The Critical Flaw in Keynes’ System
Just when the Austrians were getting ready to purge me for my “Keynesian” thoughts on interest theory, I confuse everyone by writing this. An excerpt:
So even if we mechanically assumed that a fall in labor’s money-wages would translate into a proportional fall in retail prices, labor would nonetheless have the power to reduce its real wages. For example, if laborers received a cut of 10 percent in their money-wages, then the prices of the goods they produce would only fall between 5 and 6 percent. Labor would be cheaper, even in real terms, and employers would move out along their demand curve and hire more workers.
But there are other problems with Keynes’s analysis. Consider: What is the actual mechanism through which falling costs lead to falling retail prices? We start in an initial equilibrium, where workers earn (say) $10 per hour, and the retail good sells for $100. Firms are happy with the number of workers they have employed at $10 per hour, and the amount of goods they can sell at $100 each.
Now, because unemployment is very high, the firms can get away with cutting their workers’ pay to $9 per hour. Holding everything else constant, they are making more profit than before. What would induce them to lower their retail price from $100?
The obvious answer is that they want to capture a larger share of the market. That is, they want to sell more units of the retail good to their customers. They can’t do this with their original labor force. No, in order to make it profitable to cut their retail price, they need to hire more workers and boost output. Then, in order to move the larger quantity of product, they cut prices from $100 to (say) $98. Even though they make less revenue per unit, they nonetheless make more total profit.
Other firms do the same thing, of course, until the new equilibrium settles down with wage rates at $9 per hour and the retail price at (say) $95 per unit. Thus a large part of the workers’ wage cuts have been passed along to the consumers in the form of lower retail prices. Nonetheless, in the new equilibrium, each firm is producing more units, and thus is carrying a larger workforce than before the wage reduction.
Let’s say that there is perfect wage flexibility. I don’t think this necessarily condemns Keynes’ case for socialized investment and/or monetary inflation. That wages flexibly adjust towards a new equilibrium doesn’t necessarily mean that the economy is no longer preforming somewhere inside the PPF. Wages have adjusted to a certain degree of productivity, but that productivity can still be improved through the investment of non-labor idle resources.
Neither does the argument of total price flexibility change anything, because then you have to disprove Keynes that the price of labor and the price of other goods will not fall proportionally. I’m not saying Keynes was right, I’m just saying that his argument is a bit more complicated and goes beyond wage inflexibility (even if wage rigidity plays a major role in his “general theory”).
Jonathan, I’m baffled. Not only am I trying to do exactly what you are suggesting here, but I am trying to do it in the excerpt I provided. Maybe I’m wrong, but I’m attempting exactly what you say I would need to do, to make my case.
That wages flexibly adjust towards a new equilibrium doesn’t necessarily mean that the economy is no longer preforming somewhere inside the PPF. Wages have adjusted to a certain degree of productivity, but that productivity can still be improved through the investment of non-labor idle resources.
This is a separate issue from the argument of Keynes that Murphy considered. Keynes’ argument that the free market is unable to soak up unemployed workers at lower wage rates is separate from Keynes’ argument that socialization of investment and inflation are justified to get the economy to the PPF in real terms, irrespective of the level of employment.
…because then you have to disprove Keynes that the price of labor and the price of other goods will not fall proportionally.
I can say that this is exactly what I interpreted Murphy to have done. You know, the whole 10% fall in wages but only a 5-6% fall in prices thing?
MF,
Right, but Murphy is talking about Keynes’ “fatal flaw”, not just about his views on unemployment.
But, I see that Murphy really does discuss what I said, and that next time I really should read the whole thing before commenting.
But Keynes idea is that expectations about future consumer spending can be very damaging to the economy, and it can enter a state of “unemployment equilibrium” (that’s a very weird way to say that its equilibrating forces are not sufficiently strong). If aggregate demand is slow and entrepreneurs don’t think that condition will improve, unemployment can perdure for a long time, notwithstanding flexible wages. That condition of low aggregate demand can create lots of bankrupcies and further enhance its destructive effects (we know, in the long run it may be fixed by market forces, but in that process a lot of capital was wasted innecessarily).
If aggregate demand is slow and entrepreneurs don’t think that condition will improve, unemployment can perdure for a long time, notwithstanding flexible wages.
