Minimum wage, again Wednesday, Jun 23 2010 

A little less than a year ago, I wrote a rather long post about the minimum wage. I explained the “textbook model” of the minimum wage, which many students just beginning to learn economics are taught. The basic neoclassical model tells us that a minimum wage set above the equilibrium wage in a market creates a surplus of labor or, in other words, unemployment. I disputed some of the assumptions on which such an argument rests, for example, elastic demand for labor, the “one-sector” model, perfectly competitive markets, equal bargaining power, etc. I also looked at empirical evidence that suggests that the minimum wage may in fact be beneficial for employment or, in the very least, may only have a modest employment effect (primarily for teenagers). Finally, I looked at some ideological or pragmatic reasons why people support the minimum wage and why it is more favorable than other redistribution policies (e.g. welfare). Rather quickly, this post became the most looked at article on this blog, and remained that way for quite some time. Today, it remains the second most-read post I’ve written.

Last month, King Banaian, a professor and chairman of the economics department of SCSU, wrote about about a study that concluded people who accept “enlightened economics” are more conservative than they are liberal. These “economically enlightened” folk were required to believe, for example, that a minimum wage necessarily decreases employment. I disputed this type “enlightened thinking.” Dr. Banaian has again made another post about the minimum wage, this time explaining why a minimum wage is bad policy (it prevents people from coming to “mutually agreed” wages below the minimum wage) and how there is a “consensus” among economists about this issue.

In the first post, I responded by saying there is quite a bit of evidence in support of a minimum wage, even if neoclassical theory provide none. One of the most famous example is research done by Card and Krueger, who found that the minimum wage had positive effects on employment. This seems quite stunning, considering the standard neoclassical model predicts just the opposite. So, quite naturally, one becomes rather suspicious of this research, but I think a careful review of the literature will show that the underlying conclusions that Card and Krueger come to are solid and are supported by additional research. Of course, one wonders how increasing wages can, in fact, increase employment levels. It seems counterintuitive. David Switzer, a professor of economics at SCSU, said it “goes against all of neoclassical economic thinking.”

Fortunately, neoclassical economics (as well as a little bit of intuition) does provide us with an answer. It isn’t, after all, beyond one’s imagination that an employer might actually pay its laborers a wage below the market clearing (i.e. equilibrium) wage. A firm seeking to maximize its profits has this incentive if it has the ability to do so. One scenario that might bring this about is one in which the labor market is oligopsonistic. Oligopsony is a fancy word to describe markets where there are few buyers and many sellers. (A related term that is perhaps more familiar is monopsony, where there is only one buyer and many sellers; this is the opposite of monopoly, which is one seller and many buyers.) In the case of oligopsony, the small number of firms can distort the wages in a market (in a similar way a monopoly can distort prices in a market), such that wages can be set below the equilibrium wage. Oligopsonistic labor markets reduce the welfare of laborers and creates deadweight loss. Under such circumstances, raising the wage that employers must pay their labor actually increases employment, reduces deadweight loss, and increases efficiency in the market. (A simplified graphical representation of monopsony can be viewed here.) So, in this case, the minimum wage has some extraordinary benefits.

The question becomes whether particular low-skilled labor markets are oligopsonistic or not. If the New Jersey fast food industry was oligopsonistic in 1992, that might explain Card and Krueger’s findings. However, as Dr. Banaian points out, the research in this area is not robust and is still “very young.” He may well be correct, in which case it would be helpful to look at empirical evidence and other areas that are more thoroughly understood. As I said earlier, a little bit of intuition might be able to help us explain why the effects of minimum wage may not be consistent with the standard model. In a 2008 study, David Metcalf explores why the minimum wage in Britain has “had little or no impact on employment.” Some of these include changes in hours, tax credits, compliance issues (part of the two sector model that Gary Fields discusses in previously noted research), productivity changes, price changes, reduced profits, and so on. He also considers the existence of “modern monopsony” (oligopsony) “very likely” in British labor markets. I defer you to Metclaf’s research for a more thorough discussion on how these variables can effect employment levels following a minimum wage hike. Suffice it to say, how these variable change does have an effect on employment, and may help explain why the minimum wage might have “minor negative effects at worst.”

