![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
kirisutogomenH.R. 2, the Fair Minimum Wage Act of 2007, was the second bill introduced in the new House Democratic majority's "100-Hour Plan" of legislation introduced between the opening of the 110th Congress and the State of the Union address. It would increase the federal minimum wage from its current $5.15 per hour to $7.25 per hour, over two years and in three increments. While the intent of poverty alleviation is admirable, an increase in the minimum wage is not the right tool for the problem. Minimum wages force some low-earning people into unemployment, and higher minimum wages are more damaging. Moreover, even to the extent that anyone benefits from a minimum wage, that benefit is paid for by a narrow, poorly chosen constituency.
Many countries have minimum wage laws, and there is widespread agreement of the value of such laws. A minimum wage may seem like a sensible approach for helping the poorest members of society. It would certainly be good if we could selectively target the lowest-income members of society and raise their incomes avoiding the disincentive to work that is often associated with the social safety net. Unfortunately, there ain't no such thing as a free lunch. Poverty is not the result of being paid too little; it is the result of one's work being of low value. Low wages are no more than the necessary acknowledgment of this by employers.
The most basic argument against minimum wages is straight out of Economics 101. A minimum wage is a form of price control, in this case a floor below which the price of a good (in this case, labor) is not allowed to fall. If the price at which supply meets demand is above this floor, there is no effect. For example, the supply of computer programmers meets demand at a wage above $7.25/hour, so a minimum wage increase has no effect on what computer programmers are paid.
But if the floor is above the price at which supply equals demand, more labor will be available than would have been at the equilibrium price, and less will be demanded. The result is oversupply, and if the good is labor, we call that unemployment.
The unemployment is not distributed equally across the population. The market for computer programmers is not affected. The market for truck stop diner wait staff is likely to be affected. If the market-clearing wage, that at which supply meets demand, would be $4/hour in the absence of outside involvement, a minimum wage of $5.15/hour will reduce the number of food servers needed while increasing the number of people willing to be food servers. The resulting unemployment is concentrated in the segment of the population with lower incomes, less education, less experience, and with the greatest need, making it more pernicious than other types of unemployment.
Admittedly, there will also be workers whose incomes rise as a result of a minimum wage. While there will be fewer wait staff, those that are employed will be paid more. It is not certain whether those who retain employment are sufficiently better off to compensate for both the increased unemployment as well as the increased inequality that results. What is certain is that once the employers are taken into account, society as a whole will be worse off. The employers will not be able to employ as many people, and thus will be unable to serve as many customers, and their costs for their employees will increase. The loss to the losers is necessarily greater than the gain to the winners, as the law is preventing voluntary transactions which, being voluntary, are mutually beneficial.
It has been argued that this predicted unemployment does not actually appear in practice and that the very simple economic model used does not account for the complexities of real-world labor markets. It has also been asserted that considerations of justice impel us to place greater weight on the welfare of the poorest members of society, and that the relevant metric for judging whether a minimum wage is desirable is the benefit to the disadvantaged, not the benefit to society as a whole. To the former claim, I will argue that the empirical evidence is mixed at best, and that the deviations from theory that would be necessary to result in no increase in unemployment are unreasonable. In response to the latter statement, I will further argue that even if we accept the dubious assumption that minimum wages are of net benefit to the underprivileged, there are far more just and reasonable ways to redistribute income, some of which already exist and could be easily expanded.
Proponents of a minimum wage increase are especially fond of citing Card and Krueger, whose 1997 book assembled evidence that increases in minimum wages did not increase unemployment. The criticisms of their methods and conclusions are too long and detailed to go into here. It will have to suffice that some of their methods, such as relying on telephone interviews to determine if employers intended to reduce employment, rather than on actual measures of employment, are questionable, and that many other studies before and since have found that increases in the minimum wage do in fact lead to higher unemployment, particularly among the young, the unskilled, minorities, and women.
More fundamentally, there is a very basic problem with theoretical explanations for how a minimum wage could fail to result in unemployment. The argument is based on the possibility of a labor market failure, in which the worker is paid less than the value of their labor. If the employee is already being paid the full value of their work, then requiring the employer to pay them more would leave the employer paying the employee more than their output was worth, i.e., losing value with every employee. An employer cannot employ people at a loss for long.
Deviations from the textbook economics model can create situations in which employees are paid less than the value of their output. One way in which this could happen would be if the employer has what is called market power. In the case of a producer market power is usually the result of monopoly or oligopoly; in this case it would be the consumer of the good (labor) with the market power, and this is called either monopsony or oligopsony. The classic case of monopsony would be the company town, a labor market in which there is only one consumer of labor, the single employer in town. Oligopsony would be the case of collusion between employers, in which the consumers of labor would cooperate to keep wages low, and either co-opt or exclude any potential competitors that might try to break into the labor market and bust the cartel.
