Glenn Loury interviews David Neumark about the minimum wage

From Bloggingheads.tv comes this interview of David Neumark by Glenn Loury about the effects of the minimum wage.

Is the Phillips Curve Still a Useful Guide for Policymakers?

Morgan Foy writes in the September 2019 issue of the NBER Digest on the question, Is the Phillips Curve Still a Useful Guide for Policymakers?

The Phillips curve, named for the New Zealand economist A.W. Phillips, who reported in the late 1950s that wages rose more rapidly when the unemployment rate was low, posits a trade-off between inflation and unemployment. When unemployment is low, and the labor market is tight, there is greater upward pressure on wages and, through labor costs, on prices.

The conceptual foundations of this relationship have been a subject of active debate, but for many decades, the relationship seemed well-supported by U.S. data. In the last two decades, however, the U.S. inflation rate has not been particularly high, even during periods of low unemployment. The recent data have led many to wonder whether the Phillips curve has weakened or disappeared. In Prospects for Inflation in a High Pressure Economy: Is the Phillips Curve Dead or Is It Just Hibernating? (NBER Working Paper No. 25792) Peter Hooper, Frederic S. Mishkin, and Amir Sufi examine why the Phillips curve relationship has not been evident in recent aggregate data for the United States.

The researchers study both inflation in consumer prices and inflation in wages. They test for a "price" Phillips curve using data on annual costs of goods and services, and for a "wage" Phillips curve using hourly earnings data. They allow for different relationships between inflation and unemployment in tight and in slack labor markets. Using a simple model that assumes a linear relationship between inflation and unemployment, and data from 1961 to 2018, they estimate that a one percentage point drop in the unemployment rate increased inflation by a mere 0.14 percentage points. However, when they allow for different effects of unemployment changes in tight and slack labor markets, they find that the estimated effect of a 1 percentage point unemployment decline on the inflation rate is about -0.32 percentage points when the unemployment rate is 1 percentage point below the natural rate, and -0.12 when it is 1 percentage point above it.

When examining data only from 1988 to 2018, the researchers see less evidence for a robust price Phillips curve. The linear and nonlinear slopes are both close to zero, consistent with the common view that the Phillips curve is flattening. However, the wage Phillips curve is much more resilient and is still quite evident in this time period.

The study points out that in the last three decades, the Great Recession notwithstanding, there has been less variability in the national economy than in prior decades, which makes it harder to detect the impact of unemployment on inflation. In addition, the Federal Reserve has tried to avoid labor market overheating as a way to stabilize inflation, thereby "anchoring" inflation expectations at a 2 percent inflation level and reducing the effect of unemployment fluctuations on price movements.

The researchers observe that state- and city-level data provide more variability in unemployment rates and are less influenced by federal monetary policy than the national figures. Therefore, they explore the relationship between unemployment and inflation at this level. They find a strong negative relationship between the unemployment rate's deviation from the state average and the rate of wage inflation. They also find evidence of a nonlinear price Phillips curve in city-level data.

The researchers point out that the relationship between inflation and the unemployment rate is a key input to the design of monetary policy. They note that the unemployment rate in the U.S. economy is currently near record lows, and they caution that they cannot predict whether inflation will rise in the coming years. However, they conclude that "Evidence that the price Phillips curve has been dormant for the past several decades does not necessarily mean that it is dead... it could be hibernating, and there is a risk of the Phillips curve waking up, with inflationary pressures rising in the face of an overheating labor market."

The employment effects of minimum wages

A brief summary of the state of play.

First, the evidence on the disemployment effect of minimum wages is contested, and there clearly are studies that find no employment effect – both in the United States and in other countries. However, the preponderance of evidence indicates that minimum wages reduce employment of the least-skilled workers. Earlier estimates suggested an ‘elasticity’ of about -0.1 to -0.2. Many estimates are still in this range, some are closer to zero, and some are larger. To be clear, some researchers may have reason to put more store in the types of estimates that tend to find no employment effects – typically the research designs that I have labeled ‘close controls’. I have indicated reasons I am somewhat skeptical of these designs, but also indicated that the jury is still out. More definitively, though, it is indisputable that there is a body of evidence pointing to job losses from higher minimum wages. Characterizations of the literature as providing no evidence of job loss are simply inaccurate.

