What Hawaii can learn from America’s leading minimum-wage researcher
from Grassroot Institute of Hawaii, March 31, 2022
A lot of people seem to think there is wide among economists about the potential ill effects of a higher minimum wage.
But, no, there actually isn’t, says David Neumark, America’s leading authority on the minimum wage.
Neumark was the guest this past Monday, March 28, 2022, on Keli’i Akina’s “Hawaii Together” program on ThinkTech Hawaii, and he talked about what he has learned from having read almost every study there is on the minimum wage.
An economist at the University of California, Irvine, and author of “The Minimum Wage,” published by The MIT Press, Neumark said, “The first thing minimum-wage laws do is to raise wages for some workers. I mean, that’s what the laws do, and we tend to comply with these things in the U.S. So wages are pushed up.
“Now, as far as the advocates are concerned, end of story. So that’s all good. It’s a free lunch, apparently, and people who make low wages now earn more. But what the research shows — I think most of the research shows — is that there are other consequences.”
Neumark said he and a graduate student recently surveyed the authors of “pretty much all” of the U.S. minimum-wage papers published since 1992 and found that “almost all” their conclusions were “negative.”
Sure, he said, “There were some outliers. This is not laboratory science where you replicate something under identical conditions and you get the same answer. It’s social science. People study minimum-wage effects differently. [But] actually, there’s a very strong preponderance of the evidence that says there is job loss from a higher minimum.
Meanwhile, of “all those extra dollars going to workers who are now making more because of the higher minimum,” only a “very small share goes to poor families; even a smaller share goes to what we call extremely poor families, families at the poverty line.”
In fact, he said, nearly half of the wage increases go to families in the top half of the income distribution, … because there’s a lot of young workers who are on minimum wage, just because they’re young and they don’t know anything yet. They’re not going to be poor later. They might be an MIT engineer later on, but they have to have some job in high school making coffee, … so they get the benefits when you raise the minimum.”
“It’s not clear we’re accomplishing anything from a policy perspective, and it’s even conceivable that … that the people paying the higher prices to pay those higher-income teenagers higher minimum wages have lower incomes than they do.”
Neumark said none of this means politicians should not increase the minimum wage.
“It just means there’s a trade-off, and you should think about the costs and benefits, and not just the benefits.”
Other topics Neumark discussed include his views on the infamous 1993 David Card and Alan Krueger minimum-wage study, how a minimum-wage increase could affect Hawaii, and what policies would better help low-income or poor families get ahead.
See the whole half-hour interview by clicking on the video below. A complete transcript follows.
* * * * *
3-29-22 David Neumark with Keli‘i Akina on “Hawaii Together”
Keli’i Akina: Aloha, everyone and welcome to “Hawaii Together” on the ThinkTech Hawaii broadcast network. I’m your host, Keli’i Akina, president of the Grassroot Institute.
Hawaii lawmakers are poised to hike the minimum wage. But some say that could actually hurt the people it’s supposed to help. Now, there are many arguments by those who have opposing points of view. But the research is in: The minimum wage mostly hurts the people it’s intended to help. That’s what we found.
And that’s the finding of David Neumark, economist and co-director of the Center for Population, Inequality, and Policy at the University of California, Irvine. He’s also the co-author of “Minimum Wages,” published by the MIT Press.
On today’s program, I’m going to talk with Neumark about his comprehensive review of evidence on the effect of minimum wages on employment, skills, wages and income distributions, and also on the longer-term labor market outcomes.
But most notably, we’re going to take a look at his conclusion that, ultimately, minimum-wage laws aren’t a good policy tool.
Please welcome to the program, David Neumark. David, thank you so much for joining us today all the way from California.
David Neumark: Thanks, Keli’i. Happy to be here.
Akina: I’m so glad to have you. I’ve been looking forward to chatting with you. Tell me a little bit first about your own background and how you got into the field of researching minimum wages.
Neumark: It’s a long story. Well, it’s not a long story, it goes back a long way. It’s a little funny though. I was a grad student, I finished in ’87, my Ph.D. at Harvard. Sometime — it must have been around ’84, when we first started taking our courses in our fields — I actually remember my adviser coming in and saying on the first lecture, it was kind of, “Well, what do labor economists do? What might you do research on?”