In fact, that’s what most research has concluded. The conclusion that I support is that the minimum wage has a modest adverse effect on employment, primarily for teenager workers. It may even have positive employment effect for older cohorts, consistent with research by David Neumark and Olena Nizalova. (Neumark, keep in mind, is a fairly notable labor economist who opposes the minimum wage.) I think this is what a majority of the published literature out there reports (I can provide plenty of references, if needed), and the reasons explaining these findings are quite reasonable. That isn’t to say that there is a “consensus” against the minimum wage, as Dr. Banaian contends there is. He thinks I am “wrong on this point in terms of where the profession is on the literature.” A few years ago, The Economist, the main establishment journal, actually printed an interesting story on the issue. They wrote, “Overall, economists have become less worried about the job-destroying effects of a modest hike in the minimum wage. . . . Today’s consensus, insofar as there is one, seems to be that raising minimum wages has minor negative effects at worst.” There’s a wealth of research to support these views, as I stated earlier. What there is not is a consensus against the minimum wage, as Dr. Banaian contends there is.

In defense of his position, Dr. Banaian cites research by Neumark and William Wascher, which stated, in its abstract no less, “Our review indicates that there is a wide range of existing estimates and, accordingly, a lack of consensus about the overall effects on low-wage employment of an increase in the minimum wage.” Even more stunningly, Dr. Banaian readily confessed these facts in a post on his blog post he made in 2006, stating, “Both studies find a lack of consensus on the minimum wage, which I simply find shocking.” He finds the lack of consensus among economists “shocking,” but he at least acknowledges the fact. Today, he has shrunk from the issue and maintains that there, in fact, a consensus. He cites, for example, a 1996 survey by Robert Whaples, which suggested that there is a consensus among labor economists that the minimum wage decreases employment. That’s already been established. What Dr. Banaian conveniently does not do is refer to Whaples’ 2006 survey of PhD economists from the American Economic Association, which found that only less than 47% of them disagreed with a minimum wage policy. Though he readily mentioned it four years ago, perhaps the 2006 Whaples study is too inconvenient for the Minnesota House Representative hopeful in 2010.

The question, then, becomes less about the employment effects of the minimum wage, since there does seem to be some agreement on that issue. As one study by the U.S. Congress revealed, “Historically, defenders of the minimum wage have not disputed the disemployment effects of the minimum wage, but argued that on balance the working poor were better off.” That’s always been at the heart of the issue. Richard Freeman, one of the foremost labor economists and a professor at Harvard, writes in a 1994 study, “The question is not whether the minimum distorts market outcomes, but how its distortionary effects compare with those of other modes of redistribution, or with the benefits of redistribution.” He concludes that the minimum wage is a decent redistribution tool for four primary reasons that are typically ignored in the textbook models. I think his conclusion is consistent with what a majority of Americans believe. An overwhelming majority, usually over 80%, support the minimum wage. People support policies that help those who work (you need to work to earn the minimum wage), compared to those that help non-workers (e.g. welfare). They also are comfortable with redistributing their income via higher prices to help the most disadvantaged of workers. As Gary Fields keenly points out in a 1994 study, “One’s views about the desirability of a minimum wage ought to depend on more than the size of the unemployment effect alone.” I think he’s correct.


Business ethics and economics Thursday, Oct 29 2009 

Today I attended a presentation given by Paul Neiman, an Assistant Professor of Philosophy at SCSU. The topic of his presentation was international business ethics. He focused on providing an ethical framework for conducting international business from a social contract perspective, expanding on the work of John Rawls and his “A Theory of Justice.” The presentation is based on a paper he is currently in the process of writing.

The basic premises are that social contracts should incorporate common shared presumptions that are reasonable and generally acceptable. Based on this, Dr. Neiman argues there are four restrictions that seem reasonable and generally acceptable to place on free markets:

1. Contractors should not be forced into accepting or rejecting principles.
2. Contractors should not be willfully deceived in arguing for or against principles.
3. All contractors must have an equal right to propose or argue against principles.
4. All contractors should expect that the terms of the social contract will be enforced.

