The Profit Paradox


*Copyright Princeton University Press, 2021. All rights reserved

Income inequality is certainly a topic de jour among the chattering classes, as is the growth in influence of superstar companies like Apple, Amazon and Facebook. But most people don’t see the two as related issues. Enter Jan Eeckhout, a Belgian-born, British-educated economist at Pompeu Fabra University (UTF) in Barcelona and the author of the new book The Profit Paradox: How Thriving Firms Threaten the Future of Work. The paradox in question is the grim reality that in an era of rapid innovation, the incomes of the majority of workers are stagnant and prospects for mobility are dim. Here, we excerpt the chapter explaining why workers are much less likely to change jobs today than in, say, the 1980s, and how the price of that stability is very high in terms of living standards and chances of being promoted.

Illustrations by john tomac

— Peter Passell

Published October 11, 2021


Chances are that one of your grandparents received a watch from his or her employer when they retired. A gold watch was often the reward for the loyal employee who entered the company out of school and retired from the same company 40-odd years later.

This image, of course, is associated with a different era of work. In current times, in an ailing job market with few attractive jobs and low pay, one would expect that workers are not very attached to their jobs. As a result, workers today must switch frequently. A worker at a fast-food chain that pays low wages might take a week off and look for a new job, since there is not much lost from not working. After all, the pay is low, anyway.

People tend to perceive current times as uncertain. Today jobs are unstable, whereas one or two generations ago jobs were secure and lasted forever, or at least until the gold watch, right?

In fact, the data overwhelmingly shows that the opposite is true: today’s average job duration is actually about one year longer than it was three decades ago. The average job today lasts 4.2 years, whereas it lasted 3.2 years in the early 1990s. Consistent with the increased duration of jobs, we find that the likelihood of switching each year has decreased from 3.8 percent in the mid-1990s to 2.9 percent today.

If workers switch jobs less frequently, then firms renew their work forces at a slower rate. This reflects the decline in business dynamism that has been observed since 1980. Back then, firms changed 35 percent of their work force annually. Currently, firms only turn over 25 percent.

Unionized jobs tend to pay higher wages, are more protected and are therefore longer-lasting. So with fewer union jobs the prediction would be jobs of shorter duration, not longer.

But why is it that people do not switch jobs as fast as they used to? An immediate explanation could be that the decline in dynamism is simply technological progress. A century ago, workers would turn up at the port and be hired as longshoremen on a daily basis. With the advent of container freight, those jobs became obsolete. Now, the logic goes, workers have more long-term employment relations with long-term contracts. The problem with this explanation is that this phenomenon of the decrease in labor market dynamism only started relatively recently.

Researchers have proposed other explanations for the decline, from demographic changes such as an aging population and the increase in demand for relatively skilled workers, to the advent of occupational licensing legislation that protects jobs and stifles labor mobility. However, the evidence shows that these explanations cannot account for the majority of the increase in job duration and the corresponding decrease in labor market dynamism.

At the same time there have been other developments in the labor market, such as the decline in unionization. Unionized jobs tend to pay higher wages, are more protected and are therefore longer-lasting. So with fewer union jobs the prediction would be jobs of shorter duration, not longer.

Market Power and Gold Watches

There is another, more immediate explanation for the decline in labor market dynamism. But bear with me; to get to that, we need to get into the shoes of the owners of competing gas stations. Consider Sunoco and Exxon gas stations at two corners of an intersection. Say they sell gas at $2 a gallon, which reflects the cost, including competitive profits, as compensation for risky capital investment. At that price, customers are indifferent about where they buy gas and the two firms split the market equally. If crude oil prices drop and total cost of producing and selling gasoline goes down by 10 percent, then at any price at or above the economic cost of $1.80 the firms will take no losses and stay in business.

But suppose Exxon keeps the price at $2. Then Sunoco can capture the entire market by lowering its price by a little, say to $1.95. Sunoco will sell double the amount and make an even greater profit than selling at $2 while splitting the market with Exxon, because the much higher volume compensates for the slightly lower price. Since changing gas prices is only a matter of pressing a button that changes the electronic sign, it can be done immediately and at no cost.

