Why Wages Don’t Fall During A Recession

I have finished reading a fascinating book by Truman Bewley called “Why Wages Don’t Fall During A Recession”. It’s an interesting book not only for its topic but also for the way in which the author conducted his research. Unlike most economists who conduct studies based on complicated mathematical models, Bewley did something unusual and interviewed business owners to understand more about how they run their business. Economists traditionally viewed the market as automatically self-adjusting so that wages and prices would easily change to the right level in response to market conditions. However, it has been found that wages are rigid and almost never decline so between 1992 and 1994 Bewley interviewed 336 people in the North East of the United States (the book was published in 1999). The studies were meant to be qualitative and as such are not random or representative. They provide a very interesting insight into the mind of business managers.

The data is slightly dated as it relies on a recession from 20 years ago. Our current recession is far more severe than the 1991 recession, but wages have risen by 10% in the EU since 2008 (Greece is the only country where wages have declined). Even Irish wages have not declined, instead lower demand has been reflected in higher unemployment.

A key point of the interviews was how much managers emphasises the importance of morale. They emphasised that workers couldn’t be treated like equipment or else productivity would suffer. Workers had to be treated well or else they could damage the company. A reputation for bad morale would make recruitment and retention of staff difficult. Even in recessions morale was important as unhappy staff would leave as soon as the economy recovered. Interestingly, when Bewley mentioned that psychologists found little link between morale and productivity, most managers believed the psychologists were wrong.

Most internal pay systems emphasised the importance of equity and fairness. Managers often imposed bureaucratic pay structures in order to avoid favouritism or overpayment and to provide clarity about how much everyone was earning. Contrary to economic theory, it was seen as harmful to pay workers their marginal productivity because this was a) unobservable and b) harmful to morale as people were suspicious of each other and why some were earning more than others. It was seen as more convenient to pay workers doing similar jobs the same wage to avoid some feeling underpaid.

External pay structures were less relevant than internal ones as workers didn’t have much information on other companies and much of their skills would be job specific. The strong ties between workers and their job, the difficulty of finding a new job and lack of information all contributed towards blurring of market forces. Employers didn’t want pay to be relatively too low as it would make it difficult to hire, keep turnover low and encourage the workers to join a union. One of the most interesting findings of Bewley’s research was that there was no such thing as a market wage. Employers and employees alike had only a vague idea of what other companies were paying. At most they only knew rough ranges, which Bewley found to be incredibly broad, the difference between the minimum and the maximum were anywhere between one-tenth and one-half. Interestingly the only place where a market wage exists is in unionised industries. Unions insist that workers doing the same job in different companies get the same pay. I found it amusing that unions which are usually blamed for distorting the market were the only ones who created the market conditions economists believe exist everywhere.

Workers are discouraged from seeking other jobs while still in employment as this is seen as disloyal. If a worker does receive a job offer from another company, managers are reluctant to give a counter offer even if the worker is valuable to the company. Counter offers antagonise existing staff who may want their wage to rise as well. Counter offers damage relations with the employer and even when they are made (which is usually to highly skilled staff) the employee usually leaves after a few months anyway.

A popular explanation of wage rigidity among economists is the Shirking Theory that employees are paid more in order to discipline them not to shirk or doss. Almost all managers disagreed with the theory and said it didn’t apply to them. They viewed the theory as relying on threats and punishments, which in their mind were the symptoms of bad management. Rather it was viewed that most people will work hard naturally if given the chance.

It is often considered by economists that businesses could hire new staff at lower wages in order to reduce the wage bill. However this is hampered by the need for internal pay equity, as new hires would soon become resentful at getting paid less for the same work. Existing staff would be made to feel insecure. Potential hires have little to bargain with as they rarely know what the established wage system is at the firm. New hires may get paid less during a training period but this is usually brought up to the rest as soon as they are trained in. Interestingly, Bewley found that not only did wages not fall, but in fact raises were often given. Even in a recession employers worried about their best staff leaving for other jobs or more likely, leaving as soon as the economy recovered.

The consensus among managers was that pay cuts would cost the firm more than it would save them. It was felt that pay cuts would cause damage morale and lose the business their best workers. Interestingly unions were not mentioned as a reason for resistance (probably due to the small number of unionised workers). It was felt that it was not possible to cut the wages of all but the best as this would create tension in the workforce. It was felt that pay cuts would be interpreted as an insult and make enemies out of the employees who would resent the company. It would damage the bond between employees and their work, making them feel that their company didn’t stand by them, so why stand by it?

