The Effect of Minimum Wage on Unemployment

Interesting tidbit: the original U.S. 1938 minimum wage law arose from a special interest group of Massachusetts textile companies who wanted to make things more difficult for their competition in lower wage southern states (an early example of outsourcing).

The Law of Demand

If you raise the price of something, people will buy less of it. That is the law of demand. If the price of labor goes up, businesses cannot afford as much labor as before and the result is a negative effect on employment. So if the real minimum wage goes up, the unemployment of minimum wage workers goes up. That's pretty much all there is to it. If you understand that, and understand that there is no magic pixie dust that will change this basic relationship, then you pretty much understand it all.

Any controversy on the subject derives from an extreme minority of academics. It's good that there are people around to challenge conventional wisdom. But since this very small group of academics (notably Card and Krueger) tells the press, the people, and the politicians what they want to hear, it gets reported on a lot. In Congress, for example, during a debate on minimum wage, a Democratic House aide commented that they would "throw numbers out till nobody can have faith in the numbers, and then political considerations will win. That's how we do it every time." Nice. Government sure is great.

Some mistakes that are made:

Looking at changes in total unemployment

This would be analogous to looking at the total sales of a grocery store in order to determine the effect of a change in the price of apples. The number of minimum wage jobs is an exceedingly small percentage of total jobs, less than one half of one percent. It is a near certainty that any effect of minimum wage will be overwhelmed merely by the noise of total employment numbers. Amazingly, some studies use this data.

A better proxy for minimum wage employment is teenage employment, but it is hardly perfect because minimum wage jobs account for only about 10 percent of teenage employment. Ideally, you need to look specifically at minimum wage employment in order to determine the effects of minimum wage. Even then, it's worth mentioning that sometimes these figures may include employment for waiters and bartenders, who are guaranteed to make at least minimum wage, but almost always earn more, sometimes substantially more.

Despite its flaws, teenage employment can often show the relationship quite clearly:

Not looking at sub-minimum wage employment

It may surprise many people, but there are plenty of jobs that pay less than minimum wage. In fact, there are actually about 3 times as many jobs at sub-minimum as there are at minimum. Some of these may be vaguely familiar, such as teenage workers in their first 90 days of employment, certain day laborers or people with certain disabilities. Alan Reynolds relates that this category also includes "workers on small farms or at seasonal amusement or recreational facilities, newspaper deliverers, companions for the elderly, outside salesmen, U.S. seamen on foreign-flag ships, switchboard operators and part-time babysitters." Also many small businesses with income less than $500,000 are allowed to pay less than minimum wage.

It is very telling to look at this category of worker. When minimum wage increases, sub-minimum wage employment reliably increases. Strikingly. Sub-minimum work ends up acting as a kind of safety valve for workers who lose their jobs due to an increase in minimum wage. This category of worker has been known to double in size after a minimum wage hike (for example after the 1997 hike, when among teenagers, sub-minimum work almost tripled). Since it is a bigger category than minimum wage to begin with, and because it is lumped into commonly compared totals such as teenage employment, or "minimum wage or less" employment, it is very important to separate out this category.

Not looking at above minimum wage employment

There is also an effect on workers above minimum wage. For workers that make slightly above minimum wage, they are much less likely to get a raise after a minimum wage hike. The money that might have been allocated to their wage increase (in exchange for their productive work) is instead being used to subsidize less productive workers that were making less than the new minimum wage.

Looking at individual instances or anomalies rather than the big picture

In this case, results are again getting swamped by the noise inherent in the data. Ideally you would like to look at a large sample size, over a lot of different timeframes (and compare apples to apples), if you want to see what's going on. An example would be looking at say, just one specific period in one specific town. A lot of other things might be going on. Maybe a new factory just opened up. Maybe high school just went back in session. Much better to gather information from many areas (and of course many states have different minimum wages), preferably during multiple periods of minimum wage increases, in order to minimize the effects of such anomalies. Nobody said it was easy to get good data.

Not looking at the overall condition of the economy

In a period where employment in general is increasing, any negative effects will not be as noticeable. But they will exist. The problem is that an increase in employment may be seen by some as a success. Overall, some may see it as a good thing, but what we really want to know is not just that employment increased, but that it increased just as much as it would have without a change in the minimum wage. If employment increases, but less than it would have otherwise, that represents people who would otherwise be working, or working at higher wages. This is reflected in the extraordinary increase in sub-minimum wage jobs following a minimum wage increase. Additionally, this increased supply of workers at below minimum actually creates downward pressure on the wages of already low paying jobs. Not a nice thing to do in my book.

Discussion

Much of the debate seems to hinge on Card and Krueger's study of the 1992 minimum wage increase in New Jersey, which was based on a 24 question phone survey of Burger King restaurants. Only two of the questions pertained to minimum wage, and these were ill-defined, not distinguishing whether respondents were reporting workers per shift, day, or payroll period, for example. Also, definitions of exactly what was meant by part-time and full-time were lacking. If you want to challenge conventional wisdom, you really need to have flawless data and methodology. Other studies covering this timeframe and area show results along conventional lines.

