« Understanding another's theory | Main | I Dissent (FOMC Edition) »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Why do you make wage rigidity sound like an attribute of the employed or unemployed? Price rigidity is a system attribute -- it is a function of the preferences and the market structure of all the actors together.

Wouldn't the burden fall as much on the firm? How many firms, for example, given an option to their employees to take a 5% pay cut or to trim 5% of the staff?

Generally firms opt for the latter. They *prefer* the latter, because they assume that if they do this, then the employees that accept the cut will more likely be of lower quality then those who do not. Note that this would be true even if workers were flexible -- i.e. workers would be willing to take a cut *if* they couldn't find something better at another firm. There is still a first mover penalty for doing this, so in a de-centralized system, this is not likely to happen.

rsj: what you are talking about is one version of the "efficiency wage theory" of unemployment.

If we want to use that (type of) theory explain why employment *falls on the way down*, then it is the response of the *currently employed* to a wage cut that leads the firm to choose not to cut wages. (Morale and hence productivity would fall, the good ones would quit, etc.)

But if we want to explain why employment, having hit bottom, doesn't rise back up again, we have to look at the response of the currently unemployed, *if they became employed*, to various possible wages, and the firm's counter=response in choosing wages based on how the firm thinks they would respond.

Sure, in an efficiency wage story (or perhaps most stories), we have to look at firms as well as workers. You are right to note that. But this post is about *which* workers? The (currently) employed or unemployed?

While it is true that sufficiently low wage offers would drive the unemployed out of the labor force, eliminating unemployment, it would not raise employment back to what it was. Demand is so low, the cuts would have to be so severe that the unemployed are effectively ZMP workers. It is a mistake to talk of the demands of the employed or unemployed as it is really the demands of employer who is in charge and does what he believes is in his interest to maintain morale and not allow the dismoralized to disturb his operation.

Lord: If it is a decline in the *real* demand for labour (a leftward shift in the labour demand curve where the *real* wage is on the vertical axis), that might be correct. But if it is a decline in the nominal AD curve, with the price level on the vertical axis, and hence a decline in the nominal labour demand curve, with the nominal wage of the vertical axis, that would not be correct. This debate is (mostly) about the latter. If the nominal wage rigidity theory is correct, an equal decline in the nominal wage, and the nominal price level, would leave real wages, employment, and output, all unchanged.

Or using Tabarrok's example, the 90% employed are overpaid by 10%, leaving nothing for the 10% unemployed. There are too many overpaid employed for the lesser number of unemployed to obtain any salary and there is no wage at which they are productive, assuming no monetary change.

Lord: They have all (I think) forgotten Greg Mankiw's Principle #3. Rational firms think at the margin. Average wages and average product of labour do not matter. It is the marginal wage and marginal revenue product of labour that matters.

They fear adding an unhappy worker would lower productivity. I do agree with Karl Smith though that sticky prices are important.

If the worker was previously unemployed, he should be very happy.

He probably would be but no one wants to take the chance, especially when current workers can be worked harder and there isn't much marginal demand. If prices were flexible we would be seeing money pile up among employees, but they don't appear to be since it is piling up at employers.

I agree that it is the flexibility of wages of the unemployed that matters when NGDP goes up. My point was that if NGDP doesn't go up, then wage flexibility among the unemployed doesn't help very much.

Nick, why would an employer want to hire, if there's no demand for more production, and if there's an overhang of excess capacity to the tune is 10% of GDP?

Wage rigidity is only dominant in constraining employment when there's growth or contraction, not when there's stagnation (or growth well below the potential), excess capacity and underemployment.

Treating labor as just another commodity with its own supply and demand curve is the biggest flaw of conventional economics. People are not apples or widgets. One persons employee is your buddies customer. All this wage cutting is short sighted and the number of economists who have been paid handsome sums to come up with "models" to justify it is pathetic. Especially since the marginal productivity of academic economists is close to ZERO. If they all disappeared tomorrow no one would notice a damn thing.

And how can you say this Nick .... "If the nominal wage rigidity theory is correct, an equal decline in the nominal wage, and the nominal price level, would leave real wages, employment, and output, all unchanged." .... and not recognize that the same is true for an increase in wages and prices.

The problem with your scenario of falling wages and falling prices is that debt contracts are FIXED. What businesses have borrowed and what people have borrowed is related to what they are earning, which is of course related to labor prices and goods prices. The "only" downside to lowering prices and wages is more defaults on loans............ we saw how that worked out in '08 and '09.