Instead of cutting wages, managers preferred to lay people off, though interestingly, they did not see any trade off between wages and employment. This was primarily due to the fact that labour costs were often a small percentage of total costs and product demand was not very elastic. It was felt that during a recession there is less demand and therefore less work to be done, so it makes sense to fire unneeded staff rather than cut wages. If the firm only cut wages, they would be left with people standing around with nothing to do. It was also felt that layoffs would do less damage to morale and productivity. Interestingly, the idea of cutting prices was rejected by most managers as it would either not be effective (due to inelastic demand) or start a price war. None of the companies offered their employees a choice between layoffs or pay cuts (most were astonished at the thought).

Firing staff and replacing them with cheaper workers was rejected by most managers as damaging the relations between employer and employee. They felt it would be seen as damaging to morale, heartless and disloyal. It would make employees feel like they were replaceable machines and therefore put less effort into their work. It was noted that the recession increased the labour supply and the quality of job applicants. Bewley uncovered a surprisingly refusal among many employers to hire people they considered over-qualified. It was felt that they would be unhappy with the job and quit as soon as they found something better. It was feared that they may disrupt the hierarchy and cause trouble for their supervisors.

Unemployment was generally seen as a bad sign among job applicants. Managers felt that only the worst employees get fired and good workers would have quickly found jobs. Government welfare was not believed to be a disincentive to work; the only exception to this was the employers of low wage temporary labour. Employers disliked using public advertisements for job positions as they usually received a deluge of applicants. Rather they used personal contacts and word of mouth. References from previous employers were not considered helpful as employers were reluctant to criticise past employees for fear of being sued for libel.

Bewley found little evidence to support many popular labour economic theories. He mentions that he “found no support for real business cycle theory. During hundreds of hours of contact with businesspeople, I never heard one description of an exogenous productivity decline.” Lucas-Rapping theory of unemployment (that it is caused by workers voluntarily quitting their job) is found to have little basis in reality. Unemployment is primarily caused by layoffs and quits actually decline during a recession. Nor do workers quit rather than accept a pay cut as they are never given the choice. Unemployment is an unpleasant situation that no one chooses; rather most unemployed were desperate for work. The main theories of strikes were also found to be inaccurate. Strikes do not occur as there is information asymmetries over the financial health of the company (which is usually known) or to test the profitability of the firm (unprofitable firms were as reluctant to compromise as profitable ones which had greater resources to fight the strike).

Bewley finds that most labour economic theories are inaccurate and fail due to unrealistic assumptions. The only exception to this is the morale model of Solow and Akerlof. Bewley criticises the traditional economic view of people being self-interested and having to be either bribed or coerced into performing tasks. Instead he found that employers ascribed different motivations to their staff. Instead they focused on building good relationships with their employees and encouraged them to take pride in their work and to enjoy their job. They want their employees to show initiative and to perform tasks without having to be constantly told to do so. They felt that treating workers like just another expense like equipment would be counterproductive. As intuitive it may be to ordinary people, it is surprising to economists that people resent being treated like machines whose only purpose is to make money.

In conclusion, Bewley has written a fascinating book that contains many insights in how labour markets work in reality. It reveals major holes in economic theory beyond the simple belief that wages adjust until they are in equilibrium. Bewley has made a great contribution to explaining downward wage rigidity and I would highly recommend his book.

16 thoughts on “Why Wages Don’t Fall During A Recession”

  1. If the survey is exclusively the North East of the United States of America, then it is a pretty narrow analysis. Worldwide the picture would be very different, and one should not forget that for the hundreds of thousands of people who lose their total incomes in a recession, that’s a significant drop in wages.

  2. Good review! Paul Krugman recently brought up this topic to point out that economists in general had no clue as to the source of wage stickiness but then mentioned the maverick economist who went out and did the unthinkable, asking people why. Voila!

      1. not funny (ha-ha) at all. funny (peculiar), yes. funny (surprising), no – to this skeptic (not cynic, yet), anyway. Economists try to emulate the methods of Physics (as pointed out in “The Economics Anti-Textbook) and treat Economics as a positive science, which, btw, is funny (ha-ha only). Gee, I wonder why Physicists don’t talk about emulating the methods of Economics.

