We also analyzed the impact of adopting right-to-work laws in Michigan and Indiana using county-level data from these two states and their surrounding neighbor states. Our analysis found positive results from adoption of right-to-work laws in Michigan, although many of the results did not meet the formal standard for statistical significance. For Indiana, similarly positive results were found and many did meet this standard.
Measures of the statistical significance of findings provides researchers with confidence that their findings are distinct from random fluctuations in the data. We recognize the value of these tests but will nonetheless include some state-specific findings that did not pass these conventional tests for statistical significance.
We also recognize that an increasing number of scholars have made the case that social scientists may place too much emphasis on statistical significance and thus overlook potentially valuable findings in research. The lack of statistical significance in our state-specific example may simply be a function of the limitations in our dataset due to the relatively recent adoption of right-to-work in Great Lake States. The observations in our dataset have only had about six years to be influenced by the adoption of right-to-work.[*]
That said, the model used for this Michigan-specific analysis, including data from Michigan, Indiana, Ohio and Wisconsin, found that adopting right-to-work was associated with an increase in manufacturing employment as a percentage of total private employment by about 26.1% in the typical Michigan county, although it lacks statistical significance.
The spillover effect on counties in non-right-to-work states was to reduce the manufacturing employment share by 30.0%, a result that was statistically significant. This suggests that, while overall manufacturing employment as a share of private employment was statistically unchanged when averaged out across the entire four-state region, the growth that did occur mostly happened in counties in right-to-work states, likely at the expense of counties in non-right-to-work states.
Similar patterns are found in other industrial sectors. Right-to-work increased employment shares in the construction and the transportation and warehousing industries in right-to-work counties: They were 30.0% higher in construction and 42.5% higher in transportation and warehousing. In non-right-to-work counties, right-to-work reduced employment share in these industries, by 32.8% in construction and 68.6% in transportation and warehousing.
Across the entire four-state region, right-to-work decreased employment share in manufacturing by 3.8%, but this was not statistically significant. The total effect on transportation and warehousing from right-to-work was also not statistically significant, however, the magnitude was large and negative. Our findings suggest that total transportation and warehousing employment share in the region is lower by 26.0% as a result of right-to-work.
Full results for Michigan are shown in table below.
Graphic 2: Marginal