No, no. Wage rigidity and unemployment is the observed empirics. The underlying theory of the rational firm is exactly that: the theory. So the answer to:

"If I increase employment by one worker, will I gain more in revenue than it costs me in wages?"

is in fact no, otherwise firms would hire more at the margin. Your task, as an economist, is to figure out why the answer is no, not to say that the answer should really be yes so let's toss rationality and/or the observed facts out of the window. There are fascinating models of wage setting for exactly this reason.

A point here: keep in mind that the marginal product of a given laborer is not his individual contribution; his marginal product is the margin from the entire employed force from adding the additional laborer. If all the marginal laborer does is to drink coffee whilst entertaining more productive workers, his own contribution may be negative but his marginal product positive.

A second point: if being hired at a vastly lower wage than expected should make the unemployed happy, I fear you are operating on an alien vision of the human psyche. Incidentally, why is this vision (that people are willing to accept vastly lower wages) plausible to you, whereas the flexible-wage-sticky-price NK narrative is not?

More carefully:

The unemployed are only recorded as involuntarily unemployed because they are searching for employment at the prevailing wage (of the employed). i.e., they are seeking to displace the currently employed, which would be consistent with Tabarrok's suggestion that the unemployed are simply unlucky. They could rationally expect someone else to eventually be unlucky instead of them.

If their only wage at which they could be employed is at a vastly, vastly lower level, then the unemployed may well prefer to leave the labor market. But you wouldn't know, since that would never show up on the statistics of jobseekers.

All of this seems a bit bizarre given that corporate earnings are at historically record highs.

The MRP of workers is already obscenely high, and yet it's still not profitable enough.

U.S. firms, in aggregate, have been net saving since 2003 -- e.g. accumulating claims on households rather than the other way around.

Earnings are through the roof, profits are through the roof, there has been an secular widening of the productivity-compensation gap, and yet people are *still* talking about low MRP workers and the need to cut wages.

The corporate profit share of GDP is already double the historical norm -- just how many standard deviations above normal does it need to get before we decide that wages do not need to be cut?

What sort of evidence would it take to convince economists to look out their window and conclude that the returns demanded of firms are far too high to achieve anywhere near full employment?

This is, fundamentally, an interest rate problem. But the rate demanded of firms is not the overnight interbank rate.

Firms are still promising rapid earnings growth and cutting labor when this growth does not materialize. Pension funds are still assuming 8% (nominal) CAGR as the industry norm, even though FedFunds is at zero and GDP is not growing anywhere near 8%.

There is some sort of disconnect or market failure going on here, but it is *not* due to sticky wages.

Scott: Suppose the central bank fixes NGDP, so the AD curve is a rectangular hyperbola. Assume P is perfectly flexible. Start at "full employment". Suppose the central bank reduces NGDP by half (the AD curve shifts left) and keeps NGDP down (it does not shift the AD curve right again).

If employment falls, it is because nominal wages of the employed are fixed. But now that employment and RGDP has fallen, why don't they immediately rise back up again? If the wages of the unemployed fell enough (if they halved) the firms would immediately hire them all back, so employment and RGDP would return to full employment. If this doesn't happen, it must be because the wages of the unemployed are sticky.

White Rabbit: "Nick, why would an employer want to hire, if there's no demand for more production, and if there's an overhang of excess capacity to the tune is 10% of GDP?"

You are implicitly assuming either sticky nominal prices, or an AD curve that is not downward-sloping. That's OK. But then you are violating the assumption of this post that the fall in employment in a recession is caused solely by wage rigidity.

Gizzard: you need to keep three different questions separate:

1. Why don't wages fall?
2. Would it help employment if they did?
3. Might there be some other bad side-effects, and there be a better way to help employment?

This post is about 2, and a very particular question within 2, under the assumption that wage rigidity is the only problem. Whose wage rigidity?

"And how can you say this Nick .... "If the nominal wage rigidity theory is correct, an equal decline in the nominal wage, and the nominal price level, would leave real wages, employment, and output, all unchanged." .... and not recognize that the same is true for an increase in wages and prices."

I do recognise it. ???

"The problem with your scenario of falling wages and falling prices is that debt contracts are FIXED."

Yes, funnily enough, I KNOW THAT.

