President Biden’s proposal to raise the minimum wage to $15 an hour by 2025 is prompting a backlash from the usual suspects. As we hear the cries about how this will be the end of the world for small businesses and lead to massive unemployment, especially for young workers, minorities, and the less-educated, there are a few points worth keeping in mind.
While $15 an hour is a large increase from the current $7.25 an hour, this is because we’ve allowed so much time to pass since the last minimum wage hike. The 12 years since the last increase in the minimum wage is the longest period without a hike since the federal minimum wage was first established in 1938. Few workers are now earning the national minimum wage, both because of market conditions and because many states and cities now have considerably higher minimum wages.
If the minimum wage had just kept pace with prices since its peak value in 1968 it would be over $12 an hour today and around $13.50 by 2025. Keeping the minimum wage rising in step with prices is actually a very modest target. It means that low-wage workers are not sharing in the benefits of economic growth.
From 1938 to 1968 the minimum wage rose in step with productivity growth. This means that as the economy grew and the country became richer, workers at the bottom of the ladder shared in this growth. If the minimum wage had continued to keep pace with productivity growth it would have been over $24 an hour last year and would be close to $30 an hour in 2025.
There has been considerable research on the extent to which the minimum wage leads to job loss. Much recent research finds that even substantial increases in the minimum, such as the $15 an hour minimum wage that is already in place in Seattle, have no effect on employment.
It is worth noting that even the research that finds the minimum wage reduces employment generally finds a relatively modest effect. A recent review article by prominent opponents of the minimum wage found that the median estimate of elasticity was -0.12 for affected workers. This estimate means, for example, that a 10 percent increase in the minimum wage would lead to a reduction in employment among affected workers (e.g. workers with less education or young workers) of 1.2 percent.
It is important to realize that even in this case we are not talking about 1.2 percent of affected workers going unemployed. Low-wage jobs turn over rapidly. For example, in a typical month before the pandemic hit, more than 6.0 percent of the workers in the hotel and restaurant industries lost or left their jobs. If we take the elasticity estimate of -0.12, it would mean that at a point in time we have 1.2 percent fewer people working in the sector as a result of a ten percent increase in the minimum wage.
Carrying out the arithmetic, this means that an average low-wage worker would be putting in 1.2 percent fewer hours in a year, but getting 10 percent more money for each hour they worked. That would mean that they would be pocketing roughly 9.0 percent more in wages each year. And, this calculation assumes there is an employment effect, ignoring considerable evidence that there is none.
A higher minimum wage also has positive societal effects. A recent review of the literature found that a 10 percent increase in the minimum wage would reduce the poverty rate by 5.3 percent. Another study found that a 50 cent increase in the minimum wage reduced the likelihood that formerly incarcerated people would return to prison within a year by 2.8 percent. The long-term effects of these and other benefits are likely to be quite large.
Finally, it is worth remembering that there is a lot of money on the side of those looking to stop minimum wage hikes. This can affect the research on the topic. While few researchers may deliberately cook their results to favor the fast-food industry, they know they can get funding for research that finds a higher minimum wage leads to job loss. There is much less money available for supporting research that finds no effect. (I know that first-hand in my former capacity as co-director of CEPR.)
Probably the clearest case of such bias affecting research findings was a paper by David Neumark and William Wascher, two of the most prominent opponents of higher minimum wages. Neumark and Wascher analyzed data given to them by the Employment Policies Institute (a.k.a. “the evil EPI”), a lobbying group for the restaurant industry. They used this data to replicate a pathbreaking study by economists David Card and Alan Krueger, which found no job loss associated with a minimum wage hike in New Jersey.
Neumark and Wascher’s study found that there was in fact a significant loss of jobs in fast-food restaurants in New Jersey following the minimum wage hike. However, an analysis of the Neumark and Wascher data by John Schmitt found patterns that were not plausible. It was subsequently revealed that an owner of a number of fast-food restaurants in New Jersey and Pennsylvania (the control state) had submitted fake payroll data to the Employment Policy Institute to be used in the study. (There is no reason to believe that Neumark and Wascher realized they were working with fraudulent data.) If the faked data was removed from the analysis, the finding of minimum-wage induced job loss disappeared.
This story should be seen as a warning. Most researchers are honest and will accurately report what they find in their analysis. However, we should realize that there are some pretty big thumbs on the scale in the minimum wage battle, and those thumbs want to show that minimum wage hikes will cause job loss.
 The actual story is a bit more complicated since typically these studies look at a specific type of worker, such as young people or workers with less education. It could be the case that employment in an industry has not changed, but we have seen older or more educated workers replacing younger and less-educated workers.