The Labor Market Isn't Broken

Journalists have recently brought attention to a very low unemployment rate — however, strong wage growth typically associated with a robust labor market is still missing. In the U.S., the unemployment rate has gone under 4% — the lowest it’s been in two decades. Despite this, the year-on-year change in average wage has only marginally grown from 2% in 2010 to 2.9% in 2018.

Traditionally, those familiar with the most basic of economic models apply the theory of supply and demand to the labor market. If there are lots of unemployed people, they will compete for jobs, driving wages down; conversely, if there are few unemployed people, employers must compete for the workers, driving wages up.

This model of supply and demand is so simple and logically coherent that it’s easy to understand why economists often use it to explain price behavior. After all, it’s the prevailing theory for modeling the response of commodity prices to changes in their consumption and production. When OPEC decides to manipulate oil production world-wide, this model routinely predicts the price response.

To some, however, the lack of recent wage growth characterizes a failure of the labor market. If the unemployment rate is approaching the lowest sustainable unemployment rate, why aren’t wages increasing more readily? One Business Insider article put forward a hypothesis:

Employers — and, belatedly, economists — are waking up to the fact that the old-fashioned supply-and-demand model of the labor market is dead. Employers have gained enough power in the marketplace to permanently hold down wages, even when unemployment is as low as 3.9% in the US and 4.2% in the UK.

Jim Edwards, the author of the opinion piece, placed blame on the employers for the recent failure of this model. The fault in his argumentation, however, lies in the fact that supply and demand isn’t the only factor in labor prices. While that model can be useful in limited situations, it never explains the whole picture. Supply and demand’s failure to explain the whole picture doesn’t mean the model is broken… it means it never worked in the first place.

Supply and Demand Never Worked

For his argument, the author ignores the subjective theory of value. In doing so, he neglects to mention that not all human labor is equally valuable, nor is all human labor equally productive. Instead, due to differences in ability, experience, and skill, wages are a distribution, differing significantly by sector.

This graph shows how wages differ greatly between occupations — not only in absolute terms, but also in annual wage growth.

US real wage growth

Right now we are living through the most supply-restricted wage market since the 1970s. There just aren’t extra workers available. In the US, there are 6.7 million job openings but only 6.3 million people looking for work.

Interestingly, Edwards presents theses statistics as if they’re meant to be subtracted. 6.3M unemployed - 6.7M openings = 0.4M worker deficit. But why aren’t prices going up, if the demand for workers outpaces the supply?

These statistics describe two things only: the number of open jobs and the number of job seekers. Because what they describe, however, cannot be simplified into a number, predicting wage behavior from those two numbers is rather naive.

This approach neglects differences between sectors and professions. If there were a systematic difference between what the job market seeks and what the unemployed provide, this would negatively contribute to improving wages. There would be little competition for the jobs employers seek, thus holding down wages — which is not, by the way, the employers’ fault.

Government intervention can often cause this phenomenon by subsidizing industries with unnatural (extra) demand for their goods.

Keynesians, in their relentless pursuit of increased consumption, love manipulating the market in favor of short-term gains without thought for the long-term consequences. These long-term consequences can often ruin the labor market. One example of this is the Car Allowance Rebate System, the $3 billion government program fondly known as “cash for clunkers.”

The 2009 program subsidized the purchase of new vehicles, increasing the demand for cars unnaturally. It was quite short-lived; the government burned through all $3 billion in just two months, and ultimately taxpayers will foot the bill. Regardless, in order to meet the increased demand, workers had to be hired and trained for a job that wouldn’t have existed without the subsidies. When the government subsidy ran out, they lost their jobs and sought work with job experience nobody needed.

One former worker at a car dealership wrote an article describing his personal experience in the program:

Sales began to slow down dramatically following the program’s end… Car salesmen began to be laid off. The assistant manager of the dealership was let go, as were many of the mechanics and staff running the automotive-parts department. Because I only worked a day and a half a week and was paid a dollar an hour less than the other experienced detailers, I was kept around. Once December hit, the dealership finally shut down for good.

All of the workers in car dealerships across the country then sought out to find jobs in other markets. Changing market conditions mean workers must be retrained, and this detracts from the supply and demand model. It’s not a perfect model for complex systems, and by all means, the labor market is complex.

The New York Times gave six reasons why wage growth never picked up:

The reasons for sluggish wage growth, of course, are a complex weave. Declining unionization, noncompete contracts, tepid minimum-wage increases, globalization and sluggish productivity have all played a role.

