David R. Henderson  

Minimum Wage Not Well Targeted at Reducing Poverty

PRINT
Do barriers create "bubbles?"... The Discipline of Dismissal...
Economists Joseph J. Sabia and Richard V. Burkhauser examined the effects of state minimum wage increases between 2003 and 2007 and reported that they found no evidence the increases lowered state poverty rates.

Further, they calculated the effects of a proposed increase in the federal minimum wage to $9.50 on workers then earning $5.70 (or 15 cents less than the minimum in March 2008) to $9.49. They found that if the federal minimum wage were increased to $9.50 per hour:

. Only 11.3 percent of workers who would gain from the increase live in households officially defined as poor.
. A whopping 63.2 percent of workers who would gain were second or even third earners living in households with incomes equal to twice the poverty line or more.
. Some 42.3 percent of workers who would gain were second or even third earners who live in households that have incomes equal to three times the poverty line or more.


This is from my short piece, "Most of the Benefits of a Minimum Wage Increase Would Not Go to Poor Households," National Center for Policy Analysis, January 13, 2014.

Later in the piece, I report Sabia and Burkhauser's finding that if one takes account of even small disemployment effects of the minimum wage, one finds, of course, an even smaller effect on reducing poverty.

Finally, I point out something that Professor Burkhauser agreed with in correspondence:

This estimate overstates the gains to households from increasing the minimum wage. Why? Because, to the extent they are able, employers will offset the higher minimum wage by reducing non-money components of worker compensation. Burkhauser notes that such an effect will not show up in the government data because the data do not measure these non-money parts of the compensation package. But that is small comfort to those who would find themselves with higher-paying but reduced-benefit jobs.


Comments and Sharing





COMMENTS (45 to date)
matt writes:

So how many people who are listed as poor are working? and if we put in other measurement lines like how many get above 50% of poverty line? or 75%? what would that tell us?

Also if more money in the economy increases demand which may increase the number or hours of jobs available, how can that be quantified?

Is it instead a case that this hourly rate increase is not enough?

Pajser writes:
David Henderson: ... to the extent they are able, employers will offset the higher minimum wage by reducing non-money components of worker compensation.

Employers had incentive to reduce non-money component even before minimum wage increase, right? What changed? Do you suggest that employer's profit incentive increased with reduction of their profit? Is it well known phenomenon? Is it argument for minimum wage increase?

Jon Murphy writes:

Pajser,

How is compensation determined?

Jon Murphy writes:
Also if more money in the economy increases demand which may increase the number or hours of jobs available, how can that be quantified?

Matt, that's putting the cart before the horse. Production leads demand, not the other way around.

JT writes:

Schmitt's paper at CEPR already critiqued this study:
"The Sabia, Burkhauser, and Hansen study, however, is subject to the same critique applied to Hirsch, Kaufman, and Zelenska (and Card and Krueger before them). Sabia, Burkhauser, and Hansen analyzed only one experience of the minimum wage. Even if the effects of the minimum wage were, in truth, zero, we would expect to see a distribution of estimates around zero, including both positive and negative estimates. As Doucouliagos and Stanley demonstrated in their large meta-study of employment effects through the middle of the 2000s, the minimum-wage literature on teenagers showed a range of positive and negative effects, but also a large spike of the most accurate estimates at, or very near, zero. Wolfson and Belman’s meta-study, which focused on the period from about 1990 through 2010, confirms Doucouliagos and Stanley's findings with more recent research. Given how far the Sabia, Burkhauser, and Hansen estimates lie outside this consensus range, the burden of proof would seem to fall on Sabia, Burkhauser, and Hansen to explain why their study of a single experiment with the minimum wage should outweigh the cumulative experience of scores of studies of the U.S. minimum wage since the early 1990s."

Employers had incentive to reduce non-money component even before minimum wage increase, right? What changed?

Rather obviously, the part of compensation that legally had to be in the form of money wage.

If a worker is compensated with $8 of wage and $1 of benefit, when the law mandates $9 of wage then the incentive is to eliminate the benefit of $1.

Awesome. Joseph Sabia was my econ and law professor at San Diego State. One of the best taught classes I took.