Second, there are two kinds of changes in minimum wages about which we know a lot less. The first change is the adoption of much higher minimum wages – as is happening in the United States with serious movement toward a $15 minimum. There is a great deal of uncertainty about the employment effects of a $15 minimum wage. One thing we do know is that it would impact far more workers than the current minimum wage, especially in lower-wage states and lower-wage areas of most states. More speculatively, my sense is that the costs of a much higher minimum wage are likely to be understated by simply scaling up the effects based on employment elasticities in the existing literature, because the much higher share of workers affected will reduce employers’ ability to partially offset minimum wage increases by changes in margins other than employment.

The second kind of change about which we know relatively little concerns the introduction of a new minimum wage – like in Germany. There is some evidence from the introduction of a new minimum wage in the United Kingdom. Some of this evidence points to job loss, but the evidence is mixed. And, of course, the institutional setting is not the same.

From "The Econometrics and Economics of the Employment Effects of Minimum Wages: Getting from Known Unknowns to Known Knowns" by David Neumark, German Economic Review, Forthcoming.

The employment effects of minimum wages

There is a new NBER working paper out on The Econometrics and Economics of the Employment Effects of Minimum Wages: Getting from Known Unknowns to Known Knowns by David Neumark.

The abstract reads:

I discuss the econometrics and the economics of past research on the effects of minimum wages on employment in the United States. My intent is to try to identify key questions raised in the recent literature, and some from the earlier literature, that I think hold the most promise for understanding the conflicting evidence and arriving at a more definitive answer about the employment effects of minimum wages. My secondary goal is to discuss how we can narrow the range of uncertainty about the likely effects of the large minimum wage increases becoming more prevalent in the United States. I discuss some insights from both theory and past evidence that may be informative about the effects of high minimum wages, although one might argue that we first need to do more to settle the question of the effects of past, smaller increases on which we have more evidence (hence my first goal). But I also try to emphasize what research can be done now and in the near future to provide useful evidence to policymakers on the results of the coming high minimum wage experiment, whether in the United States or in other countries.

Employment effects of Chinese imports in the US

In 2013, Autor, Dorn, and Hanson published a paper in the American Economic Review showing that rising import competition from China has been an important contributor to the recent decline in the employment rate of working age population in the United States.

Exploiting variation in exposure to Chinese import across local labor markets (commuting zones) over 1990-2007, they find that Chinese import exposure caused a large reduction in manufacturing employment: a $1,000 per worker increase in import exposure over a decade reduces manufacturing employment per working-age population by 0.6 percentage points (their Table 3, column 6), explaining about 44 percent of the actual decline in manufacturing employment from 1990 through 2007. Furthermore, the negative employment shock by Chinese imports goes beyond manufacturing and exists for nonmanufacturing workers. To be more specific, import exposure to Chinese imports caused a substantial employment decline in both manufacturing and nonmanufacturing sectors for workers without college education; while for workers with college education, import exposure caused substantial job loss in manufacturing sectors but a statistically insignificant increase in employment in nonmanufacturing sectors (their Table 5, Panel B) (Feenstra, Ma and Xu 2017: 2).

Robert Feenstra, Hong Ma and Yuan Xu have written a comment, "The China Syndrome: Local Labor Market Effects of Import Competition in the United States: Comment" (pdf), on that paper. In their comment Feenstra, Ma and Xu report results that cut the total employment effect in half, and in some groups the employment effect becomes insignificant, by controlling for the US housing boom .

The comment's abstract reads,

We re-examine the findings by Autor, Dorn, and Hanson (ADH, American Economic Review 2013, 103(6)) on the impact of Chinese import penetration on U.S. local employment by taking into account the concurrent housing boom. The responses of total employment, unemployment, or not-in-the-labor force to import exposure fall by about one-half when controlling for changes in housing prices, and become statistically insignificant in a number of cases. Results across sectors are more subtle. Noncollege workers in the manufacturing sector continue to experience a reduction in employment after correcting for the ‘masking’ effect of the housing boom, but that reduction does not occur in the nonmanufacturing sector. For college workers, their employment in the nonmanufacturing sector even rises due to the China shock, which fully offsets their reduced employment in manufacturing during 2000-2007. Our results imply that the net reduction in total US employment due to Chinese import exposure was about 0.8 million workers, or less than one-half of that implied by the estimates in ADH (2013).

