Arnold Kling  

Lectures on Macroeconomics, No. 4

PRINT
Notes From the Bailout... Mr. Caplan Goes to Singapore...

This lecture covers an important issue: why do firms adjust by cutting workers rather than by cutting wages?

So far, I have been pushing an explanation for unemployment that relies on issues with adjustment. There is heterogeneity in the labor market. Demand rises for some occupations and falls for others. Some industries expand, while others contract. Within industries, some firms are successful, while others fail. The overall effect is that individual workers frequently change jobs; moreover, the occupational mix changes as some workers exit the labor force (when they retire, for example) and other workers enter the labor force.

The picture I have of a macroeconomic recession is one in which the need for these adjustments becomes overwhelming. The gains and losses in employment, which usually offset one another, get out of balance. Declines in certain sectors are too large and/or too rapid, so that aggregate employment declines. Moreover, this decline in employment has multiplier effects, as those who are unemployed cut spending.

A question is: why do worker cutbacks take the form of job losses, rather than wage cuts? There are many possible answers.

My first thought is that most of the time wage cuts are inappropriate. The overall trend is for productivity and living standards to rise. On average, salaries are increasing. One can see, particularly in longitudinal studies, a strong tendency for people to move on a rising escalator of income. (Longitudinal studies examine the same people over time. In contrast, many people wrongly conclude that incomes are stagnating when they examine snapshots of the distribution of income at different points in time, using Census data for example. One might find that the bottom 10 percent of the income distribution in 2000 is not far from the bottom 10 percent in 1980. However, many people at the bottom in 1980 moved up, and they were replaced by new workers and immigrants.)

For the most part, the market tells workers that they can expect to maintain or improve their standard of living. Doing so may require a willingness to adapt by changing firms, changing cities, or changing occupations. Overall, however, it is not normal to have to accept a permanent wage cut.

If the long-run trend of wages is upward, then as a manager you know that cutting wages at your firm means cutting them relative to other potential opportunities for your work force. If wages are not falling generally, then an absolute wage cut at your firm means that you are reducing wages relative to the market. Your workers are likely to bleed away, and the workers you lose first are likely to be your best workers. Better to choose which workers to lose and to lay them off.

In other words, keeping the same work force and cutting wages is typically not an option. The choice is between cutting the work force directly or having your work force decline in response to a wage cut. You probably are better off making direct cuts. Most of the time, cutting wages is a bad policy.

In addition, economists have considered a number of sociological reasons for maintaining wages. For example, wage cuts may demoralize workers and therefore harm productivity.

In any case, I find it plausible that under most circumstances wage cuts are not a good way for a firm to cope with adversity. Still, from a macroeconomic perspective, there are rare circumstances in which wage cuts are called for.

For example, if there is general deflation, then cuts in nominal wages are needed in order to keep wages from rising relative to prices and productivity. This may have been an issue during the Great Depression. One would expect that if real wages were too high, employment would fall and productivity would rise as firms attempt to economize on expensive labor. This seems to be what happened.

In addition, if there is a major sectoral decline that is too big for the rest of the economy to absorb, it might be better for wages to fall for a while in that sector in order to help maintain employment. In the long run, a decline in demand for cars relative to physical therapists should lead to fewer auto assemblers and more physical therapists. In the short run, an unusually large decline in auto demand (due, say, to permanently higher oil prices) might best be met with lower wages in the auto industry, which could bring down prices and enable firms to sell more cars than they would otherwise.

Even if the macroeconomic rationales for wage cuts may on occasion be legitimate, such cuts may be very difficult to implement. How is a worker to tell the difference between an appropriate, macro-determined wage cut and a raw deal?

In fact, the inability of workers to see through the reasons for wage cuts is at the heart of why workers are paid fixed-dollar salaries to begin with. Economists have long noted that there would be fewer employment fluctuations if workers instead were paid a share of corporate revenues or profits. In theory, if worker were paid in the form of the output of firms, there would never be layoffs!

As a worker, shifting some of my income from fixed-dollar wages to a share of corporate profits is adverse for two reasons. First, it makes my income riskier and dependent on things beyond my control. I may do my job perfectly well, but the company makes strategic errors, leading to no profits and little or no pay for me. Second, it imposes an information cost on me. I have to be able to audit the corporation's accounting statements to make sure that the reported profits on which they base my pay are not misleading--perhaps the real profits are being disguised and held in a form that benefits shareholders but not me.

So here is my story for explaining why we observe swings in unemployment rather than short-term wage adjustments. Under normal circumstances, most labor market transitions are toward jobs with higher pay and/or more desirable non-wage characteristics. Millions of these transitions take place each month. Therefore, a wage cut is a destructive, unsustainable policy that will alienate the firm's work force and cause its best employees to leave.