Demand for labor and labor’s demand for goods is a part of aggregate demand. To artificially divorce this demand from “aggregate demand” makes it easy to deny what makes it possible in the first place, then imagine aggregate demand not rising!
It would be like saying the economy won’t improve unless the economy improves first.
As I document in my book on the Great Depression, average money wages fell very sharply in the depression of 1920–1921, in contrast to the much more modest decline during the early years of the Great Depression. Perhaps the change was due to enhanced union power, or to Herbert Hoover’s pleas with business to maintain wages.
Bob, do you really believe that we had 25% unemployment in 1932 because Herbert Hoover *asked* businesses not to cut wages? (I realize you also mention “enhanced union power” as a factor, but unionization in 1928 was less than 10%, lower than in 1914, so that is not a plausible explanation either).
Suppose that during a drought a high government official takes to the airwaves to plead with people to conserve water. The typical libertarian response would be to snicker at the idea that this sort of attempt at moral suasion would be at all effective. People aren’t going to significantly cut back their water usage just because a major or governor or whatever asked them too. Indeed, they might increase their water usage, as they expect the pleas may be a prelude to direct controls, and want to take as much advantage of the open system as they can while it lasts.
Yet we are supposed to believe that Herbert Hoover was so persuasive that businesses were willing to go bankrupt rather than lower wages.
re: “do you really believe that we had 25% unemployment in 1932 because Herbert Hoover *asked* businesses not to cut wages?”
I have a comment under review at the QJAE on just this topic.
Hoover *asked* businesses not to cut wages
When an 800-lb gorilla expresses what it wishes of you, it is never “asking”.
Silas,
Suppose Obama went on television tomorrow and pleaded with businesses to hire more people. Do you think this would cause the unemployment rate to fall precipitately, as the 800 pound gorilla always gets what it wants?
No, they would either:
a) shuffle into doing economically inefficient things that would (at least superficially) satisfy whatever constitutes as PC excuse for not hiring more people, or
b) if pleading for special favors (e.g. exemption from new health care mandates) or worried about politically motivated demonization of their business, hire some token useless workers as a loss leader (which would promptly be hailed as good news on the job front).
You can’t fight economics, but you can sure do economically stupid things to take the political heat off yourself.
Have you ever run a high-visibility business?
No
Good. So do you see why Herbert Hoover asking firms not to cut wages is not an explanation for 25% unemployment?
Do you see why the bully pulpit and threats of future “investigations” _are_ good explanations for why businesses do economically inefficient, recession-prolonging things? Or did you just skip right past the substance of my reply?
There more and less inefficient ways to respond to the demands of the 800-lb gorilla, and they differ between the cases.
Do you see why the bully pulpit and threats of future “investigations” _are_ good explanations for why businesses do economically inefficient, recession-prolonging things?
Did Hoover threaten anyone with “investigations”? I must have missed that one (as far as I know, those businesses that did lower wages despite Hoover’s pleas had nothing bad happen to them).
Up to a point, a business will do economically inefficient things to avoid government scrutiny (real or threatened). The question, though, is whether the particular government act (Hoover’s attempt at moral suasion) can explain the magnitude of the effect (25% unemployment). When Gerald Ford urged businesses to whip inflation by not raising their prices, businesses did not respond by doing lots of economically inefficient things to avoid raising prices. They just ignored the 800 pound gorilla.
Maybe Hoover was such a menacing figure, and his pleas struck such fear into the hearts of businessmen, that they were willing to go bust en mass rather than defy him. Somehow I doubt it.
Did Hoover threaten anyone with “investigations”?
Wow, you have a _really_ naive view of what it means when an 800-lb gorilla suggests you do something.
I suppose Chuck Schumer didn’t “threaten any investigations” when he called up Amazon and politely “asked” them if they could stop hosting Wikileaks, Purty Pleez.
And FWIW, businesses were intimidated into action by Ford, it’s just that the equilibrium policy wasn’t as harmful: they shrank product sizings whilst keeping the “same price”. (Hm, sounds familiar…) Incredible shrinking chocolate bars!
What about the “tobacco settlement” of ’97? The companies just accepted the deal because it was such a great offer, not because the government could make their businesses suffer.
And health insurance companies “supported” Obama’s health reform because they ~loved~ his ideas, not because of what would happen if they got on his bad side, right?
