Why do recessions occur?… One answer points to Productivity
Noah Smith made a thought-provoking statement on twitter, “Sometimes I wonder what actually causes recessions.” He received a slew of responses.
While one could point to many causes, including tight money and the beats of butterfly wings, I look primarily to the dynamics of productivity. Here is a graph of year-over-year productivity growth…
You will see that productivity has a tendency to accelerate through the recessions. Productivity has a natural tendency to increase due to innovation. So why are recessions part of that process to higher productivity?
- On the supply side… Some firms continue with lower productivity processes. They have contracts. They have market presence. They do not replace equipment that is less productive. So there is a force to push these firms to either increase productivity or get out of the way.
- On the demand side… Effective demand constrains productivity. In the process of a business cycle, productivity gets limited by demand. In the sense that demand wears thin for increased production. Before many recessions, effective demand will expand allowing productivity to increase.
Recessions are a normal part of the cyclical growth process to push through the demand and supply constraints upon productivity. In the graph above, you will see productivity growth spurts after the recessions. Those are times when more productive firms can establish themselves more securely in the economy.
In China… Productivity has increased greatly there, right? Well, maybe… Productivity has slowed down greatly in China, and one could make a case that productivity is the most important problem in China. (link to WSJ article)
Eventually the less productive firms feel the increasing pressure from more productive firms. There is a battle for market share and profits. Labor share normally rises, Inflation will rise. The economy will overheat in this battle for productivity. Nominal interest rates may rise to try to cool it down, but eventually the more productive firms gain advantage through the reshuffling process in a recession as profit rates become too vulnerable for the less productive firms.
Some recessions are better at supporting the advance of more productive firms. After the crisis, productivity had a nice quick growth spurt but then fell flat as less productive firms were protected within accommodative policies, most specifically monetary… and the stagnation of real wages.
“Productivity has a natural tendency to increase due to innovation…”
“Productivity” is a ratio between output and labor input, represented by GDP and hours of labor, respectively. What is going on inside that particular black box at any particular time is fertile ground for speculation.
Not sure if there is much speculation to be had. Yes you have Labor, yes you have capital in the form of machinery, and you have materials. The process employed is throughput or transforming materials into a product. Maybe, I am being to micro as I live in that black box.
“Recessions are a normal part of the cyclical growth process to push through the demand and supply constraints upon productivity”
Go push this Austrian nonsense somewhere else. The “productivity” clearly follows the cycle – it doesn’t lead it. When a recession hits, firms reduce their workforces, and naturally get rid of the least immediately productive sections first (think about it – how is the productivity of R&D and training measured). When the recovery comes they cannot maintain production without new workers, who will take time to be integrated into their processes.
Up until recently, most recessions were caused by the Fed hitting the brakes as inflation started to accelerate (as supply constraints became binding). But because of, in my view, a mistaken belief that expanding the money supply is all that matters – regardless of whether that is done by expanding private or public debt – some recent recessions have been caused by reaching indebtedness limits.
All this is well known. Why do you think you need to explain a puzzle that isn’t one. You should read more Dean Baker.
I’ve also pointed out elsewhere that a recession economy is not the same economy as a normal expanding economy. Being a “low productivity” firm in a recession is not the same as being a low productivity firm in an expanding economy because the relative value of diffirent products changes in a “batten down the hatches” economy. I don’t think the firms that close down in a recession are necessarily lower productivity firms (in fact if you have been following the wage history of the US you should realise that plenty of relatively low productivity firms have done pretty well).
Ed is off his rocker again. Productivity is useless. That isn’t what drives economies.
I think it is more reasonable to believe that youe indicator is noisy than that worker’s ability to produce changes that much. How much does the chart smooth out if you divide current GDP by lagged labor hours? How much of what people do does not show up in this measure of GDP?
“There is a battle for market share and profits.”
You do not say how this battle is waged. Let me add to this a bit. Those companies manufacturing with more efficient throughput of product have a tendency to have lower costs. They can bring to market during high demand cycles a product at a lower cost than those companies with less efficient throughput. During high demand cycles, everyone profits.
During recessional periods when demand drops off, those companies with less efficient throughput reach their breakeven point much sooner than those companies with more efficient throughput processes, labor, or machinery (yes, all three count). Less efficient companies go out of business in the end. After a recession when demand picks up again, there is fewer companies available to manufacture and there would be a spurt.
Pick up Goldratt’s “The Goal”, an interesting and easy non-economist read.
My $.02 on manufacturing.
First, Dean Baker thinks there is lots of slack left. He is delusional. We will fall well short of the slack that he sees in the past. He bases his view on historic numbers but fails to see the new dynamic of a lower effective demand limit. Mark my words…I saw the top of the economic cycle last year, while he was saying that unemployment could go down to less than 4%.