These are all well and good. They are reasonable and generally acceptable rules to impose. The problem is when Dr. Neiman ventured into what these rules would result in for two contractors, one representing a corporation and another representing a community, coming to negotiate a deal but who are completely unfamiliar with their constituents. That is, the business contractor does not know anything about the corporation he is representing other than that they seek profit maximization. Likewise, the community contractor does not know anything about the community she represents except that they care about their living standards, their culture and social norms, their environment, and so on. These two people are then supposed to negotiate a deal based on the four rules above, and we’re supposed to assume both have an equality of power (the community does not desperately need the corporation and the corporation does not desperately need the community).

Based on these rules, Dr. Neiman posits that the negotiators will come to expect that the corporation is obligated to pay a living wage, be fully responsible for any environmental damage it creates, respect cultural and social norms all the time (i.e. not just when it is profitable to do so), and so on. Clearly it seems the balance of power rests with the community negotiator, not with the corporation (and Dr. Neiman justifies this by saying it is the community that has the deal-breaking terms, as if the corporation has none of its own).

To be frank, the presentation made little economic sense to me, as someone who is minoring in the subject. That’s just my opinion. Let’s take the living wage obligation, for example. Dr. Neiman says the community won’t accept any corporation that won’t pay its population a living wage because that would decrease their wages. I assume that means they won’t take any salary below the average. Already something seems quite wrong with this ethical framework. What happens to, say, cashiers at a grocery store? They usually don’t get paid a “living wage” because they do not add at least that much revenue to the company. If the marginal cost of an additional laborer does not at least equal the marginal revenue of hiring that laborer, the laborer won’t be hired. That is, the company won’t hire someone at a cost that exceeds the benefit that hiring that person would bring. Otherwise they’re just losing money. So what does this mean when we say the corporation is obligated to pay a living wage? Well, it means this theoretical world doesn’t have any cashiers or any other job, for that matter, that would normally pay less than a “living wage.” (That’s the classical argument against minimum wages, a topic I’ve explored here [by far my most popular post for some reason]. The difference, really, is in the magnitude in setting the minimum. The hypothetical purpose of a minimum wage, as I see it, is to set wages at an equilibrium price that clears the market, essentially correcting for market failures that exist. A “living wage” minimum sets it way above this level and is based more on ethical rather than economic arguments.)

So the result is actually very high unemployment because there is a lack of jobs. Jobs that cannot afford to pay a “living wage” won’t exist. People whose labor is not worth this living wage are out of luck. And I did ask Dr. Neiman about this problem. He essentially responded that he doesn’t buy the argument and that perhaps some people will just have to live unemployed to persevere the interests of the community, namely high wages. To me, this is a very astonishing ethical guideline he is proposing. What he is saying is that it is better to receive nothing rather than something. That it is better to live in poverty and in unemployment than to receive at least some amount equal to the worth of your labor (if your labor is worth less than the “living wage”). Is that ethical? Further, having people earning zero rather than even a minimal amount decreases the average wage, what Dr. Neiman was originally against.

A look at the minimum wage Saturday, Jul 25 2009 

Yesterday the minimum wage rose to $7.25, thanks to a bill signed by President Bush in May of 2007. I figure now might be a good time to examine the minimum wage and its effects on employment. Forewarning: this is a very long post and is very economic oriented, so bear with me.

In my freshman year at SCSU, I took a course on the principles of microeconomics (with Dr. Komai, an excellent professor) where we learned about price floors including a minimum wage. I found the topic interesting, and so wrote about it in a research paper for my English course. In it, I argued for an increase in the minimum wage and indexing it to inflation. (Note this was in 2006, prior to President Bush signing the minimum wage increase bill.) This is back when I considered myself a Republican (!) and when most Republicans and other economic libertarians were arguing for an abolishment of the minimum wage under the assumption that a minimum wage causes higher unemployment. Many microeconomists and other economists would agree with that assessment (see, for example, this blog post by Gary Becker made around the same time as I wrote my paper). Now in an intermediate microeconomics course, I was again presented the classical economic argument that a price floor like the minimum wage will create a surplus (i.e. unemployment). Why do economists make this argument and does it make a minimum wage undesirable?