Of course, Exxon will not sit still and lose its entire market share, so in turn will respond by lowering its price below $1.95. Eventually, the only stable market outcome is one in which both set a price equal to the cost. In a competitive market, when the cost of the good sold becomes cheaper, prices fall by the same amount as the decrease in cost.

Economists call this complete pass-through. Faced with competition, firms pass on all gains from lower costs to the customer – not because they necessarily love the customers, but because the competitive pressure forces them to. If Exxon keeps prices higher than cost when others don’t, it will sell nothing.

Things are different when there is only one gas station at the intersection. If the next gas station is far away, this station retains some market power and pass-through of the 10 percent fall in costs will likely be incomplete. Sunoco does not at first lose many customers by keeping its price at $2. That’s because many drivers may not initially be aware of the fall in costs, and even those who are aware might not find it worthwhile to drive to the next town where prices are lower.

Eventually Sunoco will lower the price to the point where it can ensure enough customers will come while keeping the price high enough to make a profit – for example, to $1.90. When firms have market power, pass through is incomplete and the firm passes on only part of the cost savings to the customer.

Economists can conveniently measure pass-through when exchange rate fluctuation affects importers (or exporters). Those cost fluctuations are easy to calculate and affect all importers equally. Researchers have found that on average in the OECD countries, around 46 percent of the cost savings are passed on to the customer immediately, and a total of 64 percent over a longer horizon. For the United States, pass-through is even lower – 23 percent immediately and 42 percent over a longer horizon. This is evidence of market power.

Because the knife cuts both ways, cost increases are also (incompletely) passed on to the customer. So the same cost fluctuations lead to larger fluctuations in prices under competition than under circumstances in which sellers have market power. Like a bike with highly pressurized tires, a competitive market translates any pothole as a shock to your body. Market power is more like a motorbike or a car with comfortable shock absorbers; your body feels some movement, but most of the shock is absorbed by the coils near the wheels.

The fluctuations of prices also go hand-inhand with fluctuations in quantities sold, only in the opposite direction. As the price of gas rises, people drive less and therefore buy less gas; as prices go down, they buy more. Therefore, under fierce competition, sharp fluctuations in price translate to sharp fluctuations in quantity sold, whereas when sellers have market power the more modest fluctuations in prices linked to cost fluctuations translate to more modest fluctuations in how much is sold. This implies that the presence of market power leads to less volatility in how much is produced.

When there is fierce competition, every little bump in demand or in costs is passed on to prices; prices then pass it on to quantities, and quantities then pass it on to the number of workers.

Why does this matter for jobs? Firms adjust how many workers they hire in response to how much they produce. This is most apparent when we look at seasonal variation. H&R Block, the largest retail tax preparation company in the United States, hires heavily during income tax season. It has only about 2,700 full-time employees but reaches over 90,000 employees, including the temporary ones, in the months before the April 15 filing deadline. Likewise, restaurants near the beach hire only during the summer, and strawberry farmers during harvesting season. That is why many statistics are seasonally adjusted. The seasonal fluctuations are systematic (they occur at the same time each year), and they are predictable.

Most firms experience very little seasonal variation in demand, but they do adjust employment as they adjust production. With high crude oil prices, gas stations sell less. And when Stan Smith tennis shoes are in fashion again 30 years after they were first produced, Adidas hires more people to make and sell them. These non-seasonal fluctuations are driven by a plethora of factors on both the demand and supply side, from fashion to new technologies.

These fluctuations translate into how much is produced and therefore into how many workers are hired. Because of the incomplete pass-through, the extent to which those fluctuations in demand, technology and costs are translated into the number of workers hired depends on the market power that firms possess. When there is fierce competition, every little bump in demand or in costs is passed on to prices; prices then pass it on to quantities, and quantities then pass it on to the number of workers. When market power is high, dampened changes in the quantities sold lead to dampened changes in employment.