Second, read run75441 above. He sees the process whereby low productivity firms are at a disadvantage to bounce back during a recession.
Third, It is correct for the Fed to raise the Fed rate through the business cycle. And the Fed has done a good job at getting the real rate near its natural level in the last 25 years. You cannot pin a recession on the Fed. There are deeper causes within the dynamics of competition.
It seems that if you lag labor hours, you will not get a true reading of productivity. The real GDP used corresponds to the same quarter’s labor hours.
Are you suggesting making real GDP correspond the labor hours in a different quarter?
What I’m saying is only that reported productivity is not a relationship between labor input in stable units and product output in equally immutable values. It is a relationship between measurements of labor input and product output in highly elastic units. Productivity can rise and fall in response to business conditions without any change in technology. The secular trend in productivity no doubt reflects technological innovation but attributing cyclical movements in productivity to innovation is, as I said, speculative.
Technically you are correct. There is more to productivity advances than mere technological innovation. Yet, if you broaden your view of what innovation is, then you include many other factors, like innovation in governance, organizational structure, employee benefit programs, supply chain management, etc…
Yes, I have a very broad view of innovation. I am especially concerned with cost shifting, as analyzed by John Maurice Clark and K. William Kapp.
“You cannot pin a recession on the Fed.”
I believe that Fed policies brought on the recession in 2000. Rather than repeat my arguments here, please see my comments to one of your prior posts:
In my view, our post war recessions were brought on by inadequately fine-tuned corporate planning and/or greed. Corporations see an improving demand for their products. They don’t want to be caught with insufficient product to sell, so they plan on increasing production. They plan and if demand still seems to be growing then they implement those plans. They see sales increasing and that confirms their original plans and increasing production. At some point ‘growth in demand’ starts to slow. But even when ‘growth in demand’ doesn’t ramp up as fast as they thought or actually slows, they will continue production thus building inventories. (Perhaps they assume that they have mistimed their ramp up in production but maintaining market share is important.) But at some point they must face reality, acres and acres of unsold autos or warehouses full of appliances are not acceptable. First they discount or lease but eventually they are forced to drastically reduce production. (They are behind the curve.) Then comes unemployment and recession.
Perhaps they do see an increase in productivity as all their equipment and personnel are engaged. But by that time the building inventory is a warning which they are ignoring. (Assuming that any well run corporation would never allow for inadequate production capacity and that if an error exists it will be for over capacity which can be left dormant.)
But I don’t believe that productivity causes any of this.
@John Cummings: Well actually, productivity IS what drives economics. That is, our ability to produce collectively far more than one of us could individually, to store, trade and preserve it, is at the root of everything that you or I would ever describe as economics.
Our cousins the great apes, and our friends the elephants and porpoises, are as bright as we are, give or take a few grey cells, but have nothing we would describe as an economy.
The real quarrel is twofold. Firstly, who benefits from all this extra? Secondly, how do we control, direct and reduce our collective productivity so we do not ruin our pretty planet?
The two problems are linked, of course.
Yes, the Fed rate went too high before the 2001 recession. They misjudged productivity potential. There were years of rising productivity and that was uncommon. They imagined that such increasing productivity would overheat the economy but they were wrong. There was still quite a lot of effective demand space.
Yet productivity was central to the Fed’s mistake. Eventually the productivity increases would have pushed against the limits and a recession would have happened.
Recession really are caused by a blend of productivity, effective demand limits, and monetary policy… but I stressed productivity in the post because it is not given its due role in recession dynamics.
You wrote: “Recession really are caused by a blend of productivity, effective demand limits, and monetary policy… but I stressed productivity in the post because it is not given its due role in recession dynamics.”
I agree with you on effective demand limit and monetary policy, but productivity just seems a bridge too far.
Once you accept productivity as a cause, you or others will find other factors which seem just as important. But this will dilute the perceived importance of the primary causes. Thus I see productivity as a distraction and a potentially destructive distraction one at that.
Unless you are hinting at the growing problem of automation. It may raise productivity but it will lower the ‘effective demand limit’. In that case I would blame the lower ‘effective demand limit’. As a philosophical question, if corporations can move most of their production to southeast Asia, I see no reason why they would not be allowed to totally automate in this country. As socially destructive as that might be.
Productivity creates pressures that require adjustment within the economy.
Also effective demand limit is related to productivity…
labor share = real compensation/productivity
Productivity hits the effective demand limit, then builds pressure over time to push through it. The result is a recession. But effective demand expands in a recession, which allows productivity to increase.
