Bryan Caplan  

"Wages Must Fall!": What All Good Keynesians Should Say

The Marginal Cost of Perfectio... Joseph Stiglitz Crosses Nick R...
When Keynesians want to gloat, they often point to the overwhelming empirical evidence in favor of nominal wage rigidity.  For the latest example, see Krugman on the Irish labor market.  Their unemployment is 14.5%, but the nominal wage index has only fallen by about 2.5%.  Krugman's conclusion:
It is really, really hard to cut nominal wages, which is why reliance on "internal devaluation" is a recipe for stagnation and disaster.
The gloating is easy to understand.  After all, nominal wage rigidity is the driving assumption of the Keynesian model.  Unemployment is just a labor surplus; since wages are the price of labor, the fundamental cause of unemployment has to be excessive wages.  And as long as the wage rigidity is nominal, you can neutralize it by printing money or otherwise boosting demand.

What's hard to understand, though, is Keynesian neglect of - if not outright hostility to - the logical implication of their argument: Wages must fall!  If they're right about nominal wage rigidity, it seems like "Wages must fall!" would be the mantra of all good Keynesians.  But few words are less likely to escape their lips.

Why would this be so?

1. Keynesians could say that nominal wage rigidity is such an intractable problem there's no point discussing it.  That's why Krugman emphasizes that "Ireland is supposed to have flexible markets -- remember, before the crisis it was hailed as an example of successful structural reform."  If wages won't even fall in laissez-faire Ireland, what hope does the rest of the world have?

There are two big problems with this story.  (a) Even if it's true, Keynesians should still militantly oppose any government policy - like the employer health care mandate - that increases labor costs.  (b) Government doesn't face a binary choice between conventional labor market regulation and laissez-faire.  There's a third choice: Low-wage interventionism.  If wages won't adjust on their own, why don't Keynesians ask government to actively push them down?  If that sounds too brutal, see Singapore for clever ways to numb the blow.

2. Keynesians could say that monetary and fiscal policy are easier to promote than wage cuts.  But Keynesians are the first to insist that fiscal policy is a valuable supplement to monetary policy.  Why not hail wage cuts as a valuable supplement to both?  At minimum, Keynesians should heatedly resist any government policy that pushes labor costs in the wrong direction - and remind us that "wrong" = up.

3. Keynesians could - and often do - retreat to the view that wage flexibility is a self-defeating solution to the problem of wage rigidity.  The idea is that wage cuts reduce demand, which in turn exacerbates unemployment. 

But this argument is full of holes.  As I've pointed out before, there are strong reasons to think that wage cuts will increase aggregate demand, making this solution doubly attractive.  Consider: Labor income equals wages multiplied by hours worked, so the effect on labor income is ambiguous; and as a matter of pure arithmetic, lower wages imply higher profit income.  In any case, if nominal wage cuts really are as rare as a blue moon, what makes Keynesians so sure that wage cuts would backfire if tried?  Without lots of empirical counter-examples, they have every reason to stick to the common sense position: "If wage rigidity is the cause of unemployment, wage flexibility is the cure." 

At this point, Keynesians could just bite the bullet: "Wages must fall!"  But in my experience they don't - and I don't think they're going to start now.  The reason, I'm afraid, is politics.  Keynesians lean left.  They don't want to say, "Wages must fall!" They don't want to think it.  "Wages must fall!" sounds reactionary - a thinly-veiled reproach to centuries of anti-capitalist intellectuals and militant unions.  After all, doesn't it mean that every "pro-labor" regulation and "victory for the workers" has an ugly downside - more workers unable to find any job at all?

Keynesians are right to ridicule people who deny the reality of nominal wage rigidity.  But they'd be a lot more persuasive if they put leftist qualms aside and focused on the logic of their own model.  Keynesians have every reason to rant against excessive wages.  They have every reason to rant against regulation that increases labor costs.  They have every reason to rant against unions.  And there hasn't been a better time to rant since the Great Depression.  Oh my Keynesian brothers and sisters, let us rant together.

P.S. I'm doing a Stossel taping in NYC tonight (12/15).  The show won't air until January, but I'm hosting a meet-up after the show at 10 PM, at Becco - 355 West 46th Street.  Hope to see you there.

COMMENTS (28 to date)
Steve Waldman writes:

Everything you say might be reasonable if it weren't for the existence of nominally contracted debt.

Whatever contribution falling wages might make towards recovery at a lower price level is thwarted by the, um, frictions associated with the fact that no one can pay their mortgages, credit card bills, student loans, precontracted rents, etc.

In a world with nominally contracted debt, reducing real wages via increased inflation is not the same as reducing nominal wages at a stable price level.