The economic consequences of labour market regulations

A recent working paper from the Penn Institute for Economic Research reviews the empirical evidence on the effects of a minimum wage on employment. And yes, they are still negative.

The paper is "The Economic Consequences of Labor Market Regulations"  (pdf) by Jesus Fernandez-Villaverde. The abstract reads,

What do we know about the economic consequences of labor market regulations? Few economic policy questions are as contentious as labor market regulations. The effects of minimum wages, collective bargaining provisions, and hiring/firing restrictions generate heated debates in the U.S. and other advanced economies. And yet, establishing empirical lessons about the consequences of these regulations is surprisingly difficult. In this paper, I explain some of the reasons why this is the case, and I critically review the recent findings regarding the effects of minimum wages on employment. Contrary to often asserted statements, the preponderance of the evidence still points toward a negative impact of permanently high minimum wages (Emphasis added).

The traditional view among economists of the effect of minimum wages is simple: if the minimum wage is set above the equilibrium wage you get unemployment. This core understanding of how the labour market functions has been attacked, most famously, by the publication of Card and Krueger (1994) and, three years later, by a companion book, Myth and Measurement (Card and Krueger, 1997). Card and Kruger looked a the case of an increase in the minimum wage that took place in New Jersey but not in the neighbouring state of Pennsylvania. On April 1, 1992, New Jersey increased the minimum hourly wage from $4.25 to $5.05. Pennsylvania, in contrast, kept the minimum wage at $4.25.

To most economists’ surprise, Card and Krueger documented a relative increase in employment in New Jersey of 2.75 full-time equivalent (FTE) employees per restaurant. In fact, there was even an absolute increase in employment in New Jersey and a drop in Pennsylvania (Card and Krueger, 1997, Table 2.2, p. 34). While, employment at the restaurants Card and Krueger surveyed in New Jersey went from 20.44 FTE employees per restaurant to 21.03, in Pennsylvania, it fell from 23.33 to 21.17

But as Fernandez-Villaverde notes the result has not aged well.

Card and Krueger’s results were sensational because they challenged a centuries-old understanding in economics. Also, their findings rationalized a policy intervention that has had strong political backing for almost as long. But sensational results invite close examination, and Card and Krueger’s findings have not held up to that torrent.

Refs.:

• Card, D., and A. B. Krueger (1994): “Minimum Wages and Employment: A Case Study of the Fast-Food Industry in New Jersey and Pennsylvania,” American Economic Review, 84(4), 772–93.
• Card, D., and A. B. Krueger (1997): Myth and Measurement. Princeton University Press.

If economic ignorance were a natural resource, our world would be paradise

You have to love a heading like that! It comes not from me but from the ever insightful Don Boudreaux at the Cafe Hayek blog. Boudreaux is commenting on the idea that the government subsidisation of low-skilled workers’ "housing, food, medical care, and transportation" enables employers of such workers to pay them less than some "true value" of their work.

There is an obvious question as to what this "true value' is.

Boudreaux writes,

The central economic point is this: the welfare programs to which Mr. Phelps alludes (with the possible exception of transportation subsidies) reduce the supply of labor and, thus, push wages up. Far from employers being subsidized by such welfare programs, employers of workers who receive these government benefits are obliged, as a result, to pay wages that are made artificially high.

But to show just how deeply confused this Mr. Phelps is, let’s pretend that he’s correct to insist that welfare programs artificially reduce wages. Mr. Phelps then asserts that “Failure to pay a living wage gives consumers artificially low prices and increases corporate profits.” Because nearly all employers of low-skilled workers operate in intensely competitive industries such as retail and food service, workers’ artificially low wages would indeed result in artificially low prices for consumer goods, but not in increased corporate profits. The ability to hire workers at artificially low wages would attract new entrants into these markets, as well as cause existing firms to expand their outputs, until the rate of profit earned by employers of these workers is no higher than it would be if wages were higher. That Mr. Phelps is oblivious to this reality is sufficient reason to dismiss his economic analysis.

While I agree with Boudreaux's point, I do wonder just how large the effect is. Of all the things the government does to stuff-up the labour market is this a big player?

Protection and job losses

At the Cafe Hayek blog Don Boudreaux makes the point that protection does not stop job losses.

