Saturday, August 5, 2017

Seattle Minimum Wage Study Data and Methodologoical Problems

 Today, the State Journal-Register published a Guest Column editorial which cited a recent Seattle minimum wage study, which can actually be found in full at NBER via the link.  The Guest Column is a typical ideological, propaganda mish mash advocating against any raising of the minimum wage ($7.25 nationally, $8.25 in Illinois), which, if the national minimum wage had kept pace with inflation, would be $10.10, which is less than a living wage for one adult in Springfield, Illinois (Sangamon County), which is a relatively inexpensive metropolitan area, as can be seen with this Living Wage calculator from MIT (choose state then county).

While the University of Washington study has stirred attention and deserves consideration, it has also been the subject of critical economic analysis which has raised serious data limitation and
methodological concerns.  One aspect of the study would seem to indicate employers may have reduced workers to part-time work to levels which would probably disqualify them for employee benefits; this requires more rigorous study to see if that is true.

However, some of the findings appear to indicate statistical problems with the cohort (control group) used, the time lines of the different data, and unrelated pay raise patterns at the same time not associated with minimum wage (there was a surge of above $19/hour hiring in Seattle).  It would appear that the study has at least three very serious problems which need to be addressed in further study as enumerated in an Economic Policy Institute report:
1) "The employment responses estimated by the authors are well outside the bounds of most published research,"
2) "The study implausibly finds employment changes due to the minimum wage in parts of the labor market where there should have be none", and
3) "The study excludes an important group of workers, representing roughly 40 percent of the workforce: those working for employers with businesses in multiple locations".

I find the data limitation excluding 40% of Seattle employers from the study very disturbing since those employers all had multiple locations in Oregon (think McDonalds, Starbucks, etc).  Given the data limitations, control group problems, lack of testing for extraneous data, and data time lines, this study is important in pointing to how to to do better studies going forward.

As to the Guest Column in the State Journal-Register, some people have just been unable to accept the abolition of slavery.

Print Page

No comments:

Post a Comment

Share This