        Such surveys as this one, while “unscientific”, i. e., taken as assessments of data (mute, unthinking data), are excellent starting points for further analysis and provide guideposts along the way as to which questions are important, answerable and fruitful and how data for their analysis should be gathered and interpreted.

        To recall the joke about the Priest, the Philosopher and the Economist stranded on an island with a boat full of canned goods, “Assume a can opener!”

        To illustrate further the detachment of Economists from the workplace, this discussion about worker morale and its relationship to worker efficiency and effective dates back to the 1950s among U. S. business academics. There are, in broad outline, three views: Frederick Taylor’s (Theory X), Herbert Simon’s (Satisficing) and Chris Argyris’s (you guessed it – Theory Y). Theory X is that job satisfaction/happiness is irrelevant; pay motivates workers to do their jobs and fear to show up for work (no work, no pay). Their jobs are made as simple as possible to reduce error and improve speed of production simultaneously. Such work is simply a path to mechanization, which occurs (according to this theory) when marginal productivity can be improved enough to offset the initial investment in machinery. Simon’s theory of “satisficing” suggests that workers are motivated by factors other than their financial well-being (power, influence, status, peer acceptance, approval from authority figures, etc.). They will work hard enough to achieve an acceptable balance between the demands of work and the pleasures of leisure. This balance differs among individuals. Donald Trump locates his fulcrum at one point and Lao Tzu would have located his at another. As Bewley’s work points out, employers are quite aware of “satisficing” and quite unaware of each worker’s idea of “satisfactory”, especially at someone else’s company.

        Theory Y is that, you guessed it, worker productivity is related to worker happiness directly: happier workers are more productive than unhappy workers. Employers are also vaguely aware of this situation, and they struggle to find ways to keep them “happy” while promoting greater productivity. They do know it’s somewhere between paying them a lot to do nothing (or very little) and paying them very little (to do a lot). They know, too, that “it’s complicated”. Many factors affect happiness and productivity in the workplace. Awareness of this fact is neither universal nor restricted to a very few. However, few seem to have any clue about what to do with it.

        Bewley did the right thing by starting with the obvious – if we want to know how employers address this question, let’s get up out of our arm chairs and ask ’em.

        More generally, economist Richard Wolff said, during his appearance on Bill Moyers’ Journal, that in order to learn how economics really works, he had to go back to school in a business school. There, he learned that businesses affect demand in many ways (the science of this practice is pretty well-developed), that worker motivation is related to productivity, that companies discourage competition to protect their market shares, and companies discourage innovation and invention to protect their existing products. I learned this in B-school and saw it confirmed every day for 30 years (not exactly a scientific study…).

  3. I think individual business owners behave as you say.
    Big corporations and other big institutions are different. They commonly take away benefits and ask workers to take pay cuts or do without pay raises – Boeing in Seattle being the most recent example.
    And of course wages went down during the last recession due to workers being laid off and taking new jobs at lower wages.
    I don’t think “wage stickiness” is a big economic problem here in the USA or elsewhere.

    1. The data would disagree with you, if it could speak for itself. So, I’ll speak for it. There is a distinction to be made between wage stickiness and job loss. As people lose their jobs and find other jobs that pay less, average wages and annual income decline as a function of mathematics. But, for the people who retain their jobs, wages remain constant or fall only by the amount of purchasing power lost to price inflation. In recessions, the rate of price inflation is low, so wages aren’t reduced meaningfully for those who retain their jobs.

      Note this graph from the St. Louis Fed database (FRED):
      http://research.stlouisfed.org/fred2/series/AHETPI (copy this link and paste it in your browser).

      Company (large, mid-market and small) managements are aware of this problem. If they cut pay among workers they retain, those workers will probably become less productive or seek employment elsewhere (or both, as happens frequently). So, they reduce their payroll burdens by increasing the workers’ share of medical and life insurance benefits and reducing their pension benefits by reducing them directly in the case of fixed benefit pensions, by replacing them with 401K plans, by eliminating benefits altogether, by reducing their weekly hours or by “furloughing” them or laying them off (to be recalled later). They hope to mollify their employees by keeping their direct wages and salaries at historical levels while they seek to muddy the waters by reducing other forms of compensation (which are harder to compare than wages).

      1. I still stand corrected, and I do not disagree with anything that Robert Neilsen or Charles Broming have written so far. However, here are some links to additional FRED data that illustrate the complexities of the subject.