You are correct that under a regime of free trade some people, through no fault of their own, lose jobs. You are also correct that such experiences are unpleasant. But protectionism does not stop job losses from occurring. Even if Uncle Sam were to completely shut the American economy off from global markets, job losses would still occur. American consumers would still change their spending patterns such that goods and services that were in high demand yesterday would be in lower demand today. Entrepreneurs would still experiment with new products and with new, labor-saving methods of production and distribution. Economic churn would still happen, complete with its unavoidable job losses.

The difference would be that, being denied access to the creative insights and productive efforts of 95 percent of the world’s population as well as to the bulk of the world’s resources, we’d all be much poorer.

What changes in trade policy do is change who has jobs, not the total number of jobs. Change trade policy and you just move jobs around the economy without having much effect on the total number of those jobs. As Paul Krugman has written,

It should be possible to emphasize [...] that the level of employment is a macroeconomic issue, depending in the short run on aggregate demand and depending in the long run on the natural rate of unemployment, with microeconomic policies like tariffs having little net effect. Trade policy should be debated in terms of its impact on efficiency, not in terms of phony numbers about jobs created or lost.

But the other point made by Boudreaux is also important, protection makes us poorer. The more we protect, the more we have to do things we are (relatively) bad at doing and the less time we spend on doing things we are good at doing. In the process we make ourselves worse off. It is better for us to concentrate on what we are good at producing and trading that for the things we are bad at producing with people who are good at producing it. As Krugman notes efficiency is what matter and by doing things we are bad at we become less efficient, not more.

How computer automation affects occupations: technology, jobs, and skills

James Bessen deals with this issue in a new column at VoxEU.org. A common argument you see with regard to computers and employment is that computer automation leads to major job losses. A modern version of the Luddite story. Bessen argues that this line of argument, however, ignores the dynamic economic responses that involve both changing demand and inter-occupation substitution. Using US data, he explores the effect of automation on employment growth for detailed occupational categories. Computer-using occupations have had greater job growth to date, while those using few computers suffer greater computer-related losses. The major policy challenge posed by automation is developing a workforce with the skills to use new technologies.

Bessen looks at estimates of employment demand growth.

Taking these considerations into account, I estimate a simple model of occupational demand across industries that allows for changing demand and inter-occupation substitution within industries. As my key independent variable, I measure the extent of computer use by workers in each occupation and industry. These data come from supplements to the Current Population Survey. I assume that occupations that use more computers will have a higher degree of task automation, all else equal. The dependent variable is the relative growth of employment in occupation-industry cells.

The estimates contradict popular assumptions about the impact of computer automation. First, computer-using occupations tend to grow faster, not slower. At the sample mean, computer use is associated with a 1.7% increase in occupational employment per year. In other words, the bank teller example may be typical rather than exceptional.

Second, there is a strong substitution effect between occupations. Occupations tend to have declining growth to the extent that other occupations in the same industry use computers. That is, the story is not about machines replacing humans; rather it is one of humans using machines to replace other humans, as graphic designers with computers replaced typesetters.

The substitution effect largely offsets the growth effect. Counting both, at the sample mean, computer use is associated with positive employment growth but the effect is small, 0.45% per year. This association is not necessarily causal—perhaps some other factor caused computer-using occupations to grow. But this finding does show that computer automation is not associated with major job losses.

The last sentence is the takeaway result. Computer automation is not associated with major job losses meaning our modern day Luddites have less to worry about than they first thought.

A downside to increasing computer use is that while the evidence suggests that computers are not causing net job losses, low wage occupations are losing jobs, likely contributing to economic inequality. The policy problem is how to get the workers effected the new skills they need in order to transition to new, well-paying jobs. Developing a workforce with the skills to use new technologies is the real challenge posed by computer automation.

Why long-term unemployment matters

Previously I have noted papers that deal with the effects of long-term unemployment, see here, here, here and here. Now Tim Harford is writing about it at the Financial Times.

Harford says,

A recent Brookings Institution research paper by Alan Krueger (a senior adviser to Barack Obama during the recession), Judd Cramer and David Cho examines this discomfiting thesis in greater depth. The researchers conclude that people who have been out of work for more than six months are indeed marginalised: employers ignore them, bidding up wages if necessary to attract workers from the ranks of the short-term unemployed.