He said, “Let me tell you one thing you shouldn’t waste any time on is the minimum wage,” and he plopped down on the table — and it’s going to be generation-specific whether you know what I’m talking about — what looked like something as thick as the New York phone book, which used to be like 5 inches …
Neumark: It was about that high. It was the comprehensive report by the U.S. Department of Labor, by something called the Minimum Wage Study Commission. They had commissioned papers by everyone under the sun — who [were], at that time, probably eight or nine years older than me or more — who worked in labor. He said, “Look, we know everything. Forget about it.”
Then, my first job was at the Federal Reserve Board, and one of the things that happened when you work there is, they obviously keep you very up on the news — right? — [because] you’re working at the Fed.
We would get this thing called the Daily Labor Report, which was a compendium of state laws, court rulings, anything that might affect labor policy. And my co-author, Bill Wascher, and I noticed that a bunch of states started raising the minimum wages. Previously, we had pretty much only had a national minimum wage.
Akina: Well, David, you mentioned Bill Wascher; he’s the co-author of your book, “Minimum Wages.”
Akina: How extensively have you researched this topic compared to other people across the country?
Neumark: I’m pretty sure I’ve written more papers than anybody else. (laughs)
Akina: That’s right. We really appreciate that, and we’re glad to be able to pick your brain today.
Overall, just the banner question I have for you, what are the effects of minimum wage laws? We’ll dive into it a little more deeply as we go on, but just in general.
Neumark: Sure. So, first of all, let me say, I think my view is a little more nuanced than your introduction …
Akina: All right.
Neumark: … because this is the point of my research — but something that is often missed by advocates, as well as, frankly, I would say, by sort of the blanket opponents — and that is that the effects are heterogeneous.
So, obviously, the first thing minimum-wage laws do is to raise wages for some workers. I mean, that’s what the laws do, and we tend to comply with these things in the U.S. So wages are pushed up.
Now, as far as the advocates are concerned, end of story. So that’s all good. It’s a free lunch, apparently, and people who make low wages now earn more. But what the research shows — I think most of the research shows — is that there are other consequences.
There is some job loss that stems from the minimum wage going up, which doesn’t mean that everybody is hurt, because obviously, some people still keep their jobs and earn higher wages. And if your wage has gone up, and A) you have your job, and B) your hours haven’t been cut, well, you’re happier, right? But there are offsetting effects, and there are some trade-offs.
My point is always based on the research that you’ve got to think about these trade-offs, and you can’t pretend they don’t exist.
The other key point, I think, that is slightly more subtle — but we can get into it in more detail — is that the minimum wage is a really blunt, by which I mean not precise, a really blunt tool to try to affect the underlying problem.
The underlying problem is low family income. That’s what we care about. I don’t care if … I don’t know if you have a teenage son or not, but I don’t really care if your teenage son works at a low-wage job. That is not a policy problem that any business owner or taxpayer should worry about.
What I am worried about is a single mother who works at a very low-wage job and how many resources she has available for her and her kids between what she earns at work and what she gets from other government programs. So that’s the problem of poverty or low income, which we define at the family level.
And the simple fact is — and then I’ll stop — is that while obviously being a low-wage worker and being in a low-income family are related, they’re not very related.
As a result, when you try to increase incomes of low-income families by raising the minimum wage, you give a lot of benefits to people who aren’t in low-income families, like the upper-middle-class teenager, and you don’t do anything for the people who don’t work very much — right? — because you don’t get anything from the minimum wage unless you work.
Akina: I think you’ve explained it very well, that the minimum wage is a very blunt tool, and probably the wrong tool, for trying to solve the problem of overall low wages.
You spoke about the fact that the effects of minimum-wage laws are not homogenous. They’re different for different people in different categories. So, specifically, how do minimum-wage laws affect the poor?
Neumark: So, let’s just go back to heterogeneous effects for a minute. In the aggregate, some people’s wages are pushed up, and then there’s some job loss and some hours reduction as well.