Look, I know your nose is out of joint because I have had the temerity even to talk about wage rigidity. But when you are on a blog run by 4 academic economists, it is wise not to make those sort of remarks, especially when you show little understanding of what academic economists understand.

david: "is in fact no, otherwise firms would hire more at the margin. Your task, as an economist, is to figure out why the answer is no, not to say that the answer should really be yes so let's toss rationality and/or the observed facts out of the window."

Agreed. But part of doing that involves understanding accurately what the models would require in order to work. Would, for example, a model require wage stickiness of the employed, or unemployed, or both, to explain the facts? I'm arguing "both" (or, it would require something else).

"If their only wage at which they could be employed is at a vastly, vastly lower level, then the unemployed may well prefer to leave the labor market. But you wouldn't know, since that would never show up on the statistics of jobseekers."

Yep.

rsj: "All of this seems a bit bizarre given that corporate earnings are at historically record highs.

The MRP of workers is already obscenely high, and yet it's still not profitable enough."

You are confusing MRP with ARP. It's high ARP (relative to W) that causes high (average) profits. It's MRP ( relative to W) that determines whether a firm wants to increase or decrease N.

"What sort of evidence would it take to convince economists to look out their window and conclude that the returns demanded of firms are far too high to achieve anywhere near full employment?"

Again, there's a difference between average and marginal. If firms want to maximise total profits, they will hire iff MRP exceeds W, precisely because they want to get more "returns".

(Best avoid the word "returns" too, since it only leads to confusion.)

Nick, Bill Woolsey has a post explaining what's wrong with your logic. If they kept on the 50% of workers at their old wage level, and hired the other 50% back at a 50% wage cut, the total wage bill would rise by 50%. If during the recession the total wage bill was more than 66% of GDP, then a 50% rise in the total wage bill would put it above NGDP. Clearly an impossibility.

Scott: OK, but that is conceptually equivalent to the debt-deflation theory of recessions. It's an *infra-marginal* transfer payment from one group of people to another. Only it's not debtors to creditors in the normal sense; it's the firm's shareholders to its insider workers. Those insider workers are the firm's creditors. And deflation, given fixed nominal wages on those who keep their jobs (fixed nominal interest rates on past debts) changes the distribution of real wealth.

Heh, I'm not confusing MRP with ARP.

Firms have already recovered their earnings to pre-recession levels, but employment is now lower.

Look at the employment cost index/CPI:

//research.stlouisfed.org/fred2/graph/?id=ECIWAG


It's below pre-recession levels as well.

You cannot argue that nominal wage rigidity is why unemployment remains so high.

In terms of nominal wage rigidity of the unemployed -- the unemployed don't set their wage, they show to interviews and generally accept job offers. As long as there are more job seekers than vacancies, explanations of nominal wage rigidity of the unemployed are difficult.

Now it's theoretically possible that the shape of the revenue function has changed in a very funny way, so that adding one more worker today gives the same or less marginal revenue as adding another one before the recession began -- but I don't buy it.

How can firms be as profitable as they were before the recession began, with lower real employee costs, and 5% higher unemployment that appears to be "stable".

Now when you say "firm", you are thinking of a stylized entity with zero profits -- really you mean the *manager* of the firm. When I say "firm", I mean the *owners* of the firm -- the ones supplying capital to the manager. That is the source of confusion.

I'm thinking along these lines.

The owners of the firm tell the manager -- we want 3% dividend yields today that grow at a rate of 5% over 5 years. The manager then embeds that into his cost structure, and hires until he believes this condition can be met. He does not stop hiring when MRP = 0. He stops hiring when the return to capital over that 5 year time frame would dip below the 3% current + 5% growth target. If, by expanding production, he could earn more revenue today, say increasing the dividend yield by 1%, but the growth rate of this (larger) revenue stream would be only 2%, then the manager wont expand production.

Now suppose that the demand curve is such that the manager cannot meet the 5% growth target. He can compensate by delivering a 6% return with a 2% growth rate. The total cost of capital is the sum. The effect of that -- in a static analysis -- is that firms suddenly need to become twice as profitable as they were before. All sorts of workers now need to be laid off, but not because their nominal wage demands are too low -- but because the cost of capital is too high. You would get pro-cyclical markups.

I am implicitly assuming that these marginal adjustments are being made, and as firms have such enormous profits with growth rates in excess of the GDP growth rate, this can only mean that cost of capital imposed on the manager is very high.