There are occasions where, from a macroeconomic perspective, a cut in wages, in either a large sector or across the board, would help to avert a surge in unemployment. However, there is no credible way for any firm to tell its employees, "You are getting a wage cut, but don't worry. It's just for macroeconomic reasons."

If this view is correct, then in times of high unemployment macroeconomic policy should aim to boost prices, presumably by expanding the money supply. Printing money should help to avert a general deflation. With sectoral imbalances, printing money should cause prices to rise in high-demand sectors, effectively reducing real wages in low-demand sectors and maintaining full employment in the latter.

Note that if sectoral imbalances are a problem, and creating general inflation in order to reduce real wages in weak sectors is the solution, then there is a Phillips Curve--a trade-off between inflation and unemployment. This model of the Phillips Curve was articulated particularly clearly by James Tobin in his address as President of the American Economic Association.

Note, however, that the Phillips Curve trade-off does not necessarily exist every instant. In order for inflation to reduce unemployment, there has to be a sectoral imbalance that is too large to be resolved by normal market forces. That may not always be the case. In fact, it may only rarely be the case.

Further reading: Martin Weitzman's book, The Share Economy, describes the macroeconomic advantages of having worker pay that varies with corporate performance. An excerpt from James Tobin's address can be found here.


Comments and Sharing





COMMENTS (18 to date)
Manjira Dasgupta writes:

Dr Kling,

Is there room for discussing the first question referred to in Lectures ... I (2008/11/08), viz. "How is it that the financial markets affect the real economy?"

El Presidente writes:

Your analysis seems plausible for workers with options. What about workers without reasonable alternatives? Can their wages be cut without fear of worker flight or decreased productivity? If the employer constraint doesn't fit for those workers, it doesn't work well enough to explain the employment-wage balance.

Cuts in health care, pension, and other benefits are reductions in income/wages. Those have been reigned in quite steadily at most levels of employment. I don't see how firms could pull that off if your explanation holds. Can you explain?

I think the best part of the explanation is the concept of competitive equilibrium between firms in labor markets, but I would say it cuts both ways. It generally drives wages down for the sake of increasing earnings, but selectively up for the same reason. It creates the need for managerial discretion in selecting the most profitable complement of employees and paring the others. This, in turn, places pressure on ALL workers to sacrifice more (increasing output and/or forgoing justifiable wage increases), because the possible alternative is losing their job and having to accept both dislocation and a lower wage.

Don Lloyd writes:

Dr. Kling,

The most common symptom of a recession for a firm is a reduction in unit sales and revenue. Simply reducing wage rates does not result in the necessary reduction in production capacity that is needed to balance a new, lower unit demand. Paying a 20% lower wage to have all your production workers stand around, idle, for the last third of the day makes no sense.

While the firm will certainly tend to lower its prices in an attempt to prop up unit sales, whatever lower marginal cost it achieves by cutting wages is highly unlikely to be sufficient to maintain unit sales in the face of reduced customer unit demand, and production capacity will remain too high.

Every firm has a collection of workers who have a range of marginal costs and marginal productivities. It is rational for a firm to sequentially lay off workers with the largest ratio of marginal cost to marginal productivity.
For reasons described below, this will tend to differentially lay off lower wage workers in a given job function instead of higher wage ones, and the average wage of the remaining workers may well be higher than before. It is not wage rates that fall in a recession, but total wage costs as seen by a surviving firm.

There is likely to be strong correlation between time in service, productivity, and wages, over comparable sectors of the workforce in a firm. As a worker gains experience, his productivity tends to rise. This means that his wage must rise as well to prevent a competitive firm from hiring him away. However, the productivity gains with experience fall into two categories, general productivity and specific productivity. The firm need only pay for the general productivity increases to retain the worker, while getting the specific productivity increases for free because a competitor cannot receive the benefits of productivitity specific to his existing job.

Regards, Don

Diversity writes:

The classical instance of the difficulty of introducing a nominal pay cut in face of price deflation was in the public sector. At a moment when prices were falling in the 1920s, the British Government decreed a - roughly corresponding - nominal pay cut. The enlisted men of almost the whole British fleet mutinied - in a very orderly way - at Invergordon.

(The crew of the one ship that did not mutiny did so because Caoptain had risked his life to save an ordinary sailor the night before.)

y81 writes:

[Comment removed for supplying false email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog.--Econlib Ed.]

floccina writes:

You might be able to get people to accept a pay cut in a down turn if it was negotiated ahead of time with a cost of living wage increase.

spencer writes:

What about the example of Japan, Wall Street and others where a large share of workers compensation is paid in the form of a bonus. When hard times hit the firm cuts the bonus and effective pay without cutting the base pay or laying off workers.

This worked very well in Japan during their era of rapid growth and full employment. It was also an essential element of their "lifetime" employment policy. Of course during the Japanese stagnation it made it more difficult for firms to downsize and lifetime employment contributed to the lost decade.

But on balance the cost of this lifetime employment system may have been less than the benefits. I know of no studies asking this question. It is just like the question of whether the Japanese stagnation was a superior alternative to other possibilities like a real depression.

Famous J writes:

Funny you mention it. I've seen it tried.

I was working in the telecom industry during the "technology unpleasantness" of 2001, when thousand-plus layoffs were being announced every week. At one firm, management announced that effective immediately, everyone from the CEO to the janitor would have to take a 10% pay cut. That was the bad news. The good news was that there would be no layoffs.

About a month later, they laid off about a quarter of their workforce. Let it suffice to say they did not give everyone their 10% paycut back. The people still there felt like serious schnooks, since they were working for less pay and still didn't have job security.

quadrupole writes:

Think about wage cuts from the point of view of a top performer. Talent goes where it is rewarded. If employer A lays of bottom performers, and employer B does across the board wage cuts, as a top performer you want to be at employer A. So if you are employed by employer B, even if the market is such that you can't move now, you *will* move as soon as you can to avoid getting wacked in the next downturn.

It's instructive to note that cutting wages *may* make sense in industries where there is little differentiation between the employees. If the difference of productivity between your least productive and most productive worker is 1% or less (which I suspect it is on most factory floors) then wage cuts might make more sense, as loosing your best employees is not much of a threat. It is interesting to note that many workers in that situation are union protected.


quadrupole writes:

Don, you said

"There is likely to be strong correlation between time in service, productivity, and wages, over comparable sectors of the workforce in a firm. As a worker gains experience, his productivity tends to rise. This means that his wage must rise as well to prevent a competitive firm from hiring him away."

The problem with time in service is that while it can be a factor in productivity:
1) It's asymptotic... lots of studies have shown, for instance, that teacher effectiveness levels off after about the first 10 years of service. So time in service essentially becomes meaningless wrt productivity after a time.
2) There are lots of jobs where time in service is not the dominant factor. Sure, a particular employee may be more productive after 10 years than after 5 years... but a different employee, due to talent or other factors, may be more productive at 3 years of experience than another after 10 years (I see this *all* the time).

Overall, time in service is generally a *really* bad proxy for productivity for the reasons noted above. That's why you *definitely* want to avoid jobs and employers that pay for time in service if you are a high talent worker... that mechanism is guaranteed to screw you.

Don Lloyd writes:

quadrupole,

"...Overall, time in service is generally a *really* bad proxy for productivity for the reasons noted above. That's why you *definitely* want to avoid jobs and employers that pay for time in service if you are a high talent worker... that mechanism is guaranteed to screw you.'"

My claim was not that time in service was being paid for, but that pruning the high cost, low productivity employees (relative)would also be likely to prune lower wage employees at a higher rate than higher wage employees. This would likely be true even if identical, clone employees were routinely brought into the company over time. Experience only builds up over time and the fact that specific productivity gains do not need to be paid for to retain the employee means that time in service not only results in higher wage rates but higher ratios of marginal productivity to marginal cost.

Thanks, Don


Pietro Poggi-Corradini writes:

To make sure I understand: is a raise below inflation considered a pay-cut? In Academia for instance, where it's hard to fire, 0% raises are quite common. I'd suspect that highly unionized firms that have a hard time firing would also use this expedient, but I don't know. On the other hand, a flexible business would try and change strategies or maybe even the products it produces, in which case the need to attract the best people for the right job would actually make wages rise, and some jobs would disappear. Trying to retool while retaining everyone at a lower wage would seem to be costly in terms of reallocation and retraining of personnel.

Brandon Berg writes:

Would it not be possible for firms to cut wages on average by first laying off employees and then hiring workers laid off by other companies at lower wages? Or is this what actually happens, with unemployment clearing up as soon as laid-off workers realize that they're going to have to settle for lower-paying jobs?

Devin Finbarr writes:

There are a number of companies that handle recessions ok by adjusting wages. The easily way to do this is by having bonuses tied to company performance, so if a recession hits, people might lose the bonus but not get fired.

Don Loyd makes a very good point. But even this companies can deal with by lowering prices. The company I work for is renegotiating a lot vendor contracts - office space leases, servers, software, etc. - to take advantage of the recession. So in theory, a company could lower prices to keep demand constant, and then lower the salaries of the workers to make the same profit as before. I actually think this happens to some extent, which is why I don't think sticky wages are the primary cause of recession unemployment.

Unsustainable booms happen because of artificially cheap credit. Recessions happen when credit contracts. As a result, many projects that relied on the cheap credit are no longer profitable. Most of the unemployment comes from these sectors - housing, auto, durables, construction, and banking are often the hardest hit. Many of these businesses shrink or fail outright, simply because they never would have existed in the first place had credit been priced correctly.

The final factor is that some companies use the excuse of a recession to fire workers they wanted to get rid of anyways. I don't know how common this is though.


Don Lloyd writes:

Devin,

"... But even this companies can deal with by lowering prices. The company I work for is renegotiating a lot vendor contracts - office space leases, servers, software, etc. - to take advantage of the recession. So in theory, a company could lower prices to keep demand constant, and then lower the salaries of the workers to make the same profit as before...."

It's not that simple. The assumption has to be that the prices before the recession are profit maximizing and that they will again be profit maximizing after the recession hits and marginal costs are adjusted and the shape of the price-demand curve has changed. It is highly unlikely that produced quantities can be set to be unchanged while finding a new profit maximizing price.

It is even possible that if most of your price-sensitive customers go away, the new profit maximizing price will be higher, but even then the unit sales will be lower and you still have excess capacity to deal with.

Regards, Don


Dave Turner writes:

It seems to be an assumption here that wage cuts affect everyone in a company at the same time. It is clear to me that an across-the-board wage cut reduces morale and encourages your best employees to leave.

Layoffs on the other hand can be seen as a 100% wage cut directed at a few (presumably the worst) employees. I'm not sure about the effect on morale that layoffs have (it's certainly negative too) but at least you retain the best employees.

What about the middle ground? Why, in the normal course of events, don't some employees get pay cuts and others get pay rises?

Stathis Kassios writes:

"If this view is correct, then in times of high unemployment macroeconomic policy should aim to boost prices, presumably by expanding the money supply. Printing money should help to avert a general deflation. With sectoral imbalances, printing money should cause prices to rise in high-demand sectors, effectively reducing real wages in low-demand sectors and maintaining full employment in the latter."


But in the start of the article you said that in times of unemployment workers will reduce spending. So if at these times we increase the prices, aren't we making the workers portofolio/budget even weaker and living standards even harsher unless economists expect/assume that the workers will move to the high-demand sector so that they will enjoy balanced wages again?

Isaac K. writes:

Dr. Kling:
Point of order:
In the short run, an unusually large decline in auto demand (due, say, to permanently higher oil prices) might best be met with lower wages in the auto industry, which could bring down prices and enable firms to sell more cars than they would otherwise.
The problem with this assessment is that auto inventory doesn't exist, and that responsiveness to price is nigh-instantaneous.
Both of these conditions do not hold true. While a large portion of auto sales are presently geared toward smaller, more fuel efficient vehicles, it took a long period of steadily rising gas prices to bring about that shift in attitude. People will not and cannot quickly respond to rising gas by buying a more efficient car. Those that can do so with cash-on-hand are probably not the ones being significantly impacted by the rise in cost of a complementary good.
Meanwhile, all the shiny new cars are still sitting in the lots of hundreds of dealers. All of these were paid for at the higher, pre-recession price.
Car dealers cannot lower the price of '09s vis-a-vis '08 models - they would be stuck with older inventory. Since dealers have a price floor, suppliers will not lower the price. Ultimately, the conundrum of consumer purchasing will prevent price drops on new autos because dealers won't be able to clear out their older inventory, which they won't sell at a lower price because the cost of older models was/is higher than newer ones.
The resultant rigidity caused by price expectation and inventory means that dealers cannot entice more buyers for fear of greater losses, and their demand from the manufacturer will subsequently drop, regardless of the drop in costs being offered them.
As a result, it makes no sense for the manufacturer to lower it's price to the industry. Facing a lower, inelastic demand regardless of it's pricing structure means that it will simply cut production instead. Ultimately, this must entail cutting jobs rather than wages, as was pointed out above.

So, ultimately, consumerism coupled with long-term inventory [that doesn't "spoil"] forces a stable or increasing price structure for middle-men even in the face of decreasing demand. Consequently, suppliers to the market have no incentives to reduce costs directly by cutting wages, since they must cut total productivity instead.

Comments for this entry have been closed
Return to top