Pajser writes:
Jon Murphy, How is compensation determined?
By law, tradition, collective and individual negotiation ... Where are we going?
Patrick R. Sullivan: If a worker is compensated with $8 of wage and $1 of benefit, when the law mandates $9 of wage then the incentive is to eliminate the benefit of $1.
That incentive existed before minimum wage was increased.
Granite26 writes:

The switch from 'officially defined as poor' in the first line item to 'poverty line' in the 2x and 3x line items makes me a little uncomfortable.

Those terms are interchangeable, right?

Thus 25.5% of the benefits go to people with incomes between 1x and 2x the poverty line? Refresh my memory, 'income' in this sense includes what?

Finally, has anybody looked at the marginal tax rates on these groups (particularly the primary earners near the poverty line). Is it possible that the increased minimum wage would reduce the hours worked (and increase staffing reqs?) in a general way? I'm assuming that high availability of low wage labor dominates any ZMP effects here.

David R. Henderson writes:

@JT,
Schmitt's paper at CEPR already critiqued this study
Actually, JT, Schmitt’s paper did not critique this study. It critiqued, as you yourself quote, a different study, that done by Sabia, Burkhauser, and Hansen. Moreover, its critique was about their finding of a disemployment effect of a minimum wage increase. So let’s say that you agree with Schmitt that the destruction of jobs is close to zero. That still doesn’t detract from the main finding that Sabia/Burkhauser and I report: an increase in the minimum wage is not well targeted.

David R. Henderson writes:

@Pajser,
That incentive existed before minimum wage was increased.
Not for the workers whose value of marginal product was above $9 per hour.
@Jonathan Finegold,
Cool. Professor Sabia was very pleasant, at least by email. I didn’t talk to him on the phone.

Tom West writes:

Not for the workers whose value of marginal product was above $9 per hour.

Correct me if I'm wrong, but isn't marginal product an *upper* bound. My dimly recollected economics indicated that in a grossly simplified model, as soon as the supply for a particular skill was one greater than demand, everybody's wages dropped from marginal product to subsistence.

Obviously that's a huge simplification, but it does suggest that for something like unskilled labor, where supply vastly outweighs demand, the marginal product may heavily outweigh wages.

Personally, when I was working at minimum wage many years ago, I don't every recall receiving any benefits at all.

(And yes, I also remember discussions about how the burnt-out full timers were earning 4-5 times what we were. That was where I first observed that the more an employer pays for someone, willingly or unwillingly, the more valuable they are perceived by the employer. Took another 30 years before we learned that neurologically, cause and correlation is a two way street.)

Michael writes:

Even if that study was done correctly, it was based on the pre-crisis data. My guess is that things have changed quite a bit since then, and many more minimum wage jobs are nowadays done by genuinely poor people.

Jon Murphy writes:

Pajser:

By law, tradition, collective and individual negotiation

What else?

Enial Cattesi writes:

@Jon Murphy:

Supply and demand?

Interestingly those who can demand only low pay, are in low demand. Even if the supply is small, the demand is still very low.

Jody writes:

Pasjer: At the margin.

Let us assume that there is a diminishing marginal utility from each additional dollar, so that an extra dollar at 100k is worth less than an extra dollar at 90k.

In this context, the owner has marginally more incentive to find ways to cut costs after the increase in labor costs than before the increase.

Similarly, you'll see businesses be more cost conscious (Save that paper clip!!!) during a recession than during a boom. Either way, they make the same amount of money by cutting costs, but the marginal incentive to look for cost savings is greater when working from a lower baseline.

(I also have a reasonable model of this same effect from loss aversion, but both lead to the same result - post income reduction, there is an increased incentive to cut costs.)

Jon Murphy writes:

Enial, close. Very close.

What is a function of the demand?

Let me ask it this way:

I'm a baseball fan and I play in the back yard. Why does David Ortiz make more money than me playing baseball? It is just because he's a better negotiator?

Maniel writes:

To paraphrase Mark Twain, "there are lies, damned lies, and studies." No study can gainsay the fact that the minimum wage is a government distortion by LAW of the private-sector free market (for labor) and that, since it makes it illegal for me to work unless I can command that wage, is likely to result in my not working (in the visible economy).