In the case of a company town, monopsony would be a significant concern, but few areas still exist in the US that could qualify as company towns. Mobility of labor effectively prevents workers from becoming tied to a single employer, and mobility of capital prevents a single employer from dominating any given community. An employer that tried to pay wages below the value of labor would quickly find competitors moving into town to take advantage of the abnormally cheap labor.
Collusion between employers is similarly implausible in a modern economy. It is highly unlikely that Wal-Mart, for example, would collude with Target to underpay their employees, but it is unimaginable that they would also collude with McDonald's, Burger King, Hilton, Marriott, and the International House of Pancakes.
There are phenomena other than oligopsony which can cause markets to become inefficient. Jobs are not perfect substitutes for one another, and to the extent that employers are not directly competing with one another for employees, they would have market power. The weakness of this argument is that the low-wage jobs we are considering are actually close to being substitutes for one another. High-end jobs requiring a great deal of specialized knowledge and skills are far more segmented than low-wage work that requires much less specialized knowledge or skills. The sheer size of the market at the low end is also a closer approximation to the theoretical ideal of perfect competition.
Many of the ways in which labor markets are less than perfectly competitive may actually work to hold wages above their market-clearing levels, rather than below. Efficiency wage theory describes a number of ways in which it could be advantageous to firms to pay above-equilibrium wages. Turnover costs are substantial for employers -- it can cost a substantial portion of an annual wage to replace a worker, and it is likely more economical to simply pay an employee more than they are worth to avoid turnover costs. There is also evidence that paying employees above market-clearing rates can improve their productivity, possibly because of the psychological effect of gratitude, or perhaps because the resulting unemployment (caused in the same way as that resulting from minimum wages) makes it more plausible that an employer would be able to quickly and easily replace any under-performing employee with one of the similarly qualified unemployed.
A second set of arguments for the existence of and increases in the minimum wage is focused on distributional justice. These arguments may recognize that absent redistribution, a person's income is determined by the value of their output -- the value of a person's work cannot be improved by fiat. A minimum wage is thus simply another form of income redistribution. Even if the minimum wage leaves society as a whole worse off, its redistributive effect makes it worth the loss of societal wellbeing.
The question of how much societal welfare to sacrifice for a given amount of reduction of inequality is a difficult one, possibly even impossible to answer definitively. Fortunately, it is not necessary to know the answer, because there are improvements available in how we reduce inequality. A minimum wage, as a redistribution of income, rewards low-income workers, and is paid for by their employers. It is a combination of a wage subsidy for low-income workers and a tax on their employers. Transferring wealth to low-income workers is just fine, but it is perverse to impose an unfunded mandate that penalizes companies for employing the unskilled. We can agree that there ought to be a subsidy for low-income workers, but we should acknowledge that a minimum wage is only one possible way to implement it, and that other ways to pay for such a subsidy are available. If McDonald's employs many minimum wage workers, and Microsoft employs few, that is no reason to force McDonald's to fund the redistribution of wealth and allow Microsoft to dodge the costs. Even broadening the sources of funds to include all employers is still too narrow. There is no reason for us to consider Microsoft any more responsible for the alleviation of poverty than Paris Hilton. All we do by imposing the costs on only the employer is to disguise our redistribution as earned income.
There might be an argument for imposing the costs of wage subsidies on the employers of low-wage workers if we thought that the employers were responsible for the low incomes. But low incomes are the necessary result of low value of labor, which is in turn the result of poor education, disadvantaged backgrounds, and other circumstances that largely determine the value of a person's work long before they enter the labor market. The employers of low-income workers are often excoriated as exploiting their employees, but such accusations have a surplus of florid rhetoric and a deficit of substance. In fact the employers of low-income workers are among the only participants in the economy doing anything at all that benefits those workers -- they're paying them.
American society appears to be in consensus that we should assist low-wage workers. If we are to continue to subsidize their low wages, the money has to come from somewhere. Rather than foist the responsibility on some of the only people already doing anything to help them, it makes far more sense to broaden the burden as widely as possible, and fund wage subsidies out of general government revenue. These considerations of justice aside, the lowest deadweight loss to the economy results from the broadest possible funding of poverty alleviation. In fact, we are already doing just that, by means of the Earned Income Tax Credit. The EITC is a tax credit that effectively acts as a wage subsidy, and in some cases can even result in negative total tax liability. It is certainly not perfect, but it acts as arguably the largest poverty reduction program in the country, with $36 billion distributed in 2004. If Congress really want to help the poor, they should be expanding the EITC, not increasing the minimum wage.
Minimum wages almost certainly increase unemployment among the poorest members of society. They certainly injure the employers whose businesses rely on unskilled labor. Those people who are better off as a result would be better served by other poverty programs, which would help more people and impose less of a burden on society as a whole. A minimum wage may play well in a sound bite, but to genuinely help people, there are much better options.