It’s almost as if the complexities of the labor market cannot be simplified into such a black-and-white model, only to be falsely criticized when it fails to explain wages.

Sticky Wages

Additionally, the labor market is known for an interesting property called “sticky wages.” This is when current wages don’t quickly change to match the conditions elsewhere in the labor market.

To the misinformed, this means that the labor market is inherently inefficient. An all-knowing and well-intentioned central planner could perhaps make the whole system efficient — by adjusting minimum wage laws according to the economy’s health, or even price-fixing wages themselves. A country with less unemployment is always better, after all.

This again, however, ignores the complex calculations individuals do to determine the value of labor. The appropriate wage isn’t solely determined by the surrounding labor market; individuals determine the true value of labor, not professional economists, or where the supply and demand equilibrium curves intersect on an arbitrary graph.

One example of these complex calculations is well-documented in the reverse case of what this article discusses. Here, we’re discussing how unemployment can decrease without a corresponding wage increase in a recovery. Conversely, historical evidence shows that recessions often come with high unemployment without a corresponding wage decrease. This equally contradicts the supply and demand theory of labor.

Economist Truman Bewley wrote Why Wages Don’t Fall during a Recession to investigate the seeming discrepancy. He found that employers value the morale of the whole team over layoffs of the minority, explaining why they choose to fire a few instead of offer a wage cut to all.

They believed that cutting wages would hurt morale, which they felt was critical in gaining the cooperation of their employees and in convincing them to internalize the managers’ objectives for the company.

Even with holistic statistics in hand, trying to outsmart the economy by fixing a “broken” system is often misguided. Individuals act in complex ways to pursue their goals, and policy prescriptions on behalf of millions can often be mistaken in their understanding of the whole situation.

An Empirical Confirmation

Arguments in favor of this supply and demand model in the labor market regularly cite statistics and graphs to prove their claims. Unfortunately, economic statistics are often misleading.

unemployment rate vs. nominal wage growth

The above graph shows the correlation between wage growth and the unemployment rate, clearly showing that as unemployment increases, generally, wage growth decreases — and vice versa. This graph is modeled after one done in the New York Times a few months ago. But like most economic analysis, it’s ripe with misinformation and confounding. Statistics can be manipulated greatly to support any conclusion, and this graph is no exception.

New York Times wage growth vs. unemployment rate

The NYT graph ended in 1995. Mine ends at 1984, which I chose only because it helped the typical labor market failure narrative. The data set I was working with reached back to 1965, but the points inside the first two decades didn’t fit. In order to show how misleading these graphs can be, I removed them.

And just like the NYT, I used nominal wage growth instead of real wage growth. Nominal is the number price at that time — a value that changes regularly due to inflation and deflation. In the scope of history, $100 is meaningless without a time frame for reference, and so comparing the nominal prices of a single commodity has little meaning. After adjusting for inflation via the Consumer Price Index (CPI) — which takes the price of a typical “basket” of consumer goods to reflect the purchasing power of a dollar — this relationship between unemployment and wage growth disappears entirely. This is because unemployment is often highest during a recession, and recessions often mean significant changes to the purchasing power of money.

unemployment rate vs. real wage growth

When adjusting for inflation (“real” wage growth) and including all data points, it’s clear that there’s no real correlation between the two. Empirically this almost disproves the idea that supply and demand was ever a meaningful model to use when analyzing the labor market.

Additionally, wage growth statistics don’t even paint the whole picture. Today, a significant portion of total employee compensation takes the form of one-time bonuses, which aren’t included in wage data.

The Gig Economy

Unemployment is low because a massive number of new jobs today are part-time gig-economy jobs. Part-time “underemployment” has, statistically replaced the mass unemployment we remember from the 1980s. As Johnson says: “A majority of people who are classified as poor now live in a household where someone is in work. This is a complete turnaround from 20 years ago when two-thirds of the poor lived in workless households.”

Edwards’ allegation that the gig economy is at fault for wage stagnation is neither well-founded nor substantiated by any cohesive reasoning. His theory is worth disproving, as a misunderstanding of the economy leads to legislative band-aids that often hurt much more than they help.

The gig economy isn’t hurting workers — it’s helping them.

Pay rates no longer move upward as unemployment moves downward because companies like Uber, Just Eat, and Deliveroo can switch their demand for labor on and off on a minute-by-minute basis.

First, it’s already been established that supply and demand never worked in the first place for modeling the labor market on a grand scale. There are too many factors at play and empirical models fail to show a strong link.