David R. Henderson writes:

@Tom West,
Correct me if I'm wrong, but isn't marginal product an *upper* bound. My dimly recollected economics indicated that in a grossly simplified model, as soon as the supply for a particular skill was one greater than demand, everybody's wages dropped from marginal product to subsistence.
That is wrong. You can see why by drawing a graph where the demand is not perfectly inelastic, that is, demand slopes downward, and the supply curve at every wage shifts to the right by one unit. The wage will fall but the extent of the fall varies with the elasticity of demand.
The other way to see it, Tom, is to look at occupations (that’s related to skills) where the supply is at least one greater than demand. There are many such occupations today. Virtually everyone is making a wage well above subsistence.
Take an extreme case. One really good baseball player comes along and all the slots are already filled. Does pay of the stars fall from, say, $5 million a year to $20,000 a year?

JayT writes:

@Pajser, one of the reasons that the incentive may not have been there to get rid of benefits is that the employer was, for lack of a better term, accustomed to that price. However, once the price of the employee goes up significantly, then the employer will look for areas that can be cut.

For example, my company would always give out a bonus in June and in December. Through the recession the company hasn't grown at the desired rate, and as such the bonuses were cut down to once a year. They could have cut back on the bonus earlier, because the incentive to save money was always there, but they continued to increase profitability with the bonus still there. As soon as the profits didn't increase, they cut the bonus.

RJack writes:

If you just use common sense, you know what the effects of raising the minimum wage will do. If a business owner does have to pay around minimum wage--raising it substantially will leave him or her with few options. Raise prices substantially and hope customers still will pay these new higher prices. Cut hours. Automate. Reduce employment. Reduce benefits.

Mr. Econotarian writes:

People who are not familiar with the stats do not understand that a large number of poor people are making more than the minimum wage. Even most illegal immigrants earn more than the minimum wage.

Often it is a lack of hours worked that makes them poor. I suspect part of the issue is that many households in poverty are single-parent, and getting affordable and reliable child care is challenging, thus they have to spend some hours caring for their children instead of working.

Some BLS data:

Among persons in the labor force for 27 weeks or more, 4.2% employed full time were "working poor", compared to 15.1% of part time workers.

Among families with at least one member in the labor force for 27 weeks or more, those families with children under 18 years old were 4 times more likely than those without children to live in poverty.

Other data from the Urban Institute "From the simulations, we find that increasing the minimum wage to $6.15 an hour would reduce the size of the working poor population only slightly. If all prime-age, able-bodied adults in working poor families were to work full-time, year-round, incomes would rise above twice the poverty line for about one-fifth of the individuals in working poor families."

Also they report "35 percent of all poor children live in families in which either a single parent works 20 hours a week or two parents combine to work at least 35 hours a week... the poverty rate for all persons in families would fall from 12.2 to 3.6 percent if all family heads worked full-time, full-year. Her measure of family income nets out work-related expenses but adds in the value of the EITC. "

"Even though the primary earners in working poor families work about as much as those in non-poor families, non-poor families are far more likely to have additional earners working more hours. For example, the median total number of hours worked by working poor families is 2,600, substantially less than the 3,873 worked by non-poor families. In addition, the average hours worked per adult is appreciably lower in working poor families (1,508 per year) compared with poor families (1,820 per year)."

"Primary workers in working poor families tend to work full-time, full-year. However, the total hours worked by all adults tends to be much lower among the working poor than the non-poor because secondary workers work fewer hours in working poor families."

Matt asked how many who are poor are working. 60% of households in poverty have at least one person who is working. Out of 46 million poor people, 10.4 are working (but many of those 46 million are children).

Jon Murphy writes:

The point I was going for with my question to Pajser is this:

Workers are compensated based upon productivity. Yes, negotiation plays a role, but not in a major scene (or major cents! Yes, bad pun but I think I am hilarious :-) )

If a portion of one person's compensation is forced to rise by law or whatever, and their productivity doesn't rise as well, then other aspects of their compensation will be cut or, barring that, they will be cut.

The incentive to reduce compensation is easy to explain. The compensation was already as low as possible for the company and as high as possible for the worker. If this person's compensation were cut, he could easily be poached by another company by offering him more compensation.

This is why minimum wage laws are terrible for helping the poor. Productivity increased must come before wage increases, not after. No profit-seeking individual, whether he be a producer or consumer, is willing to spend more money for something unless he knows he is getting more for it.

It's just Say's Law. And I have nothing more to Say on the matter (yes, another bad one).