Additional debate turns on data following the minimum wage increases passed in 1996 and 1997. Overall, this was a period of strong employment growth. It would not be surprising to see minimum wage work holding steady or even slightly increasing, as was the case. But, as mentioned above, sub-minimum work among teenagers tripled. That should tell you something, i.e., that we forced a lot of people into sub-minimum jobs that wouldn't be there otherwise. Let's look at unemployment of black teenage males during this specific time frame (and remember that this was a period of strong employment growth), as well as the history since 1954:

Unemployment of black teenage males
at the time of increase in minimum wage:

1996
month prior to increase   36.6%
month after increase      38.6%

1997
month prior to increase   33.9%
month after increase      37.6%

As Milton Friedman put it, the minimum wage law is one of the "most anti-black laws on the statute book." If you are young, black, and urban, you're screwed.

People want to help others that aren't as well off. And the idea that we can just force low wage employers to increase wages is a very popular, vote getting idea. It fits the idea of "help the poor" as well as the dogma that "corporations are bad." Of course, most employers of minimum wage labor are small businesses, and some are forced out of business by minimum wage increases. While many people think of fast food businesses as giant corporations, at the restaurant level these are just individual franchises.

A lot of rhetoric in this area seems to be based on the idea that we can "get away" with a small increase in minimum wage because any effects are small. Often, this is based on the type of studies that use total job data. Well, they are right, in a sense - we can get away with these increases. A few hundred thousand low income workers forced into lower paying jobs or outright unemployment IS very small in the overall scheme of things. Some liberal arguments simply cast this group aside as "a small minority of workers whose employment patterns are most volatile." Wow. What elitist crap. I would argue that it's the wrong thing to do, that you're hurting the exact people you should most want to help.

If you wanted to "get away" with using minimum wage as a part of an overall strategy of subsidies, you could do it, but you'd have to add some other subsidies. In my mind you would not only need to compensate those people you pushed into lower paying jobs or unemployment, who are arguably taken care of through existing programs, but you would also need to compensate the numerous (mainly small) businesses who are forced to pay higher than market price for labor.

There is also a lot of talk about exactly who works at minimum wage. Suffice it to say the statistics can be bent a number of ways here, but overall the average worker is young, single, and not the head of a household or sole breadwinner. One example of bending statistics regards the number of people who work full-time at this level. As with many things, it depends on how you define it. You will find that arguments from the left use stats from anyone who has worked full-time for any length of time during a year, no matter how short, giving a figure of about 40% of minimum wage workers. On the other hand, the percentage of people who actually work full-time year-round at minimum wage is small, about 9% of minimum wage workers.

Summary

You can't get around the law of demand - there is no such thing as a free lunch. Increasing minimum wage has a reliably negative effect on minimum wage unemployment. The benefits go to a few, while the devastating result is that many are pushed either into unemployment, or employment at lower wages. Increasing minimum wage changes the distribution of low wage jobs, increasing the proportion of jobs below minimum, and creates pressure to lower the lowest wages even further. This is not compassion. Demographically, among the hardest hit are young black urban males. Overall, aren't we better off allowing people to get an entry-level job rather than forcing them into unemployment?

Another devastating result is that minimum wage law ends up being a penalty on small business. A better plan to help the poor would be to allow the marketplace to make its decisions as to who is employed and at what wage, and then give out some form of subsidy which does not directly target a single industry. Of course, these already exist in the form of various programs and the earned income tax credit, but for politicians, why increase those when you can get more votes by increasing minimum wage?

Finally, businesses end up compensating for the increases by reducing training, fringe benefits, and maintaining more stringent work requirements. This is an expected result and very similar to what happened in France's recent failed experiment of reducing the work week to 35 hours.

Addendum

This shows the real minimum wage has remained fairly steady for 30 years, and the percent of hourly paid workers at or below minimum wage has fallen substantially, to a current mere 3% of hourly workers.

Generalizing the principle

The same type of thing happens when you raise any wage above its market price, whether you are talking about minimum wage, unions, or living wages. And another category of problems occur when you try to keep a price below its market price, for example with rent control or the gasoline price controls back in the seventies.

At a more general level, as you increase the price of any business expense, profits decrease. This happens whether you are increasing the price of labor, in the form of wages, or capital, in the form of interest rates, or government, in the form of taxes, or raw materials, in, say, the price of oil. All of these things have a negative effect on profits, and have a similar effect in the aggregate on the economy. Thus all these things tend to turn the economy towards recession (negative growth). With less growth, or negative growth, the pie becomes smaller, and the smaller the pie, the smaller the number of people that can be employed.

Minimum Wage, Maximum Folly by Walter Williams

Home