Now I don't have a simple model as to *why* the cost of capital is so high -- but I can build a good empirical case that it *is* high. And my gut says that investors "learn" what rate is possible, and then demand that rate going forward. Moreover, if the central bank lowers the borrowing rate, and this causes the price of capital to be bid up, then perversely in my imperfect capital market "gut" model -- this can cause the cost of capital to increase, because investors begin to count on the one time capital gains as being part of their regular return going forward.

And if that is the case, then lowering nominal wages might not help. It might just serve to justify and re-enforce the excessive cost of capital.

Truman Bewley, anyone? That is, employers believe in an efficiency wage, not that there is anything that employees can do to protect employment. Even if employees agree to wage cuts, employers would not take them up on the offer because of the belief that they would be less efficient. I suspect that having 2 different wage structures, one for the currently employed, one for new hires, generally falls under the same heading, although we have seen it on occasion in unionized parts of heavy manufacturing, esp. when the union has negotiated concessions.

The practicality of 2 different wage structures likely depends on the organization of work and the firm, but my impression is that where we have seen it, it is less related to the business cycle and more to changes in the competitive situation facing the firm that requests or imposes it.

rsj: "Now it's theoretically possible that the shape of the revenue function has changed in a very funny way, so that adding one more worker today gives the same or less marginal revenue as adding another one before the recession began -- but I don't buy it."

I buy it. So do all sticky-price Keynesians and monetarists. Firms are sales-constrained. The MRP function is going along as normal, then falls off a cliff. The extra worker can still *produce* the same amount of extra output as before (the MP curve hasn't changed), but firms can't *sell* that extra output (or they find it very hard to) in an economy with deficient AD. That's what I was rabbiting on about in my post on ZMP workers and output quotas.

That's a big if! But I'll try to go with it ... I confess I have a hard time getting my head around the idea of nominal wage rigidities for people not receiving wages. Do you (or Cowen?) mean wages being offered are less than what people are willing to accept, and if only offered wages would increase then people would agree to do the job? I think there are lots of frictions that end-up looking like that: can't afford to move to where the work is (but if you paid me to move I would), house underwater so can't/won't sell and move, hope (unemployed GMU econ prof ain't gonna work as WalMart greeter), reservation wage (I have 20 years experience, so I want $100K a year), etc ... So I guess I'm agreeing with Nick. Under the given assumptions (which I don't think have much bearing on the current situation in the US), it is the unemployed's wage rigidity that matters.

Patrick: Hang in there with the big if!

"Do you (or Cowen?) mean wages being offered are less than what people are willing to accept, and if only offered wages would increase then people would agree to do the job?"

Yes, except I would add at the end of what you said "...or if there were a decrease in what people would be willing to accept."

"Firms are sales-constrained. The MRP function is going along as normal, then falls off a cliff. "

Yes, during the collapse, profits plunged and there were mass layoffs.

But now -- in terms of levels -- Real GDP is as it was before. Earnings per share are back to pre-recession levels. Employer Cost Index/CPI is at where it was at the peak. Prices have increased since then, so firms are able to sell more output than at the peak for a greater (nominal) price.

But there is still 5% less labor and nothing is happening in the labor market -- e.g. any changes to the unemployment rate, which are minuscule, are due to declines in participation.

So what makes 2011 different from 2008? The obvious thing to me, from the point of view of the firm, is that expectations of future earnings growth are now lower. Projects that have a low present return, but (in 2007) would have had an expected higher future return are now not funded, whereas they would have been funded before, so demand for present labor to work on those projects is now lower. Real Dividend yields are higher, and expectation of real dividend yield growth rates are lower. I interpret that as the cost of capital increasing for firms, even though FedFunds has decreased.

rsj: "Yes, during the collapse, profits plunged and there were mass layoffs."

That's not what I meant (my fault for not being more clear).

I am talking about moving along a given MRP curve from left to right, as well as a shift in that curve over time.

It used to look like this: (Oh, sod this keyboard)

Draw a slightly downward-sloping line. MRP on the vertical axis, and employment on the horizontal. Now something happens. The left half of the line continues to slope down, until about halfway along, then it goes vertical. That's what I meant by "The MRP function is going along as normal, then falls off a cliff."

What happened? AD fell. They can only sell half the output they could before. So that changed the MRP curve to make it cliff-shaped.

Speaking as an unemployed worker, what stops the re-employment is imputed wage rigidity. Labor markets differ from other kinds of S:D equilibria - see the Unemployment chapter in _Animal Spirits_.