        The first series shows that while nominal U.S. hourly wages remain constant, the number of hours worked has declined. This reflects what I see going on around me, which is employers limiting weekly hours to a threshold which will not trigger an obligation to pay certain benefits.

        The next series shows that compensation of U.S. employees of non-financial businesses does, in fact, fall during recessions, and the third series shows that labor’s share of U.S. GDP falls during recessions.

        Whatever may have been the case during the Great Depression, I do not think that wage stickiness is a bar to economic recovery in the United States or in any other country. Neither of you said it was, but I still think the point is worth making.

        1. First, thanks Phil. Those are excellent tables and charts. And, “standing corrected” is rare, even though commendable, in the blogosphere. I do agree that the unemployment picture is more complex than my narrative depicted, but, I’m tickled that it led to further investigation.

        2. Re your ultimate point, I agree: wage stickiness-on-the-downside (analogous to price downside stickiness) is not a bar to economic growth or recovery. Sorry I left this part out of my reply. I just agree for different reasons. One could argue persuasively that wage-rate growth is as necessary as hours-worked growth to support growth or recovery; better pay for given work is as crucial for growth or recovery as more work (conversely, that falling wages deepen and prolong recessions because they reduce demand for goods and services, thereby amplifying the rate of economic activity shrinkage during the initial phase before wages adjust). Some economists have published studies that support this alternative hypothesis (in respected, peer-reviewed journals). A more fruitful perspective may be gained by examining employment participation data as well. You will find that the rate of participation in the economy fell, in 2009, to its lowest since the 1930s. More than one single-factor explanation is possible, but, a multiple-factor explanation would be “better”. The recession of 2009-2010 is one large factor in this decrease in participation, of course, but so are the aging of the workforce (and agism among employers) and the growth of the workforce (more women, more two-earner households, more minors and young adults seeking employment, for example). This problem becomes a partial differential equation involving the rates of change of at least these variables ( y = f-prime(job loss rate) + g-prime(aging rate) +h-prime (women entering the work force) + j-prime(college students entering) + k-prime (high school students entering + …). Fun, huh? Not only is this problem challenging as a math problem, the underlying data is not as reliable as, say, data from a physics experiment.

          As far as the distinction between individual business owners and big companies’ labor practices, the anecdotal data I’m aware of is that my narrative fits the behavior of large companies as well as small companies. Let’s cut to the chase. I hear of large employers threatening to cut full-time employees’ hours to part-time in order to avoid subsidizing medical care that meets the minimum standards specified in the Affordable Care Act. In Orlando, Darden Restaurants, Inc., have threatened this option (by stating that they are “exploring” it), but, have not implemented policy to achieve this objective, yet. A cost-benefit study may suggest that to implement such policy would be costlier than increasing the subsidy of benefits. A more frightening and more interesting prospect for restaurants is the passage of a minimum wage for restaurant tipped workers that results in parity with other workers without including estimations of income from tips. As of now, the argument that the A. C. A. will result in coverage of fewer Americans is purely ideological. The data will confirm or disconfirm this hypothesis in a year or so; the data collected and disseminated to date suggests that it will disconfirm this hypothesis.

          One of Robert’s favorite books is “The Economics Anti-textbook”, which I like, too. In it, the author’s cite some studies that, if they’re not strong enough to debunk it, cast serious doubt on the veracity of the hypothesis, “A minimum wage increases unemployment among young or unskilled workers.” I use “or” in the inclusive sense (Did you eat all of your dinner or did you save room for dessert? Yes). For example, the trend among economists is toward disagreement with it. In 1979, 68% of economists surveyed agreed with it generally while 10% flat-out disagreed; in 1992, 56.5% agreed with it (generally) while 20.5% disagreed with it (flat-out); and, in 2003, 45.6% agreed with it and 26.5% disagreed. More remarkable than the trend, to me, is the fact that in 2003, fewer than half of economists agreed with this hypothesis, yet, the example of minimum wages increasing unemployment is among the canonical, textbook examples of how government intervention (“manipulation” to the 45%?) affects the workings of an economy according to the canonical supply-demand model of human resource markets. Confusion about this issue among economists has increased as their ability to gather data has improved (no surprise). The work of Card and Krueger, (The Myth of Measurement) ignited a raging controversy (resulting in some public name-calling!) among economists about this issue.

