Wednesday, October 11, 2017

Minimum Wages in Seattle

Minimum Wage Increases, Wages, and Low-Wage Employment: Evidence from Seattle.” Ekaterina Jardim, Mark C. Long, Robert Plotnick, Emma van Inwegen, Jacob Vigdor, Hilary Wething, National Bureau of Economic Research, June 2017

In yet another study of the minimum wage six authors tell readers they intend to evaluate the wage, employment and hours effects of a first and second phase in of the Seattle Minimum wage ordinance. The first phase raised the minimum wage from $9.47 an hour to $11.00 an hour on April 1, 2015. The second phase raised the wage from $11.00 an hour to $13.00 an hour on January 1, 2016. They analyze “employment in all sectors paying below a specified real hourly rate.”

The paper’s opening sentence starts with the standard obsessions economists always cite against minimum wages: “Economic theory suggests that binding price floor policies, including minimum wages, should lead to a disequilibrium marked by excess supply and diminished demand.” Economists predict a raise in the minimum wage will reduce employment of those earning a wage lower than the minimum wage. The see cause and effect as part of their doctrine.

They conclude the first phase effect was smaller than the second phase, which second phase caused a decrease in hours worked in low wage employment by 9 percent while the wage of low-wage workers’ was about 3 percent so that the cost of this wage hike outweighed its benefits for these workers. They conclude the minimum wage hurts low wage labor because hours lost makes a loss bigger than the gain from higher wages.

People leave jobs and lose jobs for many reasons, especially in low wage employment where turnover rates can be high. If the Seattle minimum wage causes employers to decrease employment, it could be useful to go out and ask these low wage employers if they recently off employees and was that because of the higher minimum wage. Typically Economists resort to analysis using large data sets filled with severe shortcomings like the Seattle study I review here.

Their data set comes from Washington’s Employment Security Division, which is produced as part of national Unemployment Insurance(UI) system administered by each state. Old timers refer to it as ES 202 data, or just “the 202” data. It is compiled and used by the Bureau of Labor Statistics in their benchmark revision of the Current Employment Survey.

ES 202 data is reported by single or multi establishment within county and Metropolitan Statistical areas coded by industry using the governments North American Industry Classification System(NAICS). Public reports of the data have monthly employment and payroll totals, but there are no occupations reported and names of employers or employees remain suppressed and confidential.

The authors inform readers that the Employment Security Division provided them the total hours worked in addition to the employment and payroll totals. Further the Employment Security Division partitioned the Seattle-Bellevue-Evertt Washington Metropolitan Division data to break out Seattle as a special and private favor to them, perhaps from their connection with the University of Washington. Because their data is a special favor and confidential the authors were required to sign a document promising not to release the data under threat of legal action against them. Therefore no one else gets to look at the data; they provide only a summary of aggregated data by quarter in their Table 3 on page 45.

They state “This unique data set allows us to measure the AVERAGE wage paid to each worker in each quarter. We compute an hourly wage rate as total quarterly payroll divided by quarterly hours worked, which corresponds to average hourly earnings. They call these numbers a realized hourly wage rate. Therefore they use an average wage of thousands of employees, an amount no one actually earns. Actual wages paid to employees will be above and below the average.

In addition their data excludes those working at establishments with more than one location. These include a variety of chain stores and franchise restaurants in Seattle and the surrounding county and metropolitan areas. Seattle’s minimum wage for a business with 500 or more employees such as McDonald’s or Costco was $13.50 an hour during the time when smaller single establishment business had an $11 an hour minimum wage. The employees included in the study had a strong incentive to move out of small business and into the large businesses excluded from the study.

In addition they define low skill employment as those working with an average hourly wage rate of $19 an hour or less. While they give an excuse for doing this, they do so without knowledge of the occupations of the employees included in the sample or the skills, experience or education needed in the unknown occupations that justify such a decision. An establishment with an average wage of $19 an hour will have many earning wages above $19 an hour, which could be occupations that need college degree skills.

In their methodology at page 16 the authors admit the hazards of their partition at $19. They write “The proxy for low-skilled employment will produce accurate estimates of the impact of minimum wage increases to the extent that a wage threshold accurately partitions the labor market into affected and unaffected components.” Their partition comes at an AVERAGE wage causing some who work with an actual wage above $19 an hour to be included in the below $19 partition while others will be in the below $19 partition who have wages above the $19 partition. There can be no assurance the 9 percent decline in total hours they cite ever worked an hours below the minimum wage or lost their job because of it.