Many countries have minimum wage laws, and there is widespread agreement of the value of such laws. A minimum wage may seem like a sensible approach for helping the poorest members of society. It would certainly be good if we could selectively target the lowest-income members of society and raise their incomes avoiding the disincentive to work that is often associated with the social safety net. Unfortunately, there ain't no such thing as a free lunch. Poverty is not the result of being paid too little; it is the result of one's work being of low value. Low wages are no more than the necessary acknowledgment of this by employers.
The most basic argument against minimum wages is straight out of Economics 101. A minimum wage is a form of price control, in this case a floor below which the price of a good (in this case, labor) is not allowed to fall. If the price at which supply meets demand is above this floor, there is no effect. For example, the supply of computer programmers meets demand at a wage above $7.25/hour, so a minimum wage increase has no effect on what computer programmers are paid.
But if the floor is above the price at which supply equals demand, more labor will be available than would have been at the equilibrium price, and less will be demanded. The result is oversupply, and if the good is labor, we call that unemployment.
The unemployment is not distributed equally across the population. The market for computer programmers is not affected. The market for truck stop diner wait staff is likely to be affected. If the market-clearing wage, that at which supply meets demand, would be $4/hour in the absence of outside involvement, a minimum wage of $5.15/hour will reduce the number of food servers needed while increasing the number of people willing to be food servers. The resulting unemployment is concentrated in the segment of the population with lower incomes, less education, less experience, and with the greatest need, making it more pernicious than other types of unemployment.
Admittedly, there will also be workers whose incomes rise as a result of a minimum wage. While there will be fewer wait staff, those that are employed will be paid more. It is not certain whether those who retain employment are sufficiently better off to compensate for both the increased unemployment as well as the increased inequality that results. What is certain is that once the employers are taken into account, society as a whole will be worse off. The employers will not be able to employ as many people, and thus will be unable to serve as many customers, and their costs for their employees will increase. The loss to the losers is necessarily greater than the gain to the winners, as the law is preventing voluntary transactions which, being voluntary, are mutually beneficial.
It has been argued that this predicted unemployment does not actually appear in practice and that the very simple economic model used does not account for the complexities of real-world labor markets. It has also been asserted that considerations of justice impel us to place greater weight on the welfare of the poorest members of society, and that the relevant metric for judging whether a minimum wage is desirable is the benefit to the disadvantaged, not the benefit to society as a whole. To the former claim, I will argue that the empirical evidence is mixed at best, and that the deviations from theory that would be necessary to result in no increase in unemployment are unreasonable. In response to the latter statement, I will further argue that even if we accept the dubious assumption that minimum wages are of net benefit to the underprivileged, there are far more just and reasonable ways to redistribute income, some of which already exist and could be easily expanded.
Proponents of a minimum wage increase are especially fond of citing Card and Krueger, whose 1997 book assembled evidence that increases in minimum wages did not increase unemployment. The criticisms of their methods and conclusions are too long and detailed to go into here. It will have to suffice that some of their methods, such as relying on telephone interviews to determine if employers intended to reduce employment, rather than on actual measures of employment, are questionable, and that many other studies before and since have found that increases in the minimum wage do in fact lead to higher unemployment, particularly among the young, the unskilled, minorities, and women.
More fundamentally, there is a very basic problem with theoretical explanations for how a minimum wage could fail to result in unemployment. The argument is based on the possibility of a labor market failure, in which the worker is paid less than the value of their labor. If the employee is already being paid the full value of their work, then requiring the employer to pay them more would leave the employer paying the employee more than their output was worth, i.e., losing value with every employee. An employer cannot employ people at a loss for long.
Deviations from the textbook economics model can create situations in which employees are paid less than the value of their output. One way in which this could happen would be if the employer has what is called market power. In the case of a producer market power is usually the result of monopoly or oligopoly; in this case it would be the consumer of the good (labor) with the market power, and this is called either monopsony or oligopsony. The classic case of monopsony would be the company town, a labor market in which there is only one consumer of labor, the single employer in town. Oligopsony would be the case of collusion between employers, in which the consumers of labor would cooperate to keep wages low, and either co-opt or exclude any potential competitors that might try to break into the labor market and bust the cartel.
In the case of a company town, monopsony would be a significant concern, but few areas still exist in the US that could qualify as company towns. Mobility of labor effectively prevents workers from becoming tied to a single employer, and mobility of capital prevents a single employer from dominating any given community. An employer that tried to pay wages below the value of labor would quickly find competitors moving into town to take advantage of the abnormally cheap labor.