I guess any amount of naivete is possible if you want to bolster your position.
Blackadder,
This isn’t a case of Hoover just going on the radio and making a general request. (My wording in the article was poor in this respect.) He had a conference with major leaders coming to Washington, and they set up “partnerships” and could keep tabs on everybody.
I have come across a theory as to what leverage the Hoover Administration actually had over businesses, and how it could play nice if they kept up their wages. (Unfortunately I don’t remember where I read it. I think it had to do with taking management’s side in labor disputes, but I could be wrong.)
Anyway here is another example of the kind of thing Silas is talking about.
In any event, the crucial point is that nominal wages fell much more sharply in the 1920-1921 depression than in the Great one (per year). So whatever story you come up–employee morale for example–make sure you explain why it changed significantly in a decade.
Bob,
Clearly you have very different outcomes from the 1921 and 1929 recessions, and asking what the difference was that caused the different outcomes is an important question. However, saying that the difference is explain by the fact that Hoover held a conference with business leaders and urged them to not cut wages doesn’t hold water. In any other context libertarians would find the idea that government jawboning could have such a huge effect laughable. The only reason libertarians take the idea seriously w/r/t Hoover, I submit, is that they need some explanation for how government prevented recovery from 1929-32, and this is the best they can come up with.
Are we operating with the understanding that a “clearing labor market” is the same as “a labor market without any unemployment”. Because I’m not sure these are the same things. Certainly what we measure and label “unemployment” – people wanting and looking for a job but that don’t have one – can exist in spades with a labor market that clears.
So if it’s THAT phenomenon that you want to explain, it’s not clear that reduced wages that clear the labor market provides any kind of answer.
This is one more reason why microfoundational preoccupations can be dangerous… it leads people to take vocabulary and terminology from microeconomics that seems to fit what we’re trying to explain in the macroeconomy, but which actually don’t.
I’m still not sure what I think of all this… still working through all these ideas…
I clearly understood most of what Bob wrote in his article. I can’t understand what the heck Daniel is talking about. I am thinking there is something that my pea sized brain is unable to take in, but what is it?
It seems like you are saying:
Employers, existing and potential, would not necessarily offer lower wage rates even if there are currently unemployed workers able and willing to work for wage rates that are lower than prevailing wage rates.
The question then becomes, why would employers not offer lower wage rates even if there unemployed workers able and willing to work for those lower wage rates?
Keynes’ belief was that it was because either worker’s marginal propensity to consume was less than unity, or because the economy was unable to accommodate more any more profitable net investment. In other words, it is because workers hoard cash rather than consume, or because the market rates of profit are so low that investors hoard cash rather than invest.
If THAT is the reason for why the free market is unable to solve high unemployment, then it’s definitely clear that government deficits are not necessary, because both explanations are deeply flawed.
DK I was using “clearing the labor market” to mean “no involuntary unemployment.”
Why would workers take a wage cut, both in nominal and real terms, especially when they have nominal debt obligations?
because it’s better than the alternative (unemployment or part-time work)
Rule #1 in economics – always think about what the alternative situation is. Obviously no worker would accept a wage cut if he could protest it without consequence.
Cuz it’s better than the alternative (unemployment or underemployment)
Rule #1 in economics – always know what the alternative situation is. Obviously no worker would accept a pay cut if they knew they could protest it without any consequences.
Largely for the same reason people often rent out a house for less than the cost of the mortgage.
Reality is not obligated to help you fulfill your obscenely-optimistic promises.
Thanks for the non answer. I guess the lack of answer is the answer.
What non answer? Look, I’ll put it differently: workers will take a wage cut because their nominal debt obligations don’t somehow make them more productive. They don’t have an altnerative but to take prevailing wages. Very rarely do people pay more for a product than its market price because, “come on, I’ve got a lot of credit card debt / mouths to feed / a big mortgage!”
(And anyone who pressures people this way for a market transaction is vile.)
If the above were not the case, bankruptcy court would be a LOT easier: “Hey, don’t worry, Mr. Doe, you’ll be able to earn a lot more now that you have big debts!”
Did you even know why people rent out a house for less than the cost of their mortgage on it?
Look, I’ll put it differently: workers will take a wage cut because their nominal debt obligations don’t somehow make them more productive.
