November 7, 2013
By Oren M. Levin-Waldman
Economics has long been divided into two warring camps: the neoclassical economists who argue that wages are set by supply and demand in competitive markets and heterodox economists who maintain that wages are also determined by the power dynamics between workers and management. In the neoclassical model, market clearing wages are achieved when the demand for labor is exactly equal to the supply of labor.
In such a market, there is no such thing as unemployment because wages either rise or fall until the demand for labor is exactly equal to the supply of labor. At the wage at which demand equals supply, all those willing and able to work at that wage will be employed. If more people are willing to work, the wage will fall further, thereby inducing firms to hire more workers, with the result being that the supply of labor will once again be equal to demand. Conversely, firms that are unable to hire as many workers as they would like are forced to raise wages as an inducement for additional workers to enter into labor market until supply and demand are again equal.
Unemployment, according to this model, is the result of market interventions, whether in the form of social programs like public assistance and unemployment insurance that causes moral hazard or wage floors and labor laws that promote collective bargaining because they artificially inflate wages above market clearing levels, thereby limiting the flexibility of workers to demand less. The heterodox, or at times referred to as the institutional, model categorically rejects this position.
On the contrary institutions, particularly labor market institutions, do matter. To paraphrase Bill Clinton’s winning campaign slogan in 1992: “It’s institutions stupid”!!! Employers are effectively able to pay their workers less, especially those lacking in skill, because they do not have market power. Institutions like labor unions give their members a measure of bargaining power which, in the absence of organizing, they would not have. Similarly, a wage floor or minimum wage offers unorganized workers in the low-wage labor market a measure of market power too. Conversely, right-to-work laws, which we might refer to as anti-labor market institutions, work to suppress wages.
Data from the Current Population Survey for 2012 shows that median individual income in the U.S. was $31,000 a year. In right-to-work states median individual income was only $30,000, while it was actually $32,000 in high union density states. On the face of it, median individual earnings in high union density states is 6.7 percent higher than in right-to-work states. Now consider that in 1992 the median individual income was $17,000 with median income being $15,200 in right-to-work states and $18,335 in high union density states. In 1992 when union density was greater overall (16.0 percent) than in 2012 (11.3 percent), median wages were 20.6 percent higher in high union density states than in right-to-work states.
This too would suggest that the 29.4 percent decline in union density from 1992 to 2012 had the effect of decreasing wages in high union density states relative to right-to-work states. Stated differently, the increase in median wages between 1992 and 2012 was 89.9 percent in high union density states compared to 97.4 percent in right-to-work states. Although it is speculation, it might be reasonable to speculate nonetheless that had there not been a drop in union density the percentage increase in median wages would have been much higher in high union density states.
Institutions also affect levels of income inequality. Income inequality was lower on the basis of the ratio of the 90th percentile to the 10th percentile in high union density states in 1992 than the nation as a whole. Although income inequality increased 21.9 percent from a 90/10 ratio of 10.5 in 1992 to 12.8 in 2012, income inequality tended to be lower in high union density states. Data for 1992 and 2002 shows the 90/10 percent ratio to be lower in high union density states than in low union density states. In 2012, the 90/10 percent ratio was higher in high union density states than in low union density states, which may be attributable to diminished union density. In 2002 union density was still 13.5 percent, which was 19.5 percent higher than it was in 2012. Of particular interest, however were the 50/10 percent ratios. These ratios also dropped between 1992 and 2012, but again they were lower in 1992 and 2002 in high union density states than in low union density states. And again they were higher in high union density states in 2012 than in low union density states, which again may be attributable to declining union density overall.
Although this is a rough cut presentation of data, more statistical testing does show that states with high union density are more likely to have lower levels of income inequality than the country as a whole, at least during 1992 and 2002 when unions were stronger. It is quite telling, then, that states with high union density are likely to have higher inequality in 2012 when unions appear to be at their lowest level over a two decade period of time.
If we want to boost the wages of working Americans, we need to reinvigorate those labor market institutions that make a difference. Unions are but one form of wage policy; minimum wages, of course, are another. Workers need a constituency that will speak for their interests and afford them a measure of market power necessary to redress the power imbalance that is inherent to the labor-management relationship. But it isn’t just a question of redressing a power imbalance, but of strengthening the economy overall. The assumptions of the neoclassical model, which have lead to the pursuit of a low-road strategy, at the end of the day cannot lead to job creation. Rather job creation stems from the grassroots level whereby workers are able to demand more goods and services because their incomes are rising. Unemployment is the result of declining demand for goods and services; not labor inflexibility. Institutions, in other words, matter.