Eventually, of course, the sea is calm either way. But a million bankruptcies and defaults are a painful way to get there, perhaps so painful that politics will have turned the sea to ice before it has time to calm.

david writes:

New Keynesians, sticky prices, rational real wage rigidity, blah blah...

And, yes, nominal debt, too.

Mark Michael writes:

If nominal wages are sticky, but if the Keynesians run a loose monetary policy long enough they'll raise consumer prices, pushing down REAL wages. Perhaps this "stealthy" approach to "Wages must Fall! Wages must Fall!" is what the Keynesians are up to? Or they're counting on the time lag between expanding the money supply and consumer prices going up.

Actually, during the 1930s New Deal, we had high unemployment without much inflation, so Hoover's and FDR's policies sustained high real wages for those workers who managed to hang onto their jobs. The unfortunate unemployed workers saw huge drops in their incomes. Keynesian strategies aggravate income inequality, not reduce it.

The article mentioned that if Keynesians say that sticky nominal wages is the problem, then isn't encouraging flexible wages the solution? makes sense to me. Logically, the Keynesians should promote wage flexibility. Remove the thicket of labor law restrictions dictating work rules, higher wages and benefits, etc.

European countries like France, Spain have highly restrictive wage labor markets - and sustained high unemployment as a result. Spain's is over 20%; France is over 10%. Both have high unemployment even during prosperous times. The Netherlands, Germany are more flexible and have rates under 8% today.

Richard Veddar's and Lowell Gallaway's 1993 (updated 1997) book, "Out of Work - Unemployment in Twentieth-Century America" shows how the increasingly restricted labor market in the US after 1933 raised the level of unemployment from what it was before then.

For example, the 1920-1922 depression saw unemployment spike to 11% for a short while. GDP fell faster during that depression that the month-by-month basis than it did in the period from 1930 to 1932. Yet, the US rebounded faster and unemployment went back to normal levels by the end of 1922.

Another obvious point: As millions of Chinese began working in factories building things that competed with US & European factory workers, that specific labor segment suddenly saw a huge increase in supply. That naturally drove down the wages for those workers. Ditto when more & more areas of endeavor opened up to Chinese & Indian, etc. competition: computer programers, technical support, telemarketing, even accounting, Xray reading, you name it. Simple law of supply & demand, I'd think.

Tough on each segment of Western workforce as it expands, but millions of very poor Asians suddenly became much wealthier - at the expense of their Western counterparts. One looking down on Earth from above, might think that's "spreading the wealth around" like a candidate for President opined as a really, really good thing to see happen.

The free market spreads the wealth around better than Keynesian economists achieve with their approaches, I'd think.

Rajeev writes:

Thank you, Steve Waldman!

Per writes:

Great post Bryan!

Peter writes:

Nominal contracts is a good reason to prefer central bank action over lower wages. But if we have a central bank who stubbornly clings to their inflation target, lower wages is better than high unemployment.

Bob Murphy writes:

Bryan, not sure if you read the comments, but as a card-carrying Krugman Kritic, I need to endorse Steve Waldman's comment above. If Krugman read this post by you, he would go ballistic at yet more free-market economists who don't even bother reading his stuff. Meaning, Krugman has said several times that cutting wages won't work, because of nominal debt.

(I'm not saying his position is right or wrong, just that it would frustrate him that you are trying to catalog every conceivable response a Keynesian could give to your objection, and you left out the response he actually does give.)

Alex Salter writes:

"Everything you say might be reasonable if it weren't for the existence of nominally contracted debt."

This is true if people are living on the razor’s edge before the downturn. It seems odd, in a world where incentives are artificially skewed towards too much consumption, to take that as a given and then worry about nominally contracted debt. Why not focus on the bad incentives in the first place? At least we should be calling attention to it.

Also, while falling wages might make it harder for some people to repay their debts, rigid wages makes it far more difficult. With flexible (falling) wages workers have to scrimp more to make ends meet. With rigid wages there is widespread unemployment; no amount of scrimping can help. Whichever way you cut it, policy which prevents wages from falling exacerbates the situation.

Steve Dallas writes:

I don't buy assumption that the cause of unemployment is the price of labor. It doesn't necessarily follow that a drop in the 'price' of labor equates to an increase on employment.

As an example, take a fully staffed business. The argument seems to be that if wages drop, staff would increase. However, without any increased NEED for staff, any sane employer would simply absorb the excess as profit.

Raymond Tseng writes:
As an example, take a fully staffed business. The argument seems to be that if wages drop, staff would increase. However, without any increased NEED for staff, any sane employer would simply absorb the excess as profit.

This is totally wrong. I'll make an analogy to what you are saying. So because I own a television already, if Best Buy lowers television prices by 25% that means I won't buy a new television. Yes that may be true for me, but there are many people out there on the margin who may. You are basically saying that if Best Buy lowers television prices there will be no increased sales which is absurd. A drop in the price of labor will result in more employment on the margin not at a fully staffed business. There are many other businesses out there who are on the fence on hiring decisions and the price will affect them.

Floccina writes:
Everything you say might be reasonable if it weren't for the existence of nominally contracted debt.

Interest rates usually fall in a downturn making debt service easier so a 10% wage cut is not so bad but unemployment is still very bad.

Floccina writes:
any sane employer would simply absorb the excess as profit.

And then she spends it going to the spa and getting a massage and her nails done.

B.B. writes:

A Wicksellian would not say that wages have to fall, he would say that the real interest rate needs to fall. At a low enough real interest rate, intended savings equals intended investment at full employment.

It is the downward rigidity of the interest rate that is the problem. If the real interest rate is too high, all you get is perpetual deflation with no gain in employment.

Moreover, falling prices/wages with a fixed (zero) nominal interest rate pushes up the real interest rate, which just makes deflation worse.

Wage cuts only work in a static model.

Silas Barta writes:


Everything you say might be reasonable if it weren't for the existence of nominally contracted debt.

This is another example of where Keynesians avoid the obvious solution: people should stop entering fixed-nominal debt contracts! Let the existing ones default, and let new borrowers and lenders learn to use adjustable-rate loans.

Really, it's depressing to see the clever ways people can rephrase whining about how hard it is to pay debt.

Nathan Smith writes:

Steve Waldman reminds us of how perverse it is that policy encourages indebtedness. Abolish the mortgage-interest tax deduction. Quit subsidizing student loans, and make student loans bankruptable. Privatize Social Security so that people have more incentive to save for old age. Then we wouldn't have to worry so much about letting wages fall in a downturn.

I'm not sure Waldman's response is as strong as it appears at first glance. Yes, nominal debt makes falling wages problematic. But how many people would really go bankrupt because of, say, a 5-10% drop in nominal wages? And how many people go bankrupt because they're completely out of work? I suspect that reducing both nominal wages and unemployment might lead to less bankruptcies. Especially if, at the new reduced wages, people had opportunities to work overtime.

Mr. Econotarian writes:

President Hoover Dec. 3 1929:

"I have instituted . . . systematic . . . cooperation with business . . . that wages and therefore earning power shall not be reduced and that a special effort shall be made to expand construction . . . a very large degree of individual suffering and unemployment has been prevented."

Shane writes:

"laissez-faire Ireland..."

...Has the second highest minimum wage in the EU:

Render writes:

This might be a plausible expectation if it weren't for the other big gun in the arsenal - regulation-driven, fabricated administrative jobs for those who produce no value, and thus tend to vote liberal. The regulations and associated bureaucracy produce a time and money impact per additional employee, in addition to per wage dollar, putting a practical upper limit on the number of employees that it is feasible to hire. The "sweet spot" for the Keynesians is roughly on the curve defined by nominal wages and administrative burden.

PacRim Jim writes:

Keynesians would love to dictate wages and prices. It's difficult to differentiate them from socialists.

Lee writes:

Winners win.
Losers lose.

Unemployment is not something that affects everyone equally. Like most personal and social pathologies, unemployment is concentrated among the underclass. As you look higher up the scales of innate ability and educational achievement, unemployment becomes virtually unheard of. Even in the depths of our permanent recession, unemployment is rare among the bright and educated.

Unemployment is not the result of excess wages, but of idiots and losers assuming their natural station in life. No matter what you do, they'll still be idiots and losers. There is no way of arranging the pieces on the board so that they will be anything other then what they are.

There is nothing that we can do for them short of paying them not to breed.

Not everyone who does not have a job is an idiot and loser of course, but there will always be a segment of the population which is either unemployed, or is working in a menial position that is soon abandoned. Pretending that they are unemployed because of some structural deficiency in the economy is ridiculous.

Render writes:

Lee, while your undisguised contempt for humanity is amusing, it doesn't really add to the discussion. While your viewpoint might be considered true-ish in an actual free market, once you have our betters "improving" things, bad things ... even *gasp* unemployment ... happen to good people, and have been doing so with some regularity.

Snorri Godhi writes:

Perhaps the original post can be (over-)simplified as follows: unemployment can be reduced only by lowering real wages, and that can be achieved either by cutting nominal wages or by inflation.
It is probably true that the 2nd strategy is easier to implement, but that is no reason to try to increase nominal wage rigidity.

So far, so good. As for the objection raised by Steve Waldman (and Krugman channeled by Bob Murphy), that looks to me as the best reason to cut nominal wages rather than increase inflation. Note that my claim is stronger than those of Silas Barta (hi Silas!), Nathan Smith, and other critics of Steve Waldman.

My reasoning is as follows: Cutting nominal wages hurts employees with (fixed-nominal rate) debt. Hence, fixing nominal wages creates an incentive to take on debt. Further, inflation decreases incentives for people to save. Hence, fixed nominal wages combined with inflation drive a wedge between "the rich" (people with enough wealth to make it worthwhile to hedge against inflation) and people with no net wealth: the latter have an incentive to borrow, not to start saving; and people with only a little wealth lose it to inflation.

In short, I claim that Keynesian policies increase wealth inequality.

Walter Sobchak writes:
Steve Waldman writes: Everything you say might be reasonable if it weren't for the existence of nominally contracted debt.

Bankruptcy proceedings under Title 11 of the U.S. Code resolve nominal debt contracting issues. Sadly, we made the law too expensive and difficult in the last set of amendments, (fie on you Todd Zywicki), but it puts another issue on the Keynesians plate besides repeal of the Wagner Act.

Carl the EconGuy writes:

Rather than depress the nominal wage, there's the other alternative. Keynes writes (GT, pp. 141-42): "The expectation of a fall in the value of money stimulates investment, and hence employment generally, because it raises the schedule of the marginal efficiency of capital, i.e., the investment-demand schedule ..." In other words, in a situation of persistent unemployment, a true Keynesian should support a *very* expansionary monetary policy that would drive real wages down, to offset the sticky nominal wage problem. So true Keynesians should be screaming HIN (High Inflation Now!). Apparently, we don't have a problem that some serious QE cannot resolve -- if Keynesians are right that it's all about sticky wages, that is.

Lee writes:


It is of course possible to create poverty, unemployment and misery. The left is especially adept at this.

However poverty, unemployment and misery are primarily self-created, at least here in the US and other functional societies. There are surely competent and intelligent people living in sub-Saharan Africa whose circumstances are beyond their ability to improve or overcome. But we don't live there.

The failure of leftists, and Keynesians by extension, to recognize the self-determined nature of these problems in functional societies is the fundamental flaw that demolishes all of their other arguments.

They claim to desire some magical government intervention that will somehow fix broken people. They talk of lifting people out of poverty, which is ludicrous. They never acknowledge the fundamental truth that most of the poor are where they are because of WHO they are, and not because of some external influence or lack of opportunity.

In truth I'm not quite expressing myself quite as clearly as I should. I talk about intelligence and competence, and these things are of course necessary, but they are not by themselves sufficient. Character is the keystone that makes success possible. A person of good character and average ability will succeed. He or she isn't going to be an astronaut, but they will enjoy a happy and productive life. A person of bad character will fail in life. People of bad character who are blessed with greater intelligence simply use those gifts to rationalize their own flaws and to blame others. This is what makes someone a winner or a loser. Character is destiny.

Character however, unlike innate intelligence, can be influenced. An individual's family and upbringing has a profound effect, but so does the culture in which they live. If the leftists really wanted to reduce poverty, they wouldn't be trying to throw monkey-wrenches into the economy and doing other destructive things, but working to improve the cultural flaws that foster poverty to begin with, particularly among the underclasses.

But of course they don't actually want to reduce poverty, but to instead use the problem of poverty as an excuse and a distraction while they work to destroy our society from within, in large part by working to degrade and destroy our culture.

Scott Bowers writes:

Debt is collected in Court. Courts garnish wages. Wage garnishments have payment ceilings of a fixed share of after tax income. A maxed out debtor will therefore have his nominal debt payments adjusted downwards as his take-home goes down.

Liquidationist writes:

The nominal debt thing is a total red herring.

If we assume that we have relatively low levels of employment because of artificial rigidities in the supply curve for labor then removing them will increase employment levels and increase real GDP.

This may cause some bankruptcy as people currently working may earn less but it will also avoid some too if currently unemployed people gets jobs and can start paying their bills again. Real GDP increases means that overall we are ahead. And that before we even start talking about all the good things that can happen in an unfettered economy...

Greg Hill writes:

"as a matter of pure arithmetic, lower wages imply higher profit income."

This is only "a matter of pure arithmetic" if you hold sales revenue constant. In a competitive economy where P=MC, a reduction in wages (MC) will be accompanied by a fall in prices. And if wages and prices both fall, higher profits need not result.

Keynes's point was that owners and workers bargain over the money wage, not the real wage. Labor is not in a position to reduce its real wage because labor has no control over prices.

Comments for this entry have been closed
Return to top