We don’t actually have great evidence on how the employment effects or the hours effects differ across the income distribution, that is, for a low-wage worker and a poor versus non-poor family. Which is sort of mystifying, because I think we should.
It just turns out, and this is a reflection of my point, that low-wage work and low-income families are by no means the same thing. That when you start slicing the data really finely, to divide it by both wage levels — who was really affected by a high minimum wage and very low-income families, unless you have really huge samples — you don’t have many people left.
But what you can do is say: What does a higher minimum wage do to the poverty rate or some other measure of the instance of low income? The poverty rate is one we use. It’s not necessarily the ideal one, but it’s sort of an agreed-upon one.
And the answer there is — and I would say, there’s been more research since my book — but essentially, you can’t find any strong evidence, I think, that minimum wages reduce poverty. It’s probably largely a wash. Whereas, obviously, a policy that is intended to reduce poverty, the instance of low income, ought to have an effect.
I mean, look, if we take TANF [Temporary Assistance for Needy Families], the current incarnation of our welfare program in the U.S., [which] you have in Hawaii as well, and raise the amount of income we gave to families that were eligible for TANF, we would clearly reduce poverty and increase incomes of low-income families.
Why? Because the program is targeted to low-income families. You don’t get it if you’re not a low-income family.
If you do simple back-of-the-envelope calculations — the most recent one of these I’ve seen is from a grad student of mine from a few years ago — you say, “Let’s imagine we raised the minimum wage.” I don’t have a calculation for Hawaii or for an $18 minimum, but he was considering sort of a $15 minimum for the U.S. as a whole. And you say, “Let’s assume there’s no response. Everyone keeps their job. No one’s hours fall.” Really simple. Your wage simply gets swept up to the minimum.
There’s a big increase in the wage bill. Employers are paying low-wage workers more, and now you simply say, “OK, all those extra dollars going to workers who are now making more because of the higher minimum, how are they distributed across families at different parts of the income distribution?”
And a very small share goes to poor families; even a smaller share goes to what we call extremely poor families, families at the poverty line. It varies slightly, but about 15%, which you use the data. In contrast, about 40% goes to families with above-median income. Median is the income level of which half are below, half are above.
Median income is not high income — it’s still a fairly low income. But the point is, nearly half the benefits go to families in the top half of the income distribution, and it’s exactly because there’s a lot of young workers who are on minimum wage, just because they’re young and they don’t know anything yet. They’re not going to be poor later. They might be an MIT engineer later on, but they have to have some job in high school making coffee, as I said. So they get the benefits when you raise the minimum, employers have to pay them more.
It’s not clear we’re accomplishing anything from a policy perspective, and it’s even conceivable that, yet another side of the equation is, that the people paying the higher prices to pay those higher-income teenagers higher minimum wages have lower incomes than they do.
Akina: David, your research findings are based on having reviewed a vast amount of studies that are out there now, and there’s tremendous research. My question is this: Why is there still so much economic debate about the minimum wage? Why do people come to so many different conclusions about it? What do you think is going on here?
Neumark: Well, let me just first attest to two premises. Is there a lot of continuing research on the minimum wage? Yes, there is. We’ll come back to why in a second. You said, “Why do people come to so many different conclusions?” They actually don’t. That’s the surprising thing.
I have a recent student of mine, Pete Shirley, and we wrote a paper which just got accepted for publication. We were asking what seems like a very simple question for a researcher to ask, which is: What do all the other papers actually say?
Because you often get this summary of literature. Oh, half the evidence finds job loss and half doesn’t, or no one can agree, or estimates are all over the map. And I was kind of interested in actually systematically tabulating that all up and seeing if that’s true or not. Which you might think I should be able to do because I study this stuff all the time. But, you know, someone sends me a paper, someone else doesn’t send me a paper; who knows what I’m seeing?
We do this very systematic survey and we actually surveyed the authors of all the U.S. minimum wage papers, pretty much all of them, published since 1992. I don’t have the picture with me — I didn’t think to prepare it for the talk, for this discussion — but we kind of plot the distribution of estimates: almost all negative.