Further they state “[The threshold wage] will overstate employment reductions if the threshold is set low enough that the minimum wage increase causes pay for some work to rise above it. This concern is particularly relevant given previous evidence of "cascading" impacts of minimum wage increases on slightly higher-paying jobs.” The previous evidence of “cascading impacts” comes from several Neumark and Wascher studies and a book, one of which is reviewed on this link at The cascading impact terminology refers to people who lose their below minimum wage job, but rather than be out of work they apply for and find work at a higher wage.

As I have suggested before people who lose their minimum wage job do not disappear, but begin looking for other jobs in other occupations with wages higher in the wage distribution. Forced to leave a sub-minimum wage job the newly unemployed increase the supply of labor in other occupations where they moderate wages in higher wage occupations and add to employment. The authors might recognize the millions of opportunities to move from low wage to higher wage employment by looking at wage distributions by occupation reported by the U. S. Bureau of Labor Statistics in their Occupational Employment Survey.

Their Summary of data in Table 3 supports this view. The table has three columns for total jobs, total hours and total payroll and a fourth column has computed average wages. The rows are for each quarter from the second quarter of 2014 to third quarter of 2016. The first column of each category has only those employers with employees that have an average monthly wage of $13 an hour or less. The second has only those employers with employees that have an average monthly wage of $19 an hour or less. The third set has an average monthly wage for all employers and employees.

These partitions allow a further partition into two additional columns through subtracting the less than $13 column from the less than $19 column, which leaves only those establishment employers with an average wage greater than $13 an hour and less than $19 an hour. Further subtraction leaves another column of only those employers with an average wage of $19 or more. Every employee and his or her employer is part of one and only one mutually exclusive column of the data.

These columns have the “cascading effects” but they show the benefit of the Seattle Minimum wage. In the second quarter of 2014 those working at establishments with an average wage less than $13 an hour total 39,807. By the third quarter of 2016 the total falls to 23,232, a loss of 16,575 working at establishments with an average wage less than $13 an hour.

Over that two year and one quarter period a low inflation rate combined with the higher minimum wage would tend to reduce people working at establishments that have an average wage below $13 an hour. Economists like to suggest that is a bad result caused by the minimum wage, but during the same period those working at establishments with an average wage above $13 and below $19 increased from 53,152 to 63,610, a gain of 10,548 jobs at above the minimum wage. Those working at establishments with an average wage above $19 increases from 199,681 to 249,675, a gain of 49,994 jobs. These are exactly what to expect if the minimum wage benefits low wage workers. In Seattle wage workers seek employment in other establishments in occupations that pay above the minimum wage.

In their 2008 book Neumark and Wascher Minimum Wages, cited by the authors in their Seattle study write on page 116, “. . . as we emphasized earlier in this chapter the potential for minimum wage increases to affect wages higher in the wage distribution is also important in assessing the effects of minimum wage policy.” It is. That is where the benefits of the minimum wage will be and that is where they are in Seattle.

This paper has no right to be a part of the public debate on minimum wages because it makes no attempt to persuade a general audience and cannot be read except by those with experience in the specialized terminology of the economics fraternity. It uses suppressed data and undefined insider terms from other studies such as region fixed effect, period fixed effect, treatment effect, idiosyncratic shock among other terms.

Business predictably opposes an increase in the minimum wage. It raises costs for businesses that depend on low wage employment and thereby pressures owners and managers to experiment with prices, jobs and work schedules. It might in some situations reduce long term profits, but that does not mean a higher minimum has no benefits to labor or the larger economy from those who will have more buying power.

Academic economists work under pressure to confirm market theory. When they do what is good for their career, the news media and the public seize on the conclusions and nothing else. They evaluate the conclusions based on academic credentials not the credibility of the work.

In Seattle I read the mayor and city council ignored the hecklers and went ahead with the next phase of their minimum wage program; they raised the minimum wage to $15 an hour. If I could get the suppressed employment data I could determine the benefits to labor and the economy I predict will continue.

No comments: