new “study” from the left-leaning Economic Policy Institute (EPI) purports to show that Minnesota has had better economic outcomes under liberal Democrat Governor Mark Dayton than conservative Wisconsin has had under Republican Governor Walker. The implicit conclusion is that liberal policies are better for economic growth than conservative ones.

While the analysis is framed as a “natural experiment” between the economic outcomes in similar states with governors of different ideologies and priorities, the conclusion does not hold up under scrutiny. Put simply, this surface-level analysis, given major attention in the media, represents the sort of shoddy work that gives think tanks a bad name.  Here’s why:

No Control Variables.  It is not a stretch to say that Wisconsin and Minnesota are similar.  It is a stretch, however, to say that they are the same. And that is exactly what EPI does. For example, Wisconsin is a state that has been historically heavily dependent on manufacturing for its economy. Minnesota, on the other hand, has a growing healthcare sector that employs more people than manufacturing. Wisconsin also has a significantly lower percentage of residents with a 4-year college degree, something that is important to consider in an era when many new jobs require such a degree. The largest cities are not the same. The Minneapolis-St. Paul metro area has experienced significant population growth. The Milwaukee metro areas growth has been much more modest. Most importantly, these trends long pre-date the administrations of Dayton and Walker.

These differences matter. Serious academic studies, like WILL’s study of the effects of Act 10 on teachers, would take these differences into account in order to put states on a level playing field before conducting policy analysis. Controlling for differences must be made before making broad-based conclusions about the impact and effectiveness of public policy.

Oversimplification of Policy under Governors of Different Parties. The EPI analysis also fails the smell test by making broad-based claims about the general agenda of conservative Governor Scott Walker in comparison to his liberal counterpart in St Paul, Mark Dayton. While it is certainly true that Governor Walker has pushed a conservative agenda forward on many issues and Dayton has pushed a more liberal one, such surface-level comparisons are more the stuff of a bar-room argument than what purports to be a serious analysis. In order to determine the impact of Governor Walker’s policies, one should look at specific reforms and specific, related outcomes: the impact of Act 10 on the teaching workforce; or the impact of right to work legislation on employment.

Selectivity in Statistics.  Another issue with EPI’s analysis is that the authors pick and choose the statistics they want to report in order to conform to push a certain agenda. If one is doing a surface level analysis, such as EPI, one could just as easily choose statistics to show how Wisconsin is doing better than Minnesota. For example, Wisconsin’s record low unemployment rate of 2.8 percent is lower than Minnesota’s of 3.2 percent. Real gross domestic product (GDP) growth has grown faster in Wisconsin than Minnesota since Walker took office. Of course, as described in #1 and #2 above, we would want to control for the variables and ideally look at specific policies before forming any conclusions.

Analyses, like EPI’s, do little to inform the debate. Instead, they muddy the waters with a set of liberal talking points masquerading as a study.

Will Flanders

Dr. Will Flanders is the Research Director for the Wisconsin Institute for Law & Liberty.