I find using graphs the easiest way to understand price floors. First of all, a price floor is when the government sets a minimum price for a product or good. This typically occurs when it is perceived that the normal (i.e. equilibrium) price is too low and that a higher price would be more favorable (such a regulation on the price is meant to be beneficial to the supplier of that good—but more on that later).

Without outside regulation, a price can be determined where the demand of a good intersects or is equal to the supply of that good. Graphically, this can be represented as:
As you can see, if the price went down, the quantity demanded would increase, but the quantity supplied would decrease (creating a shortage of goods). If the price increased, quantity demanded would decrease, but quantity sold would increase (creating a surplus). A surplus will put a downward pressure on price until the quantity demanded is equal to the quantity supplied. A shortage will put an upward pressure on price until quantity demanded is equal to quantity supplied.

If the government stepped in and made it illegal to sell or purchase a good (labor in this case) below a certain price, a surplus of that good will exist if that new price is above the equilibrium price. Graphically, this can be represented as:
If the minimum wage is set above the equilibrium wage, then there will be a surplus of labor (i.e. unemployment) because the quantity of labor supplied will be higher than the quantity of labor demanded. This is why most Republicans and many economists argue against the government instituting a minimum wage. It creates more unemployment than there would be without a minimum wage. The higher the minimum wage is above the equilibrium wage, the higher the unemployment is going to get.

That is the classical economic model of the minimum wage (also known as the “textbook model”). This explanation is short and deliberately not thorough for the purpose of space and time, but accurately represents the typical example of a price floor. In understanding economic theory, however, you should always realize that certain assumptions are being made. To gauge the reliability of that theory, you would do well to also gauge the assumptions on which the theory is being made. Sometimes you may find that, while the theory looks nice, the assumptions on which it depends can be seriously strenuous. In this case, the classical economic model makes a number of assumptions that also happen to influence how the slope of the demand curve looks (i.e. the elasticity of demand). (All elasticity means is how sensitive people are to changes as measured through percent changes. It can measure, for example, how responsive people are to changes in price. Mathematically, this can be represented as %ΔQuantity/%ΔPrice.)

If the slope of the demand curve (i.e. the wage elasticity of demand) is changed, you’re going to end up with different results. Namely, unemployment is higher the more elastic the demand. Conversely, if the demand is fairly inelastic, unemployment will not be as high. Graphically, this can be represented as:
In this example, the D1 curve is more elastic than the D2 curve. With the minimum wage set above the equilibrium wage, the quantity of labor demanded is less for the elastic demand curve (Qd1) versus the inelastic demand curve (Qd2). Therefore, the unemployment resulting from the minimum wage is smaller when demand for labor is inelastic. So we must therefore ask how elastic the demand for labor is. However, we should also realize that this model is still based on classical economic model where a number of assumptions are being made.

Gary Fields, a professor of labor economics at Cornell University, questions one of these assumptions in a 1994 article. He argues against the “one-sector model which implicitly assumes that the minimum wage applies uniformly to everyone.” Instead, he claims a two-sector model, including a covered and an uncovered sector, is more appropriate based on empirical evidence that suggests two sectors (i.e. covered and non-covered) exist nearly everywhere. Fields concludes that actual theoretical results are less clear-cut than assumed in the textbook model, which “simply cannot be relied on” because it is unrealistic. Fields also stresses the importance of empirical studies to measure the real effects of the minimum wage, because theoretical models do not suffice.

Other assumptions being made have also been challenged. For example, it’s likely the case that employers of labor have monopsonistic power—that is, where few buyers (or employers of labor) face a large amount of suppliers. Without going into detail, this would suggest a higher minimum wage would also increase employment (up to a certain point of course). Again, other assumptions can be challenged (and have been) and so it is important to not only rely on theoretical models but to also look at empirical data gathered from real-world situations.