Collusion between employers is similarly implausible in a modern economy. It is highly unlikely that Wal-Mart, for example, would collude with Target to underpay their employees, but it is unimaginable that they would also collude with McDonald's, Burger King, Hilton, Marriott, and the International House of Pancakes.
There are phenomena other than oligopsony which can cause markets to become inefficient. Jobs are not perfect substitutes for one another, and to the extent that employers are not directly competing with one another for employees, they would have market power. The weakness of this argument is that the low-wage jobs we are considering are actually close to being substitutes for one another. High-end jobs requiring a great deal of specialized knowledge and skills are far more segmented than low-wage work that requires much less specialized knowledge or skills. The sheer size of the market at the low end is also a closer approximation to the theoretical ideal of perfect competition.
Many of the ways in which labor markets are less than perfectly competitive may actually work to hold wages above their market-clearing levels, rather than below. Efficiency wage theory describes a number of ways in which it could be advantageous to firms to pay above-equilibrium wages. Turnover costs are substantial for employers -- it can cost a substantial portion of an annual wage to replace a worker, and it is likely more economical to simply pay an employee more than they are worth to avoid turnover costs. There is also evidence that paying employees above market-clearing rates can improve their productivity, possibly because of the psychological effect of gratitude, or perhaps because the resulting unemployment (caused in the same way as that resulting from minimum wages) makes it more plausible that an employer would be able to quickly and easily replace any under-performing employee with one of the similarly qualified unemployed.
A second set of arguments for the existence of and increases in the minimum wage is focused on distributional justice. These arguments may recognize that absent redistribution, a person's income is determined by the value of their output -- the value of a person's work cannot be improved by fiat. A minimum wage is thus simply another form of income redistribution. Even if the minimum wage leaves society as a whole worse off, its redistributive effect makes it worth the loss of societal wellbeing.
The question of how much societal welfare to sacrifice for a given amount of reduction of inequality is a difficult one, possibly even impossible to answer definitively. Fortunately, it is not necessary to know the answer, because there are improvements available in how we reduce inequality. A minimum wage, as a redistribution of income, rewards low-income workers, and is paid for by their employers. It is a combination of a wage subsidy for low-income workers and a tax on their employers. Transferring wealth to low-income workers is just fine, but it is perverse to impose an unfunded mandate that penalizes companies for employing the unskilled. We can agree that there ought to be a subsidy for low-income workers, but we should acknowledge that a minimum wage is only one possible way to implement it, and that other ways to pay for such a subsidy are available. If McDonald's employs many minimum wage workers, and Microsoft employs few, that is no reason to force McDonald's to fund the redistribution of wealth and allow Microsoft to dodge the costs. Even broadening the sources of funds to include all employers is still too narrow. There is no reason for us to consider Microsoft any more responsible for the alleviation of poverty than Paris Hilton. All we do by imposing the costs on only the employer is to disguise our redistribution as earned income.
There might be an argument for imposing the costs of wage subsidies on the employers of low-wage workers if we thought that the employers were responsible for the low incomes. But low incomes are the necessary result of low value of labor, which is in turn the result of poor education, disadvantaged backgrounds, and other circumstances that largely determine the value of a person's work long before they enter the labor market. The employers of low-income workers are often excoriated as exploiting their employees, but such accusations have a surplus of florid rhetoric and a deficit of substance. In fact the employers of low-income workers are among the only participants in the economy doing anything at all that benefits those workers -- they're paying them.
American society appears to be in consensus that we should assist low-wage workers. If we are to continue to subsidize their low wages, the money has to come from somewhere. Rather than foist the responsibility on some of the only people already doing anything to help them, it makes far more sense to broaden the burden as widely as possible, and fund wage subsidies out of general government revenue. These considerations of justice aside, the lowest deadweight loss to the economy results from the broadest possible funding of poverty alleviation. In fact, we are already doing just that, by means of the Earned Income Tax Credit. The EITC is a tax credit that effectively acts as a wage subsidy, and in some cases can even result in negative total tax liability. It is certainly not perfect, but it acts as arguably the largest poverty reduction program in the country, with $36 billion distributed in 2004. If Congress really want to help the poor, they should be expanding the EITC, not increasing the minimum wage.
Minimum wages almost certainly increase unemployment among the poorest members of society. They certainly injure the employers whose businesses rely on unskilled labor. Those people who are better off as a result would be better served by other poverty programs, which would help more people and impose less of a burden on society as a whole. A minimum wage may play well in a sound bite, but to genuinely help people, there are much better options.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2007-02-15 03:53 pm (UTC)By which you mean it's essentially replacing welfare benefits for people who work, but still don't make enough to live on? It seems like that would still cost the government less, even laying aside the incentive reasons...
no subject
Date: 2007-02-15 04:01 pm (UTC)All that said, though, the EITC is so much simpler than almost any other anti-poverty program that it's still likely one of the most efficient around.