          There are deeper issues, here. First, scientific reasoning is essentially “post hoc ergo propter hoc” reasoning. Events occur and are measured; the measurements (data) are organized and analyzed. Analysts posit causal relationships among events to explain the data at hand. As such, this reasoning is deductive only insofar as a hypothesis follows from a set of statements already accepted as true by a given scientific community (current theory or received wisdom, depending on how you feel about it); we, as amateurs or pros, posit candidate explanations after sorting out those candidates that are obviously inconsistent with this given body of theory. Then, we select the “most probably ‘successful’ ” candidates for empirical testing, either by experiment or analysis of data, one at a time. There are many mathematical tools available for this activity. The researcher’s choices of tools will influence his conclusions.

          Second, as Poincare points out (in “Science and Hypothesis”), there are infinitely many models or explanations (take your pick) that describe or account for any finite set of data (all data sets are finite necessarily). Social science data and the events that produce them are Gordian. The method of measuring events, especially how the decision variables are isolated from the non-decision variables, influences (some would say, “determines”) investigative results/reports of outcomes significantly.

          Third, the manner in which economists pass Economics to new generations of economists, political scientists, policy makers, legislators and citizens exhibits biases that are detrimental to sound policy. So, not only are research conclusions influenced by methods, so, too, research methods are influenced by the theoretical frameworks of researchers.

          We can and must reach conclusions about the world. But, we must be aware of their fragility, even when buttressed and cemented by some current theory. Our behavior and willingness to change it to meet the challenges of changing circumstances depends on our grasp of the fragility of our beliefs. Thanks for looking up those time series and sharing them with me.

        3. Charles, I think we have a fundamental disagreement in our approach to things that is unlikely to be resolved in this discussion. I am not a scientist. I do not base my opinion on what is going on in society on data. Rather I base it on what I see going on around me, and turn to the data for background and context.

          What I see going on around me is an economic system that is growing harsher and more unforgiving. I see wages and benefits being driven down, and economic values displacing human values. I see executives and consultants with six-figure incomes figuring out ways to squeeze more work for less money out of people with five-figure incomes.

          I see this among acquaintances working in, or recently retired from, manufacturing, retail employment, journalism, public schools, higher education and the postal service. I see unions being successfully pressured for more and more give-backs. This is not new. It has been going on for many years.

          I know of people who were laid off from Eastman Kodak Co. in the old days, and then hired back as temporary employees at lower pay with no benefits—in a couple of cases, at the exact same jobs. Talk about your Theory X and Theory Y!

          I think the claim employers are cutting back on labor hours because of Obamacare is a red herring. Employers have been minimizing benefits by capping labor hours for years.

          I do not deny the importance of economic and demographic trends, but I think human agency and public policy are equally important. Costco pays its employees decently; Wal-mart does not. The reasons lie with the decisions of individual human beings, who have names, addresses and telephone numbers. (It is true that impersonal economic factors set limits on what each of the two companies can do.)

          Two books that more-or-less reflect my own perspective are WINNER-TAKE-ALL POLITICS by Jacob S. Hacker and Paul Pierson (a poorly titled book which discusses how public policy promotes the upward redistribution of income) and THE COST OF INEQUALITY by Joseph Stiglitz.

          I do not deny the value of the technical study of economics, but it is not my starting point.

          1. I agree with everything you said, except that we have a fundamental disagreement. I regret not being clear. I, too, begin with observations about the world around me and what others say about it. Then, I react emotionally and intellectually, the outcome of which is, “Hmm, why is that true, false or neither and what does the data say about it?”

            A typical reaction for me includes some reaction to the assertion’s language, to it’s logic and to my version of the “facts”. I have found that in many cases, I didn’t have a well-formed (much less well-informed) view on such-and-such subject, sometimes to my pleasant surprise. So, in my quest for clarity of an idea and it’s fit with the world around me, I look at data that others have gathered, read and listen to others’ assessments of this world as relevant to the idea I’m exploring and attempt to formulate a position that is clear, fruitful and defensible. In other words, I do much of what you do. Economics isn’t my starting point, either. I see everything through the lenses that I have acquired with much effort from Philosophy and Mathematics.

            Btw, Stiglitz is among my favorite economists and I’m aware of (but haven’t read) Hacker and Pearson. You might enjoy Robert H. Frank, who wrote, THE WINNER-TAKE-ALL SOCIETY: WHY THE FEW AT THE TOP GET SO MUCH MORE THAN THE REST OF US and FALLING BEHIND: HOW RISING INEQUALITY HARMS THE MIDDLE CLASS.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: