Scott Sumner  

Why do booms feel good?

A Question of Organizational L... What the Swiss Vote Really Sho...

The answer to this question might seem obvious, but it isn't. Yes, booms feature lots of jobs and income, but most standard macro models suggest that booms feature "excesses," with too much employment and perhaps over-investment too.

In the sticky wage/price model, recessions are bad because workers are working less than they'd prefer to work, and booms are bad because workers are working more than they'd prefer to work (in a non-distorted economy.) So why do booms feel so much better than recessions? Indeed even better than long run equilibrium?

The answer to why booms feel good is actually quite subtle. The economy is riddled with labor market distortions that cause people to work less than the optimal amount, even when wages and prices have fully adjusted to any change in NGDP. These include taxes on labor and subsidies for not working (implicit taxes on labor.) So the "normal" or natural rate of employment is far below the optimal level. A boom pushes employment closer to the optimal level, even if the result is not sustainable. (Eventually wages and prices adjust, returning you to the suboptimal "natural rate.")

Note that one cannot examine welfare issues here by looking at the behavior of an individual. Given the array of tax and subsidy distortions in the labor market, each worker may be "optimizing." But because those taxes and subsidies lead to external effects, the optimal level of labor supply for the individual will typically be below the optimal level for society.

I've noticed that a number of liberals have overlooked this problem when discussing the job losses that will result from Obamacare (now estimated at more than 2 million full-time equivalent jobs.) They claim that this sort of job loss is a good thing, because workers will be responding optimally to the incentives created by the law. I don't doubt that some workers will respond optimally, given their situation, but that doesn't mean the loss in employment will be optimal from the perspective of society.

Keynesians tend be especially aware of the asymmetrical effect of business cycles on welfare, much more so that real business cycles proponents. They understand the role of sticky wages and prices, and that the optimal level of employment is not the average level experienced in the US, but something closer to the boom level. Thus it's ironic to see Keynesian less aware of the dangers of Obamacare than are the real business cycle economists.

Patrick Sullivan directed me to an excellent Wall Street Journal article that discusses the views of Casey Mulligan, a famous RBC economist:

Instead, liberals have turned to claiming that ObamaCare's missing workers will be a gift to society. Since employers aren't cutting jobs per se through layoffs or hourly take-backs, people are merely choosing rationally to supply less labor. Thanks to ObamaCare, we're told, Americans can finally quit the salt mines and blacking factories and retire early, or spend more time with the children, or become artists.

Mr. Mulligan reserves particular scorn for the economists making this "eliminated from the drudgery of labor market" argument, which he views as a form of trahison des clercs. "I don't know what their intentions are," he says, choosing his words carefully, "but it looks like they're trying to leverage the lack of economic education in their audience by making these sorts of points."

A job, Mr. Mulligan explains, "is a transaction between buyers and sellers. When a transaction doesn't happen, it doesn't happen. We know that it doesn't matter on which side of the market you put the disincentives, the results are the same. . . . In this case you're putting an implicit tax on work for households, and employers aren't willing to compensate the households enough so they'll still work." Jobs can be destroyed by sellers (workers) as much as buyers (businesses).

He adds: "I can understand something like cigarettes and people believe that there's too much smoking, so we put a tax on cigarettes, so people smoke less, and we say that's a good thing. OK. But are we saying we were working too much before? Is that the new argument? I mean make up your mind. We've been complaining for six years now that there's not enough work being done. . . . Even before the recession there was too little work in the economy. Now all of a sudden we wake up and say we're glad that people are working less? We're pursuing our dreams?"

Tyler Cowen also has an excellent post on this:

A lot of Keynesians try to maintain the communication of the feeling (if not the outright statement) that demand-driven employment shocks have very little to do with the choices of workers but that is closer to wrong than right. (By the way, sarcastic comments about soup kitchens causing the Great Depression belie an understanding of both this argument and of contemporary search models for the labor market.)

OK, given all that, when those workers, hit by negative shocks, do not rush to go back to work at lower reservation wages, we then read a portrait of hysteresis, despair, and soul-crushing joblessness, a psychic swamp so difficult to escape that even summoning up the strength to go back to work may be difficult.

In other words, would-be workers irrationally undervalue the benefits of having a job and they also underestimate the costs of remaining unemployed.

Now let's switch settings. A benefit shock comes along, positive for many people, and it induces many of them to work less or not work at all. How happy should we be? And here I mean happy at the margin, due to their change in employment decision.

People, it is rather difficult to have it here both ways.

All I would add here is that it's an externality problem. And that means that one cannot look at the motivations of individual workers responding to the Obamacare incentives as providing any sort of reliable estimate of the social impact of the loss of jobs.

In countries with great disincentives to work (parts of Europe), people work less than in America. In regions with fewer disincentives to work (parts of East Asia) people work more than in America. Obamacare moves us closer to the low employment European equilibrium. Liberals may like that, but then they need another explanation of why a boom feels better than an economy operating at the natural rate.

Comments and Sharing

COMMENTS (24 to date)
Richard writes:

This is an absolutely brilliant post. The point about how so-called boom times actually reflect something closer to what the economy would look like without governmental disincentives to work is stunning, thought-provoking, and original. A real paradigm-shifter.

Yancey Ward writes:

The optimal thing for me is to be paid not to work and with regular cost of living adjustments and merit increases.

What I have found interesting is that the arguments for the employment reductions due to ACA being a good thing could just as logically be applied to argue that everyone should be paid to quit their jobs.

Rob Rawlings writes:

'So the "normal" or natural rate of employment is far below the optimal level. A boom pushes employment closer to the optimal level, even if the result is not sustainable.'

I disagree with this.

Assume things like UI push the supply of labor curve to the left. Along comes a boom and shifts the demand curve for labor so that more labor is demanded at current wage rates. As a result we move along the supply curve so that both wages and labor supplied increase. This is likely to make the suppliers of labor happier. It would also move us closer to the employment level that would exist without either boom or labor market interventions.

However even if there were no interventions in the labor market then a boom would still cause a move up the labor supply curve and workers would still be happier as a result. But in this case we move away from, not closer to , the true equilibrium.

In other words its the boom causing the demand for labor to increase not the fact that we may be closer to some optimal equilibrium that increases worker happiness in booms. Surely any supplier is happy when demand for his product increases?

Rob Rawlings writes:

To clarify: I don't disagree with the bit I quoted but with the expressed view in the post that the reason booms feel good is that they moves us closer to the true equilibrium.

Glen Smith writes:

Those that underestimate the costs of remaining unemployed are not those unemployed.

Andrew_FL writes:

@Yancey Ward-

What I have found interesting is that the arguments for the employment reductions due to ACA being a good thing could just as logically be applied to argue that everyone should be paid to quit their jobs.

This was my reaction, too. Saying that people work too much because they want to acquire healthcare, is like saying people work too much because they want to acquire food.

Well, I guess we aren't to far from government provided food, then.

Nick Rowe writes:

One more reason booms are good: monopoly power. This is central in new Keynesian macro.

But there is another side of me that says booms don't really exist. They are just the absence of recessions. And booms are good, simply because they are not bad. Plucking model. Arnold's PSST.

Brandon Berg writes:

The left-wing defense of Obamacare's disincentives makes sense if you ignore the costs to taxpayers. Which is a pretty good model for the left-wing mindset.

Joe Teicher writes:

Maybe I'm weird, but I don't get this idea that booms "feel good." To me a boom means more crowded train rides and fewer good deals in lots of markets. It also means a loss of relative status for people in less cyclical industries. What aspect of booms are actually supposed to feel good?

Travis Allison writes:

I think one can differentiate between a person who wants a job or is not working enough regardless of Obamacare and a person who would like to work less if they had health insurance. So I think Mulligan's point is irrelevant.

Well, Joe Teicher, maybe the ease of finding a better job? During the late 90s boom there were stories about jobs that young mothers could bring their infant children to, or even their pets. Ex-convicts and the elderly were employable too.