What you are overlooking is that nominal wage cuts can make workers less productive (by reducing morale, increasing resentment against the boss, etc.) If an employer is forced to choose between cutting wages for all workers 10% or laying off 10% of workers, it might be individually rational for the employer to lay 10% off and have 90% grateful to still have a job than cut pay by 10% and have all his employees angry with him.
Yes, I’m familiar with that effect. If you kindly follow back to the beginning of this thread, you’ll see I was only replying to Mammoth’s question about why workers’ nominal debt obligations wouldn’t somehow make them less likely to insist on above market wages.
I don’t think the sociology of the layoff/wage-cut tradeoff is responsive to that question.
I never mentioned above market wages. That’s your own biased interpretation.
Fine. Remove that from my statement. Your question is still confused: the fact that I have obligation does not make more able to demand more for my labor. The fact that I have a big mortgage does not mean I can charge more rent for the house I rent out. Etc etc etc. Get a clue.
You are still wrong. I am not asking why they should ask more, but rather why would they accept less. A question you didn’t answer.
Yes, I did. A hundred times. Why does a house owner “accept” less than the cost of his mortgage as rent when he rents it out?
Because that’s all the market will bear!
Yes, I did. A hundred times. Why does a house owner “accept” less than the cost of his mortgage as rent when he rents it out?
Because that’s all the market will bear!
I agree with your arguments, but they seem to be too short term oriented. Is it sustainable for rental prices to be perpetually lower than the costs of the corresponding mortgage?
I think if rental prices continued to be lower that mortgage costs, then eventually you will see either rental prices going up, or home mortgage costs going down, or else you will see a disappearance of rental properties that carry a mortgage.
Right — I didn’t mean to imply that something like that can last forever. Usually, it’s temporary, until the owner can sell the house or otherwise get rid of the mortgage. But the point is, it’s a rookie mistake (one financial talk shows warn midlife folks all the time about) to think that you somehow ought to hold out for rent = mortgage, “cause that’s only fair!”
Yes, I did. A hundred times. Why does a house owner “accept” less than the cost of his mortgage as rent when he rents it out?
No. That explains hundred times why they don’t get more as rent, but not a single time why would they accept less than at the present.
You are 100% right MamMoTh, there is no reason for any worker to accept lower than market wages.
However, in a depression market wages are not all that good.
What you are overlooking is that nominal wage cuts can make workers less productive (by reducing morale, increasing resentment against the boss, etc.)
What this overlooks is that layoffs on the basis of too high a wage rate can make workers even less productive than that.
If an employer is forced to choose between cutting wages for all workers 10% or laying off 10% of workers, it might be individually rational for the employer to lay 10% off and have 90% grateful to still have a job than cut pay by 10% and have all his employees angry with him.
Why do wage earners get pissed off with the people who pay them less, but employers don’t seem to get pissed off by their customers paying them less?
In other words, why aren’t producers of electronics treating their customers like crap on the basis of consumers paying less for electronics over time? Why is the antagonism and strife only from worker to employer? Sounds like a doctrine that seems plausible only on the basis of already widespread Marxist sentiments.
If a worker is given the option of being laid off or making 10% less, then the worker is far better off than being fired outright because the employer fears the employee will go postal. If a worker is given the option of less pay or being laid off, then the worker themselves can choose to stay or leave for another company.
It is always better to be given options.
I mean really, if you worked for me, and I gave you the choice between 10% pay cut or else you can look elsewhere for employment, or just being outright fired because I won’t even offer you a 10% cut, then wouldn’t you be MORE pissed off for not even being given the option of a 10% pay cut?
It would be like a seller of electronics totally removing their electronics from the market rather than selling fewer electronics for 10% less, and then they tell you “We didn’t want to piss you off by selling you fewer electronics.”
What this overlooks is that layoffs on the basis of too high a wage rate can make workers even less productive than that.
If you mean that the laid off workers won’t be productive, that’s true, but since the employer no longer has to pay them, that’s not really an issue. If you mean that laying off some employees makes the rest of the employees less productive, then no, that’s not what generally happens.
Why do wage earners get pissed off with the people who pay them less, but employers don’t seem to get pissed off by their customers paying them less?
That’s an interesting question. Regardless of why people get pissed off, however, it is a fact that they do tend to do so.
If you mean that the laid off workers won’t be productive, that’s true, but since the employer no longer has to pay them, that’s not really an issue.
Well when the problem is how to increase employment, not increase employment of happy workers, then it’s kind of an issue, yes?
Well when the problem is how to increase employment, not increase employment of happy workers, then it’s kind of an issue, yes?
Yes, exactly.
In Aus they we a habit of offering voluntary redundancy which means that all the people who know they are completely useless make absolutely sure they stay no matter what, and all the people who know they can easily get another job take the redundancy package, jump ship and pop up next week working for a compeditor.
However, at the emotional level I can see that when faced with a real pay cut caused by inflation, workers find it hard to blame their immediate boss so psychologically there’s a diversion in the blame chain.
@Tel: Yes, the American carmakers tried that in ’08, since union workers were entitled to continued employment as long as they met the bare minimum requirements. It backfired the same way: anyone who was any good took the money and ran, and they were left with all the deadwood.
Not a good idea, you would think.
Seriously, it really P/O’s me that half the debate is essentially, “Wah!!! You have to fundamentally restructure the economy because I have a right to get enough money to pay of a stupid debt I took out! Gimme gimme gimme!”
Seriously, it really P/O’s me that half the debate is essentially, “Wah!!! You have to fundamentally restructure the economy because I have a right to get enough money to pay of a stupid debt I took out! Gimme gimme gimme!”
You’re in luck! There is no need to be angry, as this is not Keynes’ argument (if you don’t believe me you can ask Bob Murphy, Keynes’ argument may be empirically and/or theoretically flawed, but it is not a gimme gimme argument any more than Zimbabwean creditors objecting to having the real value of their loans inflated away is gimme gimme).
Believe it or not, Keynes is not participating in the debate I’m referring to, nor does he comprise the totality of people who engage in policy debate on this matter — the set I was referring to.
I’m sure Keynes has a more nuanced version of this attachment to the inalienable right to free repayment of stupid loans, but he wasn’t the target of my criticism.
It’s better than losing job.
And cutting profits is better than losing your company.
I agree.
So it’s only a matter of who is going to cut his earnings first.
Why would workers be the first?
Who says its the workers? Companies cut headcount because it already has negative earnings or its earnings are falling rapidly.
The option is not profit cut or wage cut.
Paying workers a lower wage rate, in the depths of a depression, will almost certainly accompany a rise in total wages paid to workers, since investments were postponed up until that time, as well as lower relative prices of goods, for the reasons Murphy laid out above. This increase in real wages will accompany a rise in aggregate profits. Both workers and employers can rebound together. It does not require one to be sacrificed for the other.
Mammoth, if I have to service a debt, I can do so more easily if I have a job than if I am unemployed. So that’s why an unemployed worker should be willing to go back to work, even for a paycheck less than what he was accustomed to.
And then since there are millions of unemployed workers willing to do that (presumably), then the workers with jobs should have to do the same, lest they will be replaced.
It’s kind of funny, usually it is the interventionists who say the working man has no power when it comes to wages. And yet on this issue, workers can tell employers how much they are going to get paid, even if unemployment is 25%.
It’s kind of funny, usually it is the interventionists who say the working man has no power when it comes to wages. And yet on this issue, workers can tell employers how much they are going to get paid, even if unemployment is 25%
This is a great point.
Unless unions happen to control that particular industry, in which case there is a monopoly seller in the labor market, and under those particular circumstances it could very easily be more value for the monopolist to sell less at a higher price than selling more at a lower price.
There’s a simple way to test whether wages are in fact downwardly rigid. If Keynes is right, then there ought to be a discontinuity in nominal wage changes around zero, i.e. lots of people getting very small increases, very few getting small decreases. If the classical picture is right, then there should be no such discontinuity.
Here is a graphical representation of data on wage growth from the mid 1990s. It is exactly as Keynes would have predicted.
Now, as Bob pointed out, it could be that wages are downwardly rigid not because of some inherent flaw in the market or in human rationality, but because of government interventions (e.g. the minimum wage, unemployment insurance, unions, etc.) In that case, though, Keynes’ conclusions would still hold so long as those interventions were in place. If you believe this, then presumably you would want to devote your energies to things like repealing the minimum wage and so forth, which would render the whole question of Keynesian policy moot. What won’t do, however, is to simply deny that wages are downwardly rigid in the current environment.