There’s some outliers, sure. This is not laboratory science where you replicate something under identical conditions and you get the same answer. It’s social science. People study minimum-wage effects differently. Actually, there’s a very strong preponderance of the evidence that says there is job loss from a higher minimum.
Again, that doesn’t mean you shouldn’t do it. It just means there’s a trade-off, and you should think about the costs and benefits, and not just the benefits.
Akina: I like the fact that you point out that it’s a trade-off. That it’s not either-or. What, if you could summarize for us, are some of the fundamental downsides of imposed minimum wages?
Neumark: I think it’s most useful to go back to the economic model. How do we teach? How do we think about it? How do — I just finished teaching my undergrad labor course — how do we teach this?
The standard story, which I think is largely borne out of the data, is the following: The minimum wage goes up. Government raises the minimum wage — federal government, state government, city government now. That, of course, forces up the wage of people who were previously making less.
That makes that input — call it low-skilled labor — relatively more expensive compared to the other inputs firms use, which can be materials, physical capital and. perhaps most importantly, more skilled labor, labor whose price isn’t affected directly by the minimum wage.
So that leads firms to substitute away from the labor that’s now gotten expensive. Just like if we tax cigarettes, people smoke less. Same idea. That inevitably raises the cost of production for firms, which raises prices. And it’s that effect — prices go up, consumers demand less — that leads to an additional source of employment reduction.
So the fundamental downside is that we have a type of worker, namely the low-skilled labor, the low-skilled workers, the ones who are affected by minimum wage laws, and we are trying to help them — although I would argue we should be trying to help poor families, and some will have low-wage workers and some won’t. But, we’re trying to help low-wage workers.
But what a minimum wage does is tax the use of low-skill labor. That’s all. From the point of view of a business as a business, it’s no different if I say, “Keli’i, you’ve got to pay that guy $7 more, or you’ve got to pay me a $7 tax when you hire that guy.” Absolutely no difference.
And what we know taxes tend to discourage the use of the things you’re taxing unless it’s something you absolutely can’t do without. So as long as there are substitutes for lower-skilled workers, a higher minimum wage effectively will — the fundamental source of the downside is it makes using that labor more expensive, and that’s what creates the trade-off at the end of the day.
Akina: That makes sense, David. We’re going to take a quick break right now. I say to our viewers, we will be right back, so don’t go away. We’re with David Neumark. We’re talking about minimum wage. We’re on the “Hawaii Together” program on the ThinkTech Hawaii broadcast network. We’ll be right back.
Akina: Thanks for coming back. I’m Keli’i Akina, on “Hawaii Together,” on the ThinkTech Hawaii broadcast network. My guest, David Neumark, is a co-author of “Minimum Wages,” published by MIT Press. We’re talking about some of his findings he has produced in the last several years on minimum wage. We’re going to go right back to it.
David, I’ve got a question for you about one of the major studies that used to be cited all the time. It was produced in the 1990s by David Card and Alan Krueger. It supposedly showed that the minimum wage doesn’t always cause job loss. What has your research found about this quite-often-cited study?
Neumark: First of all, it is often cited, and it’s not just that it used to be cited; it’s still cited a lot. It was written in 1992, so, you know, it’s not as old as me, but it’s getting old.
I want to say a couple things. First of all, what people often forget, and I’m not always sure it’s unintentional on the part of advocates, is the study didn’t just show that the minimum wage didn’t reduce employment. It actually found whopping large positive effects on employment. People often forget that. I’m not cherry-picking here. Go back to the paper.
You know, we argue, in the literature, about the employment effects, about what we call elasticities of around minus 0.2, which means that if minimum wage goes up by 10%, low-skilled employment falls by 2%. They had elasticities of about positive 0.7, which means you raise the minimum wage 10%, you raise employment — in their case, fast-food restaurants they were studying — by 7%, which is patently absurd. It’s not possible. And there’s essentially no research to support that.
So that was written in ’92. It sort of took labor economics and the policy world by storm. I never quite understood why. Without regard to the quality, it was a study of one minimum-wage increase, in one industry — fast food — in one pair of states, New Jersey and Pennsylvania.