Tom West writes:

David, you are, of course, correct that demand is not inelastic, but given comparisons in unemployment rates between minimum wage and non-minimum wage jurisdictions, it doesn't appear that the curve is all that elastic, either.

After all, Canada, with a ~$10 minimum wage, has an out-of-work-and-school (NEET) rate for young people is actually lower than America's.

I have to admit that among all the freedom-limiting legislation out there, the concentration on minimum wage seems awfully odd from a political perspective.

If there was a single piece of legislation that Libertarians could attack that increases the public perception that Libertarians secretly believe "the poor aren't suffering enough", this would pretty much be it.

The statistics for the effects of a reasonable minimum wage are all over the place, but even cherry picking for the worst effects, the effects are simply not very large for the perceived benefits. After all, there's a reason a majority of economists support a minimum wage. It's just not *that* damaging.

Surely there are better policies to go after that do more damage to freedom and don't play right into the public's worst stereotypes of the Libertarian movement?

genemarsh writes:

Arin Dube released a long paper this month that found
a negative relationship between minimum wage and poverty across 12 different studies, even digging up data out of one of David Neumark's studies that didnt confirm the minimum wage opponent's preferred conclusions:

http://www.washingtonpost.com/blogs/wonkblog/wp/2014/01/04/economists-agree-raising-the-minimum-wage-reduces-poverty/

http://www.washingtonpost.com/blogs/wonkblog/wp/2014/01/04/economists-agree-raising-the-minimum-wage-reduces-poverty/

[Broken html fixed--Econlib Ed.]

Jon Murphy writes:

Tom West:

Couple of things

1) 1/3rd is not a "majority."

2) I cannot speak for Prof Henderson, but what I have heard from Dr. Mark Perry (Carpe Diem blog) and Dr. Don Boudreaux (Cafe Hayek) is the focus on minimum wage is because it is a good litmus test to one's economic understanding. Those who argue that minimum wage has limited/no effect on employment prospects of the poor are, for lack of a better term, economically illiterate. The evidence, both empirically and theoretically, is pretty overwhelming that minimum wage does actually harm employment opportunities and/or compensation of the low-skilled workers. Rather, the debate is whether or not the benefits exceed the costs. The perceived benefits, to use your word, are just that: perceived. Given the known effects of minimum wage on employment, the increased purchasing power of those who keep their jobs likely does not offset the decrease from those who are no longer employable.

To put it another way, why are price controls so unpopular among economists for every commodity except labor? Why is no economist out there saying "There should be a minimum price on everything! The increased demand from companies will offset the higher prices faced from consumers and everyone will be better off!" No one makes that argument because it's obvious results: a mal-allocation of resources would result, leading to waste.

Jon Murphy writes:

Oh, one other thing:

That argument assumes that demand drives the economy. That is not correct. Production must rise before demand, not the other way around. Say's Law.

David R. Henderson writes:

@Tom West,
David, you are, of course, correct that demand is not inelastic, but given comparisons in unemployment rates between minimum wage and non-minimum wage jurisdictions, it doesn't appear that the curve is all that elastic, either.
Right. I was simply addressing your claim, because you asked me to, that an addition of one person supplying his labor can bring the free-market wage down from a high level to subsistence.
You’re free to change the subject, of course, which is what you did, but just recognize that that’s what you did.
If there was a single piece of legislation that Libertarians could attack that increases the public perception that Libertarians secretly believe "the poor aren't suffering enough", this would pretty much be it.
That has not certainly not been my experience. If all I said was “Let’s not increase the minimum wage because business and consumers will suffer but low-paid workers will be better off,” then, yes, you’re right that it would increase the public’s perception in the direction you claim. But, of course, as you know, that is not what I have said, either in this post or in previous ones. My argument, rather, and the argument of many other economists, is that it hurts many of the very people whom it is allegedly designed to help.
Indeed, when I make this case to people in person, I get a lot of people reconsidering their views and making an important distinction between helping the poor and appearing to help the poor.
After all, there's a reason a majority of economists support a minimum wage. It's just not *that* damaging.
Your claim is simply wrong, Tom. A majority of economists does not support a minimum wage. I’ve seen you make that claim before. I’ve never seen you back it.