Also, Keith Hennessey has a related post up, Ladder v. Safety net;

The costs of all three of these policies [Extended UI, Minimum Wage raises, subsidized health insurance] include higher structural unemployment and fewer people building additional skills to move up the income scale over time. When the U.S. economy eventually recovers fully, our unemployment rate should be in the low 5s. Because they keep layering on “protections” and “assistance,” France’s comparable rate is around 10 percent. Imagine if the U.S. steady-state unemployment rate were 10 percent. We’re not there yet, but all of President Obama’s policies push us toward a European-style model.
I think movement in that direction is a huge mistake, but my point today is a more basic one. These trade-offs must be considered and debated openly, and the Obama Administration is doing a disservice by suggesting that no trade-offs exist, and that those who oppose these programs do so because they are mean. Extending unemployment insurance benefits, raising the minimum wage, and ObamaCare have long-term labor supply costs that must be weighed against their more immediate benefits. There is no free lunch here. Do you want a stronger ladder or a higher safety net?

Nathan Smith writes:

It seems to me there's a big difference between not working because one can't find a job, and not working because working means losing health care subsidies. Both seem bad and inefficient, but they would "feel" pretty different, both to the people experiencing these different kinds of joblessness, and to society looking on.

The default attitude to the unemployed is pity. Poor guy, he can't find a job! And the unemployed guy feels useless and anxious about the future. The default attitude to a voluntary non-worker living off the government is anger. Lazy bum, he won't work, he'd rather mooch off the rest of us! And the guy living off the government may like his life well enough.

Yancey Ward writes:
I think one can differentiate between a person who wants a job or is not working enough regardless of Obamacare and a person who would like to work less if they had health insurance. So I think Mulligan's point is irrelevant.

But then you can make the same argument with regards to food, shelter, clothing etc.

ThomasH writes:

An excellent post and explains why Republicans to the man have demanded an increase in the earned income tax credit to offset the disincentive to work created by allowing people to access the subsidy for health insurance without being employed full time.

James writes:

"A boom pushes employment closer to the optimal level, even if the result is not sustainable."

What evidence leads you to believe this?

I'm sure that someone could describe a scenario where, assuming the right utility functions, production functions, tax policies, etc., a boom can cause a shift from a low utility world to a high utility world because some new distortion has effects that go against the effects of previous distortions. But this post offers no evidence to believe that this is what is happens during actual booms, nor any citation of any other work offering any such evidence.

Andrew M writes:

Is there any evidence that wages are more or less sticky in downward moving NGDP scenarios than upward?

To illustrate
Upward- 1. Companies see rising NGDP and lucrative investment opporuntities -> 2. companies hire more employees -> 3. Economy approaches optimal employment (the upper bound, if you will) -> 4. Workers may now ask for raises, higher starting salaries, etc -> 5. due to higher wages, employment starts to fall back to natural rate

Downward- 1. Companies notice falling NGDP and poor future investment outlook -> 2. knowing workers will respond poorly to wage decreases, companies reduce employment -> 3. Economy falls below natural rate of employment -> 4. raises are typically frozen, conservative starting salaries maintained, -> 5. Recovery

The stickiness question comes into play mostly in step 4, in both cases, however, in a boom environment step 4 can gain traction a lot easier, plus you have the upperbound on employment typically faster because you're not so far away to begin with.

Andrew M writes:

In the same vein of thinking as "lots of less hours worked being different if it's voluntary or not," does it not count for something that people who make this voluntary decrease in traditional employment may still be quite productive, especially domestic and DIY work? "Leisure" time is not necessarily leisurely. Whereas with involuntary decrease, finding a new job is now your full time job and that's definitely quite unproductive.

Basically stop being grouchy about people not needing to hit the magical 40 hour work week just for health insurance, that's silly. Don't be a jerk.

Scott Sumner writes:

Thanks Richard.

Rob, You said;

Surely any supplier is happy when demand for his product increases?

"Not a supplier with sticky wages."

Glen, I agree.

Nick, Why isn't it symmetrical? If wages are sticky wouldn't booms cause workers to work more overtime than they prefer?

Joe, You may be right in areas with negative externalities like transport. But labor has positive externalities due to all the tax wedges.

Travis, The point is that workers face multiple distortions, and Obamacare is just one of them.