It was never clear to me why you would draw really strong policy conclusions from one paper, especially when there’s lots of other papers. (laughter). Yeah, you might have had differing opinions on the quality, but there are lots of other good papers that found very different effects.
That doesn’t mean we should ignore new studies. It doesn’t mean we should ignore new evidence. But we should proceed slowly. When a study is surprising, there’s often a reason it’s surprising. Sometimes it’s right and we were all wrong before, but we should always give some weight to the alternative: that it’s surprising because it’s wrong.
So in this case, Wascher and I went back to not the data, because we wanted to collect better data. They had surveyed managers at restaurants and asked them a very vague question, which literally, I don’t know the exact quote, but pretty close, was, “How many people work in your store?”
They did that twice. Before and after the increase in the same period in the state, Pennsylvania, where the minimum wage didn’t increase. That’s a really strange question to ask a fast-food restaurant manager.
If I went to the New Jersey Turnpike, where McDonald’s is open 24 hours a day, 7 days a week, and I asked you, “How many people work in your store?” do you say, “Oh, there’s 130 workers on the payroll to cover our 24/7 shifts,” or do you say, “Eight,” because there’s eight people here right now? I have no idea.
And then you come back eight months later and ask a different manager, whoever answers the phone, the same question, and who knows what they tell you that time?
We actually went back and got the payroll records from the restaurants back in the day of computer paper — you’ll remember this, with the holes down the side. They mailed it off to us. I actually have it in my file cabinet, believe it or not. Went through the data and found modest, small effects, which is kind of what you get in most of the research literature.
Neumark: So the study was based on flawed data. I think most researchers, if they find a really surprising result that really overturns, or may overturn, a lot of other studies, they sort of say, “This is really interesting. We should think a lot more about this.”
I think the problem with this study was, and it wasn’t so much the authors, but others, with skin in the game, just said, “This is the be-all and end-all, and forget everything that came before it, and frankly, everything that came after it.”
Akina: And that’s really why your approach worked so well in terms of doing a comprehensive review of the studies that are out there.
Let me bring this home to us here in Hawaii, where our Legislature is in session now. It’s looking at a couple of bills. HB2510 would increase our minimum wage in Hawaii to $12 per hour this year, and gradually increase it to $18 by 2026.
So what do you think the downsides are that could occur with such a policy? And do you think that delaying the $18 amount until 2026 reduces the harm that could be caused, particularly due to inflation?
Neumark: OK, a lot of questions I want to come back to. One thing is, just before you asked that question, you talked about me doing reviews of the evidence. Most of my work is original papers on minimum-wage effects. The book has a lot of reviewing, of course. My studies, they don’t all find strong evidence of negative effects. Sometimes they do. Different approaches, weaker, stronger. I think the consensus to the evidence is you reduce jobs.
To the Hawaii question. First caveat is: I’ve been to Hawaii, and I love it. I don’t know much about its economy other than whatever is true of other economies. So I don’t have any strong reason to think that minimum-wage effects are very different in Hawaii than elsewhere.
I will say, I think there’s one potential qualifier to that, and that is: Remember what I said before, that what kind of gets the ball rolling on the job loss is that the higher minimum wage pushes up prices, and that reduces demand, and that’s one of the reasons employers use less labor.
The other is, you can substitute away from low-skilled labor, right? Now, what that means, there’s two implications to that — well, there’s one implication to the questions: Minimum-wage effects will be less severe in terms of job loss for the low-skilled, where two things are true. One is, there’s not as much scope to substitute towards low-skilled labor, and the other is, where consumers are not sensitive to prices.
OK, so it’s conceivable, just to raise the possibility that — I know the tourism industry, I know that’s not the whole Hawaiian economy, but I know it’s obviously a disproportionate share relative to other places, for obvious reasons, right? It’s beautiful.
You know, the tourism industry may be an industry where there’s fewer substitution possibilities. I can walk into a McDonald’s, there’s a lot of capital back there that has replaced workers. Very capital-intensive production of food. It’s not clear. Now, you can reduce service in the tourist industry, but… maybe you can’t reduce it nearly as much. You can’t automate. Yes, you can have self-check-in, and this, that and the other, but there may be some limited scope for that.