Jody writes:

Tom: "I have to admit that among all the freedom-limiting legislation out there, the concentration on minimum wage seems awfully odd from a political perspective."

Now my perspective may be limited (so correct me if you have empirical counts of blog posts to the contrary), but my perspective of the timing of libertarians' critiques of the minimum wage is that it correlates well with when minimum wage increases are being proposed / considered.

Which is to be expected as people are more responsive to ideas when they're on point.

So it doesn't seem odd to me to be chiming in on the national conversation when others have chosen the topic. (e.g., SeaTac, Fastfood strikes...)

Dale Carville writes:

RJack: "If you just use common sense, you know what the effects of raising the minimum wage will do."
Maniel: "No study can gainsay the fact..."
Jon: "This is why minimum wage laws are terrible for helping the poor...It's just Say's Law."
Jon: "The evidence...is pretty overwhelming."

Tom is right. It's a terrible issue to dig your heels in on, as the old theory has been proven inapplicable to hundreds of city state and national minimum wage increases around the globe over the last decade. By clinging to your theory over the course of its long debunking, perhaps now just hitting its tipping point, libertarians risk self-defining as non-serious scholars, intractable ideologues, or insincere about their putative empathy for the minimum wage worker. If there is no evidence that can justify revision of theory, if wages must not be raised in a time of massive profit taking, for the worker's own good! , the litmus test begins to backfire, repurposed to test for business class crackpottery

Here's a poll which shows a 50/50 split among economists
polled by Univ. of Chicago. At the very least, one can no longer call the evidence "overwhelming". The days of "it's common sense" and "everybody knows" have passed.

http://www.igmchicago.org/igm-economic-experts-panel/poll-results?SurveyID=SV_br0IEq5a9E77NMV

Jon Murphy writes:

Dale, where are you getting 50/50?

All I see is 32%.

old theory has been proven inapplicable to hundreds of city state and national minimum wage increases around the globe over the last decade.

That is empirically not true. Countries with minimum wage have higher unemployment rates (and lower labor force participation rates) than countries without minimum wage (I cite ).

Furthermore, many of the studies looking at minimum wage in the States look only at immediate effects, ignoring effects on future hiring and the fact that wages are increased in increments, allowing businesses to adjust via attrition. This fact has been discussed multiple times on this blog, as well as elsewhere.

The body of evidence is quite heavy on the negative effects of minimum wage on those it supposedly intends to help (I cite).

All due respect, Dale, to say that minimum wage has been "debunked" is to ignore a lot of research and common sense.

David writes:

http://money.cnn.com/2014/01/14/smallbusiness/minimum-wage-hike/index.html?iid=HP_LN

Quotes from the business owners are interesting.

Maniel writes:

@Dale
It's a terrible issue to dig your heels in on.
My issue was that, being unskilled and unable to qualify for the so-called minimum wage, I would be unable to find work in the visible economy (you know, the one which all those studies address). However, in the larger scheme of things, you are right - there are indeed many alternatives: work in the less visible economy; the government dole; the parents' or relatives' dole; etc.
BTW, you may count me as an "intractable ideologue" WRT laws which take away my freedom (as well as WRT to the law of supply and demand).

Tom West writes:

My argument, rather, and the argument of many other economists, is that it hurts many of the very people whom it is allegedly designed to help.

I do agree that is your argument, and your posts have made me consider that the disemployment effects are not considered strongly enough in general. (My support for a minimum wage is based on my observations of the social effect of paying very low wages on employers and the citizenry in general.)

I was just struck by the optics of minimum wage fights while I was writing my post. It wasn't directed at your posts in particular.

As for my claim about economists supporting minimum wage, it was Bryan's article Can 600 Economists Be Wrong About the Minimum Wage?" from which I took it. However, my memory of majority was *not* accurate (that was Bryan's guess ("I'd say "near-majority" rather than "clear majority," even if there were no option to abstain.")). Thanks for making me dig it out.

(Note, it's pretty clear that most economists who do support the minimum wage do believe there are disemployment effects.)

(And yes, this entire 'political' business is an aside - my apologies if it's unwanted. I tend to treat comments like cocktail party conversation so I will occasionally post comments that are only tangentially relevant to the original topic if they might spur an interesting conversation.)

john hare writes:

It seems to me that a pair of problems could possibly be teamed up to form a solution. Since the welfare state is here to stay in some form or another, why not simply allow welfare recipients to accept anything offered with only minor reporting requirements. If one can enter the workforce at sub-minimum wage without it affecting the government support to any serious degree, there would be incentive for the disenfranchised to take what they can get from two directions.