Yeah, good point, we should ignore all other forms of remuneration and spending on employees when investigating whether employers can reduce payment for labor in response to its diminishing value.
Keynes actually identified unions as a major reason for wage rigidity.
Except he didn’t identify the source of the power to prevent non-union workers from competing. The classical economists which Keynes was criticizing used a context of free market competition, which was not exactly present at the time Keynes wrote the GT.
There’s a simple way to test whether wages are in fact downwardly rigid. If Keynes is right, then there ought to be a discontinuity in nominal wage changes around zero, i.e. lots of people getting very small increases, very few getting small decreases. If the classical picture is right, then there should be no such discontinuity.
Here is a graphical representation of data on wage growth from the mid 1990s. It is exactly as Keynes would have predicted.
Classical economics does not predict that wage decreases should take place with the same frequency and intensity as wage increases, such that you see a normal shaped curve centered at 0% change.
Which classical economist did you read that suggested wage rate decreases should occur with the same frequency and intensity as wage increases?
To see a discontinuity centered at 0% change is also consistent with flexible wage rates, if the change in wage rates tends towards general increases rather than general decreases due to inflation of the money supply and hence in the demand for labor.
I bet that if the supply of money stayed constant, then we’d all be here listening to interventionists claiming that wage rates are rigid upwards because the same graph shows a general tendency of wage rates to decrease if they change at all.
The graph does not vindicate the Keynesian worldview, BA. It only shows historical occurrences. It doesn’t confirm nor falsify economic laws or principles.
To see a discontinuity centered at 0% change is also consistent with flexible wage rates, if the change in wage rates tends towards general increases rather than general decreases due to inflation of the money supply and hence in the demand for labor.
Did you look at the chart? A tendency towards nominal wage growth (whether due to increases in productivity, inflation, or both) can explain why you would have more wage increases than decreases, but it doesn’t explain why you have a discontinuity at zero.
Did you look at the chart?
No, silly. I knee jerked and made an off the hip type comment that does not take into account any information you have provided, because my intention to obfuscate, not elucidate.
A tendency towards nominal wage growth (whether due to increases in productivity, inflation, or both) can explain why you would have more wage increases than decreases, but it doesn’t explain why you have a discontinuity at zero.
Why not? Why can’t I argue that the implications of inflation tending to be gradual in one direction only, leads to changes in wage rates concentrated in one direction only?
Why should wage rates changes be equally distributed around 0% change, when inflation is not 0%, but continuously positive?
If inflation of the money supply were 0%, then I might take this graph to mean what you think it means, because then it would make sense to expect equally distributed instances of wage rate decreases and wage rate increases. But because inflation is typically positive and gradual in one direction, we should expect, or at least I would expect, to see a high concentration of instances of gradual wage rate increases, and relatively few instances of any wage rate decreases.
I guess what I am asking is: Why SHOULDN’T there be a “discontinuity” at 0%? Aren’t you just imagining that an “efficient” market should lead to prices changes that look all nice and symmetrical around 0%? I find it highly prejudicial to insist that an efficient market has to resemble nice smooth normal curves, as if humans think and behave according to mathematical interpretations of non-thinking phenomena like dice throws.
Why should wage rates changes be equally distributed around 0% change, when inflation is not 0%, but continuously positive?
Again, the problem is not that wage changes aren’t equally distributed around zero. The problem is that wage rate changes get more frequent as you approach zero (despite the fact that inflation is positive), and then fall off a cliff once you get there.
Again, the problem is not that wage changes aren’t equally distributed around zero. The problem is that wage rate changes get more frequent as you approach zero (despite the fact that inflation is positive), and then fall off a cliff once you get there.
Again, that is you saying that wage rate changes should be more equally distributed around zero.
Of course one would expect that in an inflationary economy, wage rates will overwhelmingly change in the positive direction, and change in an amount consistent with the amount of inflation, which means many small jumps, and fewer and fewer larger jumps.
Why should wage rates do anything other than what you see, given what you know about what happens to money and prices?
More than that, in an environment where inflation is normal, all parties come to expect that sooner or later the dollar will devalue and they build that fact into their negotiation strategy.