Thomas, If only they had, they might once again be taken seriously.

James, OK, I cite Milton Friedman's Nobel lecture.

Andrew, I believe wages are more sticky in a downward direction. I agree that people should not have to work to achieve health care.

Rob Rawlings writes:


You say "Not a supplier with sticky wages." implying that a worker will not be happy if demand for his labor increases as he will not be able to get a higher wage as a result.

Assume this as true (though in reality workers may be happy to have better job opportunities/ more overtime etc).

Assume also that we start from a position "wages and prices have fully adjusted to any change in NGDP." and then have a boom created by a sudden and unexpected increase in NGDP (that the CB chooses not to suppress).

Why will a worker feel better about this in an economy with lots of dis-incentives to work (high UI etc) than in one where the market is free ? In both cases the amount of labor supplied just moves along a (flat) supply curve.

While it is true that in the case of the fettered labor market the amount of labor supplied will (as a result of the boom) move closer to the level that would exist in a unfettered market I do see any reason to think that this is the reason that people like booms. They either like the fact that the demand for labor has increased or (if you are right) they are indifferent between boom and non-boom irrespective of what disincentives exists in the labor market.

Michael Byrnes writes:

I would say that a very important reason why workers are happier during booms is they have more opportunities to "fire" the company they work for. (Mitt Romney isn't the only guy who likes to fire people - everyone does.)

During a recession, it's much harder to "fire" your company (if you need to find another job before you leave), so employees are more likely to remain in a job they would rather leave. Thus, less incentive for companies to keep their staff happy. Thus the staff is less happy.

James writes:


Since you did not protest my characteristic of your position, that booms bring new distortions that offset preexisting distortions, I assume I did not misunderstand you. Friedman's lecture does not support that position, so far as I can tell.

Was there a specific paragraph in Friedman's lecture that convinced you that the real world works this way?

Scott Sumner writes:

Rob, They like booms because it's not a zero sum game. During a boom living standards rise as the extra goods produced more than compensate for the disutility of losing some leisure. And that's because of all the distortions in the labor market. If there were no distortions then the utility from extra goods being produced would be less than the disutility of working more.

James, I meant that the Friedman lecture explains why the results are not sustainable. Regarding your other point, there is a mountain of evidence that labor market distortions exist, and theory predicts this would lower employment below the optimal level. Of course we cannot measure utility directly, so this theory is hard to test. It just seems to be the most reasonable interpretation of the evidence. In any case, my post was addressing a different problem, the asymmetry in the way many liberals regard booms as being good and also the lower labor market participation triggered by Obamacare as also being good.

James writes:


I never asked for a reason to believe that booms are unsustainable and I don't doubt that there are plenty of distortions in labor markets.

The only objection I've raised is to your claim that booms feel good because of offsetting distortions. But you've claimed that when a boom causes hours worked to move back toward the optimal amount of hours worked, this actually makes people better off.

Well, I can tell a story where the opposite holds: Everyone in my scenario works less than n* hours because under the existing income tax, their consumption-leisure utility is maximized when nlower indifference curve in their consumption/labor choice set. This is Friedman's story and it contradicts your position.

Do you have some similar story in which people could be on a higher indifference curve but they choose to be on a lower indifference curve most of the time, but then during booms they do something they could have done all along and get to that higher indifference curve? If you can't sketch out a scenario where this is possible, you probably don't have adequate reason to believe that the offsetting distortions increase welfare.

I'm only addressing your macro claim because you are a macro guy, and because that's the part I believe you got wrong. I could not agree more with your point about the inconsistency of how people react to Obamacare's employment effects. But you don't need to support your argument with a very speculative macro theory. Hard boiled mainstream micro is sufficient.

James writes:

Correction to my third para:

Everyone in my scenario works less than n* hours because under the existing income tax, their consumption-leisure utility is maximized when n is less than n*. Then some unexpected inflation is misperceived as an increase in the real wage so people work more. But the actual purchasing power of their wage has not risen at all so they wind up on a lower indifference curve in their consumption/labor choice set. This is Friedman's story and it contradicts your position.

Comments for this entry have been closed
Return to top