One study I looked at was a 1994 paper by Richard Freeman, a respected labor economists at Harvard. He found that pre-1980 minimum wages had a “modest adverse effect on employment,” where a 10% increase in the minimum wage would increase joblessness by 1 to 2%, implying the minimum wage is an “effective redistributive tool.” During the 1980s when the minimum wage was held stagnant and the real value of minimum wage decreased, unemployment of low-skilled labor actually increased. In surveying three studies on employment after the early 1990s minimum wage increases, Freeman found that two reported insignificant unemployment effects and a third reported an increase in employment (suggesting employers were monopsonistic). Freeman goes on to cite other empirical studies that show moderately adverse or even positive effects on employment through minimum wage increases, again suggesting the minimum wage is an effective redistributive tool.

A highly influential paper on the effects of the minimum wage on employment was done David Card (UC Berkeley) and Alan Krueger (Princeton University) in 1992. In it, they found that employment went up after a minimum wage increase from $4.25 to $5.05 in New Jersey (while Pennsylvania’s stayed at $4.25). Classical economic theory would have suggested employment would decrease. Of course, Card and Kreuger’s work was heavily criticized, for example by George Mankiw, an influential Harvard economics professor, and Gary Becker (who I noted earlier in this post). Others, such as Joseph Stiglitz and Paul Krugman, both influential Nobel laureates, accepted Card and Kreuger’s findings. Surveys of economists have shown a decreasing amount of economists who believe the minimum wage result in significant negative employment effects, partly due to Card and Kreuger’s research (see, e.g., here). (Still, however, a majority agree a minimum wage does increase unemployment.)

Finally, in a 2004 paper by David Neumark (UC Irvine) and Olena Nizalova (MSU) find that in states with higher minimum wages, workers of ages 16 through 19 experience higher unemployment; in the 20 through 24 age range, however, the unemployment difference is smaller; and within the 25 through 29 age range, there is even higher employment in states with higher minimum wages. This suggests that young teenagers, ones most likely to live with their parents or be second income earners, take the brunt of any unemployment caused by the minimum wage. William Maloney and Jairo Mendez affirm in a 2004 paper, “[unemployment] effects [from the minimum wage] have traditionally appeared to be weak in the US, perhaps with the exception of young workers.” Those who are in most need of the minimum wage increase—those more likely to be older and have children—are the ones who seem to be the least affected by any unemployment effects the minimum wage has. The minimum wage, therefore, is a fairly well targeted policy that has the potential to help low-income families when it is raised. Indeed, Sara Lemos proclaims in a 2006 paper, “The minimum wage has a concrete potential to serve as a policy to alleviate inequality and poverty without undesirable side effects.”

Many Americans happen to agree. One reason why people tend to favor the minimum wage over other redistributive tools and transfer payments is because it creates an incentive to work. As Richard Freeman points out, a minimum wage requires no taxes or government borrowing. This is contrast to other policies such as earned income tax credits or subsidies for employers of low-skilled labor, which must come out of the government’s budget.

Critics, on the other hand, maintain the minimum wage causes prices to increase. Daniel Aaronson and Eric French, writing for a paper published by the EPI in 2006, explicitly find that prices rise about 0.4 to 1.5 percent with a ten percent minimum wage hike. This makes sense because a higher minimum wage imposes higher costs for firms who hire at the minimum wage. “Higher labor costs are pushed on to consumers in the form of higher prices,” they write. However, this reason alone should not be used to detract the minimum wage. In a 2003 paper, Sanjiv Sachdev finds that most people prefer to redistribute their income (e.g. through increased prices) in such way that rewards those who work, not to those who do not. The minimum wage acts like a redistributing tool for the poor, but unlike other means of doing so (e.g. welfare), it rewards those who work. We must therefore weigh the cost of higher prices to the rewards that the minimum wage brings.

Millions of people today work at the minimum wage, and many more are affected by the minimum wage by so-called “knock-on effects,” and whose collective bargaining is weak (Stephen Bazen, 2006). With researchers suggesting ambiguities in the classical theoretical model of the minimum wage and empirical studies showing only moderately adverse, insignificant, or even positive effects on employment, I think the minimum wage can be used as an effective policy for raising living standards of low-skilled workers. It effectively targets those who work and those most likely to be breadwinners of the household. With the new minimum wage hike passed by President Bush, I hope new studies can come out that can either confirm or deny these findings. Time will tell.