The other is people may be somewhat insensitive to the price for the same reason that I’ll pay a lot more for a beer in an airport than I will anywhere else. Because I’m in the airport. I’m stuck. I can’t go anywhere.
To the extent that Hawaii has a unique idiosyncratic product — which to some extent it does, there aren’t a lot of great substitutes — you know, demand may be somewhat less sensitive. The tourism industry is one where one might expect somewhat weaker value; the direction is the same.
I’m just saying they may not be as pronounced as they are for all the other parts of the economy. And obviously, there’s a lot of other people in Hawaii doing a lot of things other than tourism, and I think those sectors of the economy are probably no different from anywhere else.
If I’m running a restaurant for locals, I might try to economize. If I’m running a Starbucks, I might make the line longer. I might automate. If I’m running a factory, obviously, there’s more scope for automation.
Akina: David, many people point to the fact that employment in Hawaii increased — or so they say — after the state minimum wage was increased in 2015. What are your thoughts about that?
Neumark: Well, I don’t know the facts. We probably don’t need “or so they say” on whether employment increased or not. We must know that. So let’s take it as true. [But] that’s not how you estimate the effects of policy.
I actually had a question like that on my labor-demand midterm a month ago. An example, like you cite, and somebody says, “Ahh, that proves minimum wage doesn’t reduce employment.” Good or bad argument?
The reason it’s a bad argument and bad evidence is — let’s go back to the Card and Krueger paper. The premise of the Card and Krueger paper, the idea was a good one, right? If you just look at one economy, call it Hawaii or call it New Jersey’s fast-food industry, and just see the minimum-wage change and say, “Oh, what happened to employment?”, you don’t learn anything because a lot of other stuff happens to the economy.
The minimum wage is not the cause of the business cycle. There are far, far bigger influences on the economy than the minimum wage.
So, if Hawaii was booming in 2015 — and a lot of states were booming in 2015; I’d be surprised if Hawaii wasn’t also was — and we happened to raise the minimum wage and you might have seen employment go up, the question is:, What would happen absent the minimum wage?
Now, of course, I can’t see Hawaii in 2015 with and without the minimum wage. I can only observe whatever was actually the case. So the beauty of the Card and Krueger paper is not just —people have been doing this for decades before they wrote their paper — is to compare two economies where we think developments are similar, one in which the minimum wage goes up, and one in which it doesn’t.
So Card and Krueger did New Jersey, and kind of Eastern Pennsylvania, that borders New Jersey. It may not be a good — we call this the counterfactual, what would have happened absent the policy change? We can only estimate that in a real experiment — a vaccine experiment, with a treatment and control group. There, I know exactly what would happen if you got the vaccine and if you didn’t.
In social science, we don’t have that. So one would have to find a comparable economy to Hawaii, if one exists, and say, “What would have happened otherwise?” I think you’re stuck with the evidence from the other states, which are a lot more similar.
Akina: David, we’ve got just a quick minute. I want to come up for air and go back to the original question about helping the poor. What’s a better way to help the poor than raising minimum wages?
Neumark: I’d say two things. One is: Actually create a program that targets benefits based on being poor or low income. You might have a somewhat looser definition of the word “poor.” The second is — and this is my own perspective, and maybe it’s more partly philosophical, partly economic — think about something that actually encourages work rather than discourages work, because we know that people’s earnings and wages grow with labor market experience.
I am a huge fan of the earned income tax credit. We have a federal earned income tax credit. Many states do. I think Hawaii does, but I should have looked that up before this interview.
Akina: On that note …
Neumark: Sorry. It subsidizes work and it targets benefits very well.
Akina: OK, very good. On that note, maybe we’ll have you back someday and talk about earnedincome tax credit. but thank you, David, for being on the program today. I appreciate it. Everyone, you’ve been watching “Hawaii Together,” on the ThinkTech Hawaii broadcast network. I’m Keli’i Akina, with the Grassroots Institute. Until next time, we bid you all aloha.