There is ample evidence that the shocks have not changed. Firms are not experiencing higher fluctuations in demand or costs now compared to the 1980s. What has changed is the response to those shocks. Due to market power, workers switch jobs less frequently and the duration of the jobs is longer.

This brings us back to the Gold Watch Myth. Workers now stay in their jobs longer – not shorter – than they did a couple of generations ago. And the effects are large, as we saw earlier. Today’s workers are about only two thirds as likely to switch jobs as in the 1980s.

From Labor Mobility to Migration

It may appear that there is no connection, but the rate at which firms adjust their work force also has implications for the rate at which families migrate between cities and states, and internationally between countries. One of the main reasons that households relocate is job opportunities and promotions. While most of us dream of moving to places where the weather is pleasant, the air is pure and the views are spectacular, the primary reason to move for people of working age is another job.

Firms offer those job opportunities because they are adjusting their production needs. If there is a sudden spike in output needed, hiring increases. Now we know that when firms have market power, they translate those spikes less into output, because with more market power the shocks are absorbed. As a result, firms with more market power adjust their work force less frequently, leading to a decreased need to move between cities. The rise in market power therefore leads to a decline in migration.

This is, indeed, what the data on migration reveal. In 1980, the migration rate between states was 3 percent per year. By 2016, that rate was just 1.5 percent. In less than four decades, the relocation rate of households has halved. Researchers have proposed explanations other than increased market power, such as population aging and the advent of new technologies that reduce the misallocation of people to places. But those explanations cannot fully account for the sharp decline in migration rates.

What makes the changes even more striking is that we see this decline at a time when the cost of migration, both nationally and internationally, is lower than ever. My great grandparents had five children and lived on a shoestring budget in the three decades covering the two world wars. They were modest farmers in a rural village, keeping a few animals and growing chicory. But with the onset of World War I, life was tough.

When the children started to get married and were having children of their own, the family felt the pressure at the table. There was scarcity of land to work on and an abundance of mouths to feed. At different times, three of my great-uncles decided to leave. Saving every franc they could, they each bought a third-class ticket on the Red Star Line from Antwerp straight to Ellis Island, where they passed the medical exam and were added to the Ellis Island registry of immigrants.

Two of them immediately made it to Detroit, where they ended up working at the Ford factory. The other brother ended up as a farmer in Montana after a spell in Canada. One of my great-uncles went to America twice – first to settle and find work, and later to bring over his wife and children.

One of his children, Suzanne, became a bit of celebrity in her birthplace when she came back to Belgium in the fall of 1944 during the liberation by the Allied troops. She had joined the U.S. Army as a nurse and eventually rose to the rank of lieutenant colonel. When she walked into her paternal village, the family members were in awe to see that all American soldiers would stand at attention and salute her wherever she walked.

There is a world of difference between mi94 The Milken Institute Review gration then and now. Moving long distances then was akin to the trip of an explorer to unknown continents. Even as recently as the early 20th century, moving west within the United States meant moving to towns and land that were barely settled, let alone developed. Add to that the high cost and inconvenience of slow and costly transportation by rail or horse-drawn cart, and the fact that people, who were a lot poorer without access to any funds, faced enormous risks.

Things became a lot easier in the decades following World War II with the development of the highway networks, faster trains and air travel. The 1960s was the epoch of massive migration in Europe, within countries in terms of urbanization, and between countries, mainly from the Mediterranean rim to Germany, France and Belgium. In the United States, it was not uncommon for families to move between states every few years. So with cheap flights and extensive travel options, we would expect to see much more migration.

The potentially catastrophic consequences of climate change have brought national planning back to the fore, and with it the question of whether five years is still a useful horizon.

We do, of course, see a lot more travel. According to the UN World Tourism Organization, the total number of annual arrivals in all locations around the world [pre-Covid-19] was estimated to be 1.3 billion, up three-fold in the past 25 years. But the decline in the cost of travel has not led to an increase in people relocating. We move less because we change jobs less – the Gold Watch Myth, now fueled by market power.

International migration is of course very different. For a start, there is typically no free mobility of people across countries. This generates potentially large differences in economic opportunities. By contrast, when there is free mobility between cities within a country, those differences in economic opportunities disappear because people move – the principle of arbitrage.

Wages are higher in New York than in Janesville, Wisconsin, but so is the cost of living. People will move to where they are best off, given wages and housing prices. As a result, in theory people within countries must generally be indifferent when it comes to choosing a location. That is not the case between countries because there is no free mobility. In medieval times there was no free mobility between cities either because only documented citizens had access to enter the gates. This generated similar economic differences that made some cities more attractive than others or made living in cities more appealing than living in the countryside.

The greater the economic differences, the greater the sacrifices people will make to move. Anyone who visited Israel in the late 1990s and early 2000s and rode in a taxi may have had a good chance to encounter a driver who was a reputable concert pianist, a professor of mathematics or a top-notch nuclear physicist.

Many of the Russian Jews who had emigrated had completed many years of schooling and obtained important positions in education, research and culture. An Israeli colleague who is the son of Russian Jewish immigrants told me the running joke among immigrants: “How do you know a Russian Jew in Israel is not a concert pianist? If they carry a violin case.”

When they arrived in Israel, the new immigrants faced three barriers that affected their job prospects and hence their earnings. The first was that many did not speak Hebrew fluently. This put them at a disadvantage for jobs such as teaching, for example. Second, the poor performance of the economy before the collapse of the Soviet Union had left them with little savings, so they needed to get jobs quickly to feed their families and rent places to live.

Third, and most importantly, they all came Fourth Quarter 2021 95 with extremely good credentials, both in terms of their education and past work experience, for the best jobs and for high-ranking positions. As a result, the largest competitors of Russian Jews arriving in Israel in the 1990s were the other Russian Jews living in Israel. With this excess supply of nuclear physicists and concert pianists, you could hire the best at a low salary, and most ended up switching to all kinds of jobs where they could make no use of their qualifications.

Is the Decline in Labor Market Dynamics Bad?

The fact that people change jobs less frequently, and the fact that households are less likely to switch towns, is great for social stability. Children do not have to change schools and leave their friends behind, and one spouse does not have to find new opportunities because the other spouse’s career requires them to relocate. Job stability is the objective of many costly social policies. And as a result of market power, job stability comes for free. But does it?

Customers are paying too much for what they buy. Market power leads to low wages due to the decline in sales that leads to a decline in the demand for labor. Are workers willing to accept a lower income to gain job security?

While the job stability is desirable, we should not forget that the reasons for the decline in job-switching is an increase in market power. Firms are pricing their goods too high, and, as a result, customers are paying too much for what they buy. In addition, market power leads to low wages due to the decline in sales that leads to a decline in the demand for labor. Are workers willing to accept a lower income to gain job security?

It is unlikely that the job security gains do, in fact, outweigh the costs. And the loss does not end there. Job security is great if all workers have the same job and don’t care about promotions, or for those who are lucky and sit at the top of the job hierarchy. But job security is damaging if it reduces the speed of promotions and prevents upward mobility for those at the bottom of the ladder.

For reasons independent of market power (mainly labor regulation), most of Europe, and in particular the Mediterranean countries, are notorious for low job mobility with slow promotions and strong job security. A Spanish friend of mine had worked for years in the financial sector in London while his wife stayed in Barcelona. When they started a family, he decided to return to Barcelona and found a job at one of the large Spanish banks. Locally it is considered the Rolls-Royce of jobs: plenty of benefits, infinite job security and a decent compensation.

But soon after he started, he realized that the opportunities for promotion were exclusively linked to the retirement of his older colleagues.

He came from a dynamic work environment in London where people constantly moved on to new opportunities both inside and outside the bank, which in turn created promotions inside. He had moved up through the ranks very quickly. But at the bank in Barcelona, promotions were rare. The great Paul Samuelson, a Nobel Laureate in economics, said the following about scientific progress around a table of economics professors, but I guess it could be paraphrased around the table of any human resource meeting in a Spanish company: promotion happens funeral by funeral by funeral.

main topic: Workforce
related topics: Books