I'll mention a point I've been making at various places during this discussion--there is a substantial literature in the job search literature and it demonstrates that reservation wages of unemployed workers are not, in fact, downward-sticky. The unemployed do reduce their reservation wages, and substantially, as the duration of unemployment lengthens.

Much of this discussion seems to me have proceeded as if a job applicant and an employer in fact bargain directly over compensation. While this may be true in some jobs (such as those that economics professors (that'd include me) have, in most cases the employer makes an offer and the applicant says yes or no--there is no explicit bargaining over wages. And, again from the job search literature, almost all offers are accepted.

So if there's a downward stickiness in the wages of the unemployed, it originates in the offers made by employers, not in the reservation wages of the unemployed. A recognition of that would change the terms of this discussion from "Why don't the unempployed wake up and offer to accept lower wages? That would take care of the problem" to "Why are wage offers by employers downward-sticky?" That's a question that the structuralists--Cowen among them--don't seem to recognize as the correct question, and don't seem interested in addressing.

"Why are wage offers by employers downward-sticky?"

Can we tell a story where the average cost of shirking goes-up in a recession, so even though wages might want to go down, the decrease is offset by having to pay larger 'bribes' to limit shirking?

Donald: good point. Again it suggests some sort of efficiency wage type theory. But that raises the question: if there's a certain efficiency wage in normal times, when unemployment is normal, why doesn't that efficiency wage drop in abnormal times, when unemployment is abnormally high? The Stiglitz Shapiro model says it should. The Spence(?) "hiring costs/turnover" model also says it should.

Patrick: logically correct. But: I can't think of any such story that is very plausible; it would need to be a very big effect indeed to offset the increased penalty from being fired at a time of high unemployment.

And even then, that gives you a theory of sticky *real* wages, and you need an extra step to get sticky nominal wages.

It's pretty clear that the problem is with the supply of jobs, not the demand for them. Companies that want to hire can, with few exceptions, find someone willing to do a job at whatever salary they feel like providing. But a company deciding to hire often goes through an extended decision-making process. First they decide that demand for their product is sufficient and stable enough that it makes sense "at the margin" to hire more workers. Then they have to figure out where in the company to hire more workers and create "space" to hire them (physical space as well as whatever equipment they need to make a worker productive). The company needs to figure out who will manage and train the new hire. Often, a budgeting process is included as well where departments of a company decide whether to accept the added cost of personnel. Only once companies go through this process do they hire people. Think about it this way--the stock market is probably going to go up at the moment, so why don't you invest in stocks? Probably because you need your money where is it, or are too busy to transfer money and manage investments at the moment.

Donald is right. In every job I have ever had, the employer simply made an offer and I accepted it. There was no wage bargaining. Maybe previous generations engaged in it; Jacques and others have provided evidence of a sort of job-abundance that is simply foreign to me. But that isn't today's reality.

Eric: Yup. Hence "The Big If" disclaimer.

:)

Grasping at straws here: Put aside hoarding for a moment.

Before the recession, the firm's workers have a range of MVP. Recession hits, so they dump the lower MVP workers. Now say the contraction is over (maybe no growth, but contraction is done). They are left with hoarded labour and high MVP labour - the people actually holding down the fort.

At this point, the firm then thinks about maybe hiring (maybe people are getting worn out, or more sales are anticipated but have not materialized). First it looks at hoarded labour. Probably it finds what it wants. Yay! Now they're getting something for the wage they were paying anyway. But what if the firm doesn't find what it wants in its stash? It looks at the MVP to figure out the wage it should pay - which is now higher than it was before the recession - considers it light of currently depressed sales, all the frictions we've mentioned already, and so they say "forget it".

This is totally off the cuff after a long day fixing broken industrial control software (no, I didn't break it), so my brain might be failing me ...

Nick--The job search literature looks mostly (entirely, if I recall correctly) at nominal wages. So that body of research is saying that nominal reservation wages are flexible downward, but nominal offer wages are not. With (eventually) flexible prices in recessions, this could explain sticky downward real wages. I'll admit that I don't remember much in the way of efforts to provide explanations of that (but it's been a while since I was reading a lot in the search literature). But I think an explanation of this has to look at models of search (both applicant and employer search) and be grounded in an understanding of what happens in that process.

Nick, I see your point, but then the more I think about it the more the model seems incomplete. Taken literally, the firm would continually fire insiders, and as soon as they were fired hire them right back as outsiders. Obviously that would mean, de facto, that there is no nominal wage rigidity at all. I'll have to think about this a bit more.