I’ve written before about an experiment conducted by a young economist, Rand Ghayad. He mailed out nearly 5,000 carefully calibrated job applications, using a computer to tweak key parameters. He found that employers were three times more likely to call an applicant with irrelevant but recent employment experience, than someone who had relevant experience but had been out of work for more than six months. Long-term unemployment had become a trap.

In Ghayad’s experiment, the long-term unemployed were identical in every other way to other applicants. In reality, of course, it may be that people also become demotivated after a long spell of looking for work. The “benefits culture” at work? It seems not. Earlier research by Krueger and Andreas Mueller tracked job hunters over time and showed them becoming ever less active in the job market – and ever more depressed. They could not rouse themselves, even when unemployment insurance payments were about to expire. It wasn’t that the people joined the ranks of the long-term unemployed because they were demotivated to start with: the long-term unemployment came first, and the unhappiness and the lack of drive came later.

The important question, not answered above, is Why do employers ignor the long-term unemployed? Loss of human capital? Do employers see long spells of unemployment as a signal of some bad characteristic about the person? Or something else? Or, as Harford puts it, are

the long-term unemployed [...] people who are not very different to the rest of us – merely unluckier.

Without this question being answered its not clear how to deal with the problem.

Unemployment and qualifications

Previously I have noted papers that deal with the effects of long-term unemployment on re-employment, see here, here and here. These papers show that being unemployed for a long period of time lowers your probability of get a another job. Do does long term unemployment affect different levels of qualification differently?

There is a paper, by Alexander Mosthaf, forthcoming in the Scottish Journal of Political Economy which asks Do Scarring Effects of Low-Wage Employment and Non-Employment Differ BETWEEN Levels of Qualification? The paper finds that for those with low-qualifications being in a low-wage job incurs a lower risk of future unemployment when compared to being unemployed. But for those with higher qualifications the risk of unemployment is no different given that you are in low-paid worker or unemployed.

The idea underlying the results is that if employers interpret the job search behaviour of workers as a signal for their future productivity - more job search implies lower productivity, if you were high productivity you would have gotten a job quickly - taking up a low-paid interim job as well as being unemployed is associated with negative signalling effects. As the incidence of unemployment and low-wage employment is more typical for workers with lower qualifications, the paper argues that negative signals may be lower for workers with low qualifications and stronger for workers with high levels of qualification.

The paper's abstract reads:

This study investigates how the effects of low-wage employment and non-employment on wage prospects vary depending on qualification. Based on theories on signalling effects, human capital and job search, we discuss why there may be heterogeneity in state dependence in both labour market states. We find that episodes of low-wage employment incur a significantly lower risk of future non-employment than episodes of non-employment for low-qualified workers. In contrast, for workers with a middle or high level of qualification the risk of non-employment is not significantly different when being low-paid instead of not employed.

Do employers use unemployment as a sorting criterion when hiring?

An important question since if they do and you are unemployed then you are likely to remain unemployed. This leads to a group of long term unemployed who find it almost impossible to become re-employed. I have discussed this idea before, see for example here and here.

Now addition evidence on the issue comes from a paper in the most recent issue of the American Economic Review (Vol. 104, Issue 3 March 2014). The paper is "Do Employers Use Unemployment as a Sorting Criterion When Hiring? Evidence from a Field Experiment" by Stefan Eriksson and Dan-Olof Rooth. The abstract reads:

The stigma associated with long-term unemployment spells could create large inefficiencies in labor markets. While the existing literature points toward large stigma effects, it has proven difficult to estimate causal relationships. Using data from a field experiment, we find that long-term unemployment spells in the past do not matter for employers' hiring decisions, suggesting that subsequent work experience eliminate this negative signal. Nor do employers treat contemporary short-term unemployment spells differently, suggesting that they understand that worker/firm matching takes time. However, employers attach a negative value to contemporary unemployment spells lasting at least nine months, providing evidence of stigma effects. (Emphasis added)

So again we see that the longer you are unemployed the less likely you are to become employed.

Economic freedom and labour market conditions

An interesting new paper on Economic Freedom and Labor Market Conditions: Evidence from the States by Lauren R. Heller and E. Frank Stephenson. The abstract reads:

Using 1981–2009 data for the 50 states, this article examines the relationship between economic freedom and the unemployment rate, the labor force participation rate, and the employment-population ratio. After controlling for a variety of state-level characteristics, the results from most specifications indicate that economic freedom is associated with lower unemployment and with higher labor force participation and employment-population ratios.