I believe that you are wandering into the weeds. (Smiling here)
“The real GDP used corresponds to the same quarter’s labor hours.”
I think an awful lot of the things you bought this quarter were not made this quarter. Your susbsequent comments make it clear that you understand the role that excessive inventory plays in the dynamics of recessions.
The last two recessions followed bubbles. When the bubbles popped (or just started hissing), people saw they could not sustain consumption and demand decreased. The feedback when businesses did not need as much labor created a recesssion.
Perhaps I will figure out how to create a chart of consumption per labor hour.
I figured out how to create the chart I wanted and it does not support my assertion. Personal Consumption Expenditures did not drop off until the start of the recession. Hours worked does drop off before the official start of all of the last 5 recessions (mildly in 80 and 81 compared to the more recent 3).
Construction drops off a year before the 2007 recession. It may slow a year before the 2001 recession, but not any more than at other times during the late 90s.
not so easy.
There were huge increases in manufacturing productivity in the two or three decades before the Great Depression thanks to the introduction of small electric motors and what we would now consider simple sensors and feedback devices like electric eyes and thermostats. In fact, productivity continued to rise through the depression.
There is a definite link between rising productivity and recessions. In fact, you’d expect periodic recessions as productivity rises. The idea is that when productivity rises, wages rise, and that rising wages increase demand. Needless to say, these don’t all rise in unison, so there will be booms and busts.
We’ve been seeing rising productivity as small computers permeate our businesses. Unfortunately, labor lost its pricing power in the 1980s, so each recession has a slower and weaker recovery that the previous one.
Lambert you should also think that the ultimate effect of the recession on productivity is not captured by the immediate measure, but the defining feature of recessions is unemployment. Low productivity workers no longer having work, may increase productivity, but the productive capacity of the economy is only increased if these unemployed workers can find more productive works afterwards. You can’t just close your eyes (as I have the feeling that (g)Libertarians often do), and make the rejected disappear. Similarly when a firm goes bankrupt that is a financial event (and you should remember reverse causality – often firms are less efficient because they have financial problems not only vice versa), the resources often are not destroyed, they just change hands without the debt hanging over them. They may be closed down in a recession, but may well be activated again afterwards. It is what happens in the recovery that is really telling for the long run productivity story, not what happens in the recession.
In fact this whole theme here annoys me. Financial limitations and real limitations are two completely different sorts of limitations (the one being current, the other often historical). Maybe it makes more sense to solve financial problems (like a lack of purchasing power in the right hands) with financial measures (like giving them money) rather than trying to change the rest of the system to adapt (including trying to pretend that financial problems are real problems).
And another fairly obvious, but not stated point – if you want to get rid of less efficient producers then higher wages (through tight labour markets), or lower prices (through increased competition) are the way to do it.
And finally one point to show that I am not totally one sided in my thinking. One thing that recessions might do for long run productivity, is make it easier for innovative startups in the recovery (but leaving cheap resources lying around). i.e. Schumpeter, did have a point. i.e. It might make the economy more versatile (not more efficient in aggegate, since many startups fail). Short run efficiency is not everything.
Sorry Edward for calling you Lambert above. Not intended.
I hear your point that low productivity worlers will just be hired back on after a recession but if you look at the graph, the high growth spurt of productivity is not undone by an equal negative fall in productivity. Far from it, the large productivity spurt is maintained.
That is a key piece of evidence.
No its not. (You can’t exactly tell from the graph, but the rate of growth falls back always after the brief spike, often staying low. And althought the spike is high, it is short.
P.S. It is even worse if unemployment just ratchets up. Even low productivity workers produce more when employed than when not.
P.P.S. If you want to make this argument, don’t map the rate of change. Map the level, minus the trend. Present that, then make the argument.
Productivity is output per labor hour. It doesn’t matter how many hours you have. The quantity of hours does not matter. Productivity looks at the quality of those hours.
After a recession the economy grows into that new quality of hours by increasing hours. That is after the growth spurt with increasing hours. That is when productivity gained through the recession will put real GDP onto a path toward a higher level of produvtive capacity.
You can do a search for my posts on the attactor level of productive capacity. Or I can find the link.
But what evidence (real evidence) do you have that you need to throw people out of jobs (rather than attracting them into new ones) to do this?
Whats more, why do you need to do it all at once, when there are no new jobs to go to, rather than steadily as part of normal churn.
It is nice to have new readers at Angry Bear. I know it can be hard to remember everything in a comment and once it is up, you can not change a comment. Can you nest your posts please? A common tactic of trolls is to post comments in rapid succession thereby flooding a thread with (sometimes) meaningless comments. Thanks!