"What happened? AD fell. They can only sell half the output they could before. So that changed the MRP curve to make it cliff-shaped."

OK -- that is a static analysis during the decline, but the numbers don't add up for the expansion.

Since the trough of the recession:

1. Prices have gone up: PCE deflator has increased by 4.5%
2. markups have gone up: unit profits have increased by 63% (nominal), and 55% real
3. Real GDP has increased by 8.3%
4. Real labor costs have fallen: Real Employer costs per hour worked has declined by 2% (taken from Employer Cost Index, total compensation for private industry)
5. Hours worked have only increased by 1%
6. Non-financial Business sector equipment and software stocks have *decreased* by 4.3%

Perhaps I am missing an obvious explanation -- but it seems to me that the only logical explanation for the capital stock declining during an expansion is an increasing real cost of capital. Firms have been investing in equipment and software, but not enough to replace its consumption. Effectively this means that they have been returning capital to their shareholders.

And this happening over a 2.5 year time frame during which prices have not been constant.

Where this extra output came from is a bit of puzzle -- perhaps there are increasing returns, or the output consisted of Apple ordering more units from China, so we didn't need more labor to produce the extra output -- it was enough that Apple "designed" the iPhones, and the value-add, or difference between paying the Chinese $5 to make the phone, and selling the phone for $600 was imputed as additional domestic output that didn't require a lot of domestic labor input.


But it seems to me that there is something more going on here than just a demand failure.

If it was just a demand failure, then firms would not have sky-rocketing per unit profits, and they would not be able to produce and sell 8% more output at 4% higher prices using only 1% more labor at a 2% lower real wage.

"Where this extra output came from is a bit of puzzle"

How about: the workers who kept their jobs are working harder than before. Workers aren't machines; they have some control over their productivity.

Patrick: "It looks at the MVP to figure out the wage it should pay - which is now higher than it was before the recession - considers it light of currently depressed sales, all the frictions we've mentioned already, and so they say "forget it"."

1. Why should the wage it needs to pay to hire a new worker be *higher* than it was before the recession?

2. When you say "currently depressed sales" you presumably mean "currently depressed demand for its output". But that implies the problem is with the output market, rather than simply sticky wages in the labour market. (Remember the big IF).

Donald: one line in the search literature that has some promise to my eyes is the one that Roger Farmer is following. In a search model, there's typically an indeterminacy of the wage when the job-seeker actually meets the potential employer. Due to bilateral monopoly/monopsony. There's a gap between the two "threat points" in the Nash bargaining game. If we reject the Nash bargaining solution, the wage can be anywhere between those two threat points. Anything could act as a focal point. It might be past history of the nominal wage.

Scott: Yes, I think it might be worth exploring the dangers of having insiders and outsiders paid different amounts.

rsj: "Perhaps I am missing an obvious explanation -- but it seems to me that the only logical explanation for the capital stock declining during an expansion is an increasing real cost of capital."

*Part* of the explanation would be that you are looking at the expansion in GDP *since the trough*. But if the capital stock reflects the desired capital stock *at the previous peak* of GDP, there is less of a puzzle.

But if you re-did your numbers from previous peak, I think you would still have some puzzle left. Maybe fear that the recession will get worse?

I think you are maybe onto something possibly important, rsj.

Max *might* have part of it.

Nick - I don't know if makes any sense. But the idea is that the firms workers are very productive, so they command high wages. That sets a sort of standard. To make it concrete, say the firm is hiring engineers. At the end of the recession only the really good ones are left. They are all paid lots of money. The firm decides to evaluate if it should hire another engineer. They say "Yes, if we can pay her $X". Then they look at their existing staff. They all earn much more than $X. In my experience, a firm would be extremely reluctant to pay two employees doing the same job very different wages. In some cases, it might even be illegal. So they decide not to hire anyone.

I’ve been banging on about the basic point in Nick Rowe’s article above for decades. My basic point is that it is possible to reduce unemployment by means of an employment subsidy that cuts the marginal cost of labour. See:

http://mpra.ub.uni-muenchen.de/19094/

A slight problem involved in trying to cut the marginal cost of labour is distinguishing between employees who are genuinely marginal and those who are not. I.e. how does one prevent employers claiming the subsidy in respect of employees who are not marginal? The answer is to limit the time for which allegedly marginal employees stay with a given employer. Employers do mind losing valuable or key employees. But they don’t mind losing employees who are marginal or relatively unproductive.

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad