecdotal, but I have lots of friends involved in this transition from construction. I know several people doing side work (non reported), some people scraping by on the little work available at greatly reduce wages, and some people waiting it out on unemployment. Some of those people on unemployment are actively trying to find jobs that will suit them (sometimes a little too picky if you ask me), some of them are hoping for a comeback in their previous employment, and some are trying to retrain/new education. Why does it surprise people that it takes a long time for this to shake out. A carpenter can’t go out and switch to being a paralegal overnight.
Aug 11 2009 at 11:44am
You don’t have to reinvent the concept of structural unemployment.
If there is a shift in needed allocation of resources, painful structural unemployment is normal, and if the natural unemployment rate rises, potential income falls. And so, we should get slow growth or even negative growth along with higher unemployment.
This is all standard in thinking about international trade. Painful structural adjustments. And when thinking about international comparisons, isn’t it obvious that one reason why European potential income is low is because their insitutional unemployment is high?
I think it is obvious that structural unemployment can rise and that this will depress potential income.
But, the other side of the coin of a reallocation of resources is that there are sectors of the economy with rising prices, rising profits, more production and more employment. The other side of the coin of structural unemployment is bottlenecks. The unemployed workers from the shrinking industries are somehow not right for the growing industries.
Of course this happens. Surely, the housing bubble suggests that resources need to be redeployed out of housing. But if this reallocation is the whole story, where are these industries with rising prices, profits, production, and employment?
Now, if you want to instead say that there are no growing industries, because people don’t want to buy as much of some of the things the firms were set up to produce, and they don’t know what they want to buy, then, you have a different sort of problem. This is exactly what the Keynesians (and especially post-Keynesians) focus on. And that leads to hydralic macroeconomics.
You are saying that people don’t want to consume as much, they just want to save.
The obvious response is that interest rates should fall until dissaving by some households and investment by firms matches this increased saving. Interest rates should clear this up. Some people buy more of some of the consumer goods that we are set up to produce. Presumably some firms will buy at least some capital goods appropriate to those consumer goods.
Sure, there can be strutural adjustment difficulties. But there will be firms with growing demand, rising prices, rising profits, rising production, and rising employment.
Suppose the demand to hold money rises with the lower interest rates and the quantity of money doesn’t rise? In the limit, suppose we hit the zero nominal bound on interest rates, so that excess demands for some assets shifts to a demand for zero interest currency?
If prices are perfectly flexible, then this situtation of people wanting to sell resources, earning income, and not buy anything now (including firms not wanting to buy capital goods now) results in lower prices and nominal wages, higher real money balances, increased real wealth, and more real consumption. Somebody will buy more of aome of the goods currently for sale. The shortages will match the surpluses, and we return to equilibrium.
Again, with this new, lower level of prices and wages, the real volume of expenditures will result in some industries having rising demand, risig relative prices, rising real profits, rising production, and rising employment.
Because prices and wages are sticky, that process is not working. And so, we have just about every sector shrinking. Real income is below an already depressed potential income. The unemployment rate is greater than an already higher natural unemployment rate.
My view is that a 3% growth path of nominal income is the best environment for needed adjustents in the allocation of resources. If potential income grows more slowly or even shrinks because of extra large adjustments, then price inflation in the face of stable nominal income growth is the least bad alternative.
While some nominal interest rates have been near zero, not all are. While the first best option is to have negative nominal interest rates on low risk and short term assets (like T-bills,) lower nominal interest rates on the riskier and long term assets, with a yield structure and risk premia reflecting market expections of risk and future interest rates, zero-interest currency makes this impossible. And so, because the Fed issues zero interest currency and all the rest of the money in the economy is redeemable into it, the Fed is going to have to buy risky and longer term assets, lowering those nominal interest rates, and shrinking the risk premia and yield curve. A committment to do this however much is needed to keep nominal income on its growth path will reduce the need to do it. That is because falling nominal income compounds anything else that might be causing drops in the nominal interest rate on short term and low risk assets.
It is possible that the Fed would buy up everything and all nominal interest rates are zero. Then, either the zero interest currency boil must be lanced, or else, the Fed must lower real interest rates by committing to higher future prices.
Fiscal policy can also fix the problem of “too much saving” or an “excessive demand for money.” Government borrowing is negative saving and at the zero nominal bound, the government can supply lower risk short term assets (like T-bills) and spend the proceeds. The problem is that rather than producing consumer goods or capital goods, government goods get produced. And interest must be paid on the added national debt.
Aug 11 2009 at 12:04pm
The thing about Sumner’s focus on NGDP that irks me is that it seems to wave off changes in values extrinsic of the general price level.
For instance, the collapse in the valuation of dot-com stocks in the early part of the decade seems to me to have very little to do with NGDP. Rather, there were a series of shocks as investors began to realize that many of these companies could not have generated profits under almost any circumstances.
At most, I would grant that the realization that NGDP might go into decline in the near future catalyzed this awareness and caused the adjustment in values to occur over days rather than months.
I imagine Sumner might argue that dot-com stock prices were a micro phenomenon and that other areas of the economy (health, finance) were going up as rapidly as they were coming down, and thus what I’m looking at all washes out, leaving NGDP as the supreme driver behind the curtain.
At base, this seems to me like an argument that we can’t all possibly go mad at the same time. I think that the natural human tendency to herding combined with the speed of modern communications makes madness something of a default behavior. 15 years ago, if you thought about refinancing you walked into a bank to talk to someone and waited for the paperwork to move through the mail. Decisions at every level took weeks or months. Now much of it happens in minutes. While this doesn’t change the underlying fundamentals per se, it does increase the potential for bubbles to develop based on sentiment.
FWIW, talking about economics and finance without talking about the effects of transaction speed seems to me like talking about the changes in the economy over the past 30 years without talking about the historically-low inflation of that period.
Aug 11 2009 at 12:44pm
A.K. quotes someone saying:
in the mean time, why should good and willing labour go unused? What digging out is being done, that involves millions of workers sitting around doing nothing while they’d like to be working?
This is absurd. Do you think that just because
they are unemployed millions of people are
Is this used as an argument for government
Those millions of people are actually, you know,
people. They do stuff with their time.
They will repair their homes and vehicles.
Spend time with their families. Take odd jobs
that are unreported. LEARN a new skill. Sleep.
Watch TV. Try to figure out where they can
best earn a living – i.e. Reallocate themselves.
If they are not “idle” then they won’t be able
to “reallocate themselves”.
I was “laid off” several years ago. It was one
of the best times in my life.
Aug 11 2009 at 7:46pm
I thank billwoolsey for the counter-argument.
The construction industry is the most cyclical of the major industries. It is impossible to eliminate cyclicality from it. The recession we’re having is first and foremost a real estate and construction industry recession. The last nine months manufacturing inventories data shows that to a large extent the output drops in manufacturing have been produced by drops in inventory levels as opposed to drops in actual end-demand — inventory drops that were in anticipation of an economy-wide recession induced by the real estate and construction sector.
If you agree that the recession has been induced by a cyclical downturn in real estate and construction primarily or exclusively, the hydraulic model remains attractive. As I see it, for the most part we’re having a garden-variety cyclical recession with a deeper trough. The structural adjustments that are in play are not much more profound nor much different than the structural adjustments that were happening in the good times before the recession.
Aug 11 2009 at 11:00pm
[Comment removed for supplying false email address. Email the email@example.com to request restoring your comment privileges. A valid email address is required to post comments on EconLog.–Econlib Ed.]
Aug 11 2009 at 11:50pm
I don’t think I agree with you about everything, and you neglect some important points and focus too much on others. But, I just wanted to say, this post is one of the best I have seen you write here, in my opinion. It’s as though something about your thinking on recent events has went through an underlying change in the last few weeks.
Aug 16 2009 at 7:12pm
One reason wages are sticky is that costs are sticky. My mortgage payment (or food or gasoline) didn’t go down 20% even if my salary at my next job does. So I have to decide whether to hold out for my reserve wage, accept a lower wage with my existing cost structure, or do something radical about my cost structure.
In the case of those of us in California, the best cost decision is “flee” — to lower housing costs, lower taxes — but not a decision that people make lightly. Just as your situation in Montgomery County, exit is very expensive — actually more expensive, since you could theoretically move to N. Virginia while realistically folks in LA and SF can’t move to NV or AZ and keep their same friends and jobs.
Comments are closed.