Besides, his argument is illogical. Supply and demand would work just as well on an hourly basis; competition doesn’t need months to beat wages back up. If Deliveroo pays workers more during lunch and dinner time, drivers will seek to work there first, before moving to Uber for the more lucrative evening hours. If anything, these gig economy jobs have made the job market more liquid, which only enables workers to find the highest paying jobs more readily.

Additionally, he ignores the fact that these three companies don’t randomly switch their demand for labor on and off; instead, the companies are simply proxies for their customers’ demands. Treating these companies as traditional employers is misleading and creates a bias towards regulating them. If Uber turns off its demand for labor, it’s because nobody seeks to be driven around nearby. Conversely, surge pricing attracts more drivers when demand is high. This isn’t a part of some sadistic CEO’s plan to bankrupt the workers; it simply isn’t efficient to pay for labor when nobody is using it. The market can use it more efficiently elsewhere during that time.

… In the old days when you lost your job you claimed unemployment benefits and lived “on the dole.” Today, you deliver pizzas or stuff packages in an Amazon warehouse for 20 hours a week.

Here is the alternative option Edwards puts forward: being completely unemployed.

When someone is unemployed, their labor is left unused and renders them completely unproductive to society. Additionally, they tend to spend less, as they must hedge against a long job search — further causing issues in the economy. Indeed the unemployed are even counterproductive, because they consume welfare benefits paid for by productive workers.

When working in a part-time job while seeking a full-time one, workers put their labor to a fruitful use (not just digging holes and refilling them, as Keynes suggested), and spend more elsewhere in the economy. It’s clear that this is the superior situation, not living “on the dole.”

Working is no longer a guaranteed way of getting ahead. Instead, it may keep you poor. You cannot get rich working for Uber. You cannot get rich working for Deliveroo.

This, unfortunately, is where the economic misunderstandings turn into mistaken entitlement. Working was never a “guaranteed way of getting ahead,” and anyone who thought that fundamentally misunderstood labor. The labor market is where people freely exchange money for services; it’s not some idealistic machine where the hard-working are rewarded for their efforts, and not their results.

You can get ahead by increasing the value of your labor. Go back to school, learn a skill, build your experience, grow your network. Skilled workers still make immensely more money than unskilled workers — this is a market incentive to grow your abilities instead of sitting at the bottom rung of your company’s ladder, waiting for a raise to be handed to your entitled self. This is how workers get ahead; not by going from grunt to CEO, but unskilled to skilled.

The obvious function of the so-called gig economy is to create inequality. “The unemployed” now barely exist. In their place are millions of people stuck in jobs that offer too few hours to get by or jobs that don’t come with a career ladder offering pay advancement.

There is no “function” of any economy. An economy is just a phrase to describe people trading of their own will. Besides, nobody created the “gig economy,” let alone with the intent of creating inequality; that’s laughable. Uber, Deliveroo, and Just Eat are there to make profits in a market niche, not ruin the poor.

A career ladder with pay advancement is best done with growing your skillset, not sitting and waiting for raises. Additionally, unemployment is clearly worse off than being partially employed.

What About the Minimum Wage?

Fortunately for the abused part-time workers, Edwards offers a solution that ought to at least partially fix the gig economy, stagnant wage growth, and inequality:

… The minimum wage suddenly becomes the most important political issue for workers. It’s the only way to get a raise in an economy that does not create full-time jobs or jobs with regular pay increases.

In light of his complete misunderstanding of the economy earlier, it’s hard to believe that the minimum wage is an easy “fix” to stagnant wages. He might not know who most often works at the minimum wage level: suburban teenagers. These aren’t the oft-discriminated workers he purports to describe. One report from The Heritage Foundation wrote:

… Most minimum-wage earners are young, part-time workers and that relatively few of them live below the poverty line. Their average family income is over $53,000 a year. A hike in the minimum wage primarily raises pay for suburban teenagers, not the working poor. If Congress and the President seriously want to help the working poor, they should look elsewhere.

Raising the minimum wage isn’t the only way to get a raise. This economy does create full-time jobs (6.7 million jobs, to use his own statistics). There is no political solution to this problem… this is a personal problem that workers must solve on their own. Everyone must take responsibility for their own earnings, instead of waiting on the government to come and save them from driving an Uber for the rest of their lives.

This entire critique of the gig economy, labor market, and part-time employment is nothing but an attempt to excorciate some aspect of the free market before lobbying for more legislation, solving some arbitrary problem. But the problem doesn’t even exist, and it’s clear Edwards doesn’t even fully understand the market he attempts to fix.