The conclusion to the paper states,

Our findings have clear implications for policymakers. Adopting policies that increase economic freedom by one point in the EFNA index would reduce the unemployment rate by as much as 1.3 percentage points. Likewise a one point increase in economic freedom would increase the labor force participation rate by up to 1.9 percentage points and the employment-population ratio by as much as 2.3 percentage points. Moreover, the estimation results for the EFNA sub-components offer policymakers some guidance about the most beneficial ways to improve state EFNA ratings. The effect of area 1 (size of government) is generally larger than the effect for areas 2 or 3. Hence, our results suggest that reducing the size of government would have the largest effect on labor market outcomes.

So economic freedom does decrease unemployment and increase the labour force participation rate but reducing the size of government is the big winner for reducing unemployment.

Unemployment is bad for employment 2

I have made the point that the longer you are unemployment the less likely it is that you will become employed before but more evidence confirming it comes from this article in the Washington Post:

Here’s one big reason why America’s unemployment crisis may be here to stay. Thanks to the lasting effects of the recession, there are currently 4.7 million workers who have been out of work for at least 27 weeks. And new research suggests that employers will almost never consider hiring them.

Matthew O’Brien reports on a striking recent experiment by Rand Ghayad of Northeastern University. He sent out 4,800 fake resumes at random for 600 job openings. And what he found is that employers would rather call back someone with no relevant experience who’s only been out of work for a few months than someone with more relevant experience who’s been out of work for longer than six months.

In other words, it doesn’t matter how much experience you have. It doesn’t matter why you lost your previous job — it could have been bad luck. If you’ve been out of work for more than six months, you’re essentially unemployable. Many companies won’t even consider you for a job.

One obvious question this raises is, Are companies irrationally discriminating against the long-term unemployed or do they have good reasons for screening out these applicants? The Washington Post article writes,

Privately, many employers worry that someone who’s been out of work for six months “may have outdated skills, or may be a short-timer who is desperate enough to take any work now but will leave when something better comes along.”

One worry with this is that the current cyclical unemployment problems could become structural and very long-lasting.

Unemployment is bad for employment

No really!

The following is a summary by Lester Picker, from the latest (March 2013) NBER Digest, of the findings of the paper, Duration Dependence and Labor Market Conditions: Theory and Evidence from a Field Experiment (NBER Working Paper No. 18387) by Kory Kroft, Fabian Lange and Matthew J. Notowidigdo. Lester writes,

According to a recent report by the Congressional Budget Office, long-term unemployment may "produce a self-perpetuating cycle wherein protracted spells of unemployment heighten employers' reluctance to hire those individuals, which in turn leads to even longer spells of joblessness." Policymakers and researchers alike tend to believe that this adverse effect of a long spell of unemployment undermines the smooth functioning of the labor market and entails large social costs. Economists refer to the phenomenon as "negative duration dependence."

In Duration Dependence and Labor Market Conditions: Theory and Evidence from a Field Experiment (NBER Working Paper No. 18387), authors Kory Kroft, Fabian Lange, and Matthew Notowidigdo confirm that the likelihood of receiving a callback for a job interview sharply declines with unemployment duration. This effect is especially pronounced during the first eight months after becoming unemployed. Their estimates suggest that this effect is quantitatively important, and that duration dependence is stronger when jobs are relatively abundant. These results imply that employers statistically discriminate against workers with longer unemployment durations and that employer screening plays an important role in generating duration dependence.

To study duration dependence, the authors submitted fictitious resumes to real, online job postings in each of the 100 largest metropolitan areas in the United States, and then tracked "callbacks" from employers for each submission. In total, they "applied" to roughly 3,000 job postings in Sales, Customer Service, Administrative Support, and Clerical job categories, submitting roughly 12,000 resumes. The resumes they created characterized the "applicant's" employment status and, if unemployed, the length of the current unemployment spell, which ranged from 1 to 36 months and was randomly assigned. As a result, this experiment directly uncovered duration dependence arising through employers' beliefs about unemployed workers.

In short, the longer you are unemployment the less likely it is that you will become employed.