Over a period of time, some would become skilled workers for decent wages, while others would continue getting government support at lower levels as it would be a supplement instead of a complete living.

The employers that exploited the system would act as filters for those willing to pay for quality workers. A poor solution such as this just might be better than a worse solution that will exist if the welfare/unemployment scenario continues on the current path.

Jon Murphy writes:

Tom West:

Nothing is that article says 50%, or anything like that.

Mark V Anderson writes:

Tom West is correct that being against minimum wages is bad politics. This is because most people are economically illiterate, including most journalists, who generally portray disputes about minimum wages as between the poor and the rich. Minimum wages are generally treated as re-distributions from the rich to the poor, and thus those against minimum wages simply don't believe that the poor deserve the extra money.

But not everything is about politics. Minimum wages are really such a bad thing for the poor that we must fight them, regardless of the political ramifications. Sure we'd like to win elections (not that libertarians do anyway), but what's the point if you give away your principles?

Minimum wages will increase unemployment amongst the poor, especially for the vital first jobs that poor teenagers need to get into the market. These enforced higher wages also greatly reduce the power of workers when on the job, because it is so easy for employers to replace workers when they are priced above the market.

Why is it that labor markets don't seem to work like other markets, where the buyer and seller generally have equal power in the transaction, because either side can deal with another player in the marketplace if they are dissatisfied with the offer received? A big part of the reason labor markets don't work this way is because of minimum wages and other "protections" of workers, which make the demand for labor perpetually lower than supply. At a pure emotional level, I would like to lose all my "protections" as an employee, so that when I deal with employers myself, I can deal with them as an equal party. I have something they need; it is just a matter of determining the terms. Labor regulations, and particularly minimum wages, make me feel like I'm begging for a job when I go on interviews.

Jon Murphy writes:

Mark Anderson said:

But not everything is about politics. Minimum wages are really such a bad thing for the poor that we must fight them, regardless of the political ramifications. Sure we'd like to win elections (not that libertarians do anyway), but what's the point if you give away your principles?

Hear hear!

Maniel writes:

@John Hare
Congratulations! You have just invented the Negative Income Tax (or words to that effect).
Here's Milton Friedman explaining that:

john hare writes:

@Maniel,

Thank you for the link. It helps me with a project in progress.

Enial Cattesi writes:

It is hard for me to understand how people could think that raising the minimum wage doesn't affect the poor. It is not only a minimum an employer may pay, but, and this is most important, it is a minimum a potential employee can demand. It a barrier to entry the market for the low skilled, which poor.

I've seen this explanation many times:

We should set a high minimum wage in order for us to create a high paying jobs hub here. If we allow a low minimum wage we will be invaded by sweat shops which will put us in an everlasting circle of poverty.

This is an excuse for high minimum wages provided by economists. Sad state of affairs. Probably Bryan Caplan can see some signaling theory in here ...

Jon Murphy writes:

@Enial Cattesi

A point of clarification: economists who argue for minimum wage do not argue it has no effect on poor workers. They argue the benefits to those who keep their jobs outweigh the costs to those who lose.

Also, an argument that is often made is monospony. In a monospony market, raising the minimum wage can actually increase hiring. However, I find that argument hard to swallow at the national level as it would mean that the ~8 million firms in the United States are colluding with each other.

Enial Cattesi writes:

@Jon Murphy:

How can an increase in the minimum wage increase hiring in a competitive market? Are there any real monopsony cases?

And I don't get this:

They argue the benefits to those who keep their jobs outweigh the costs to those who lose.

What kind of argument can be offered for this line of reasoning? Do those economist also cherish "social justice"?

Tanya Thomas writes:

[Comment removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Jay writes:

Even supporters of a the minimum wage often fail to justify why we need it at the federal level. The idea that a rural Idaho business needs to pay the same amount as a New York bistro (excluding states laws) is ludicrous to me. Yes, states are free to raise it above the federal level and often do, but then why have the federal (or the debate) at all?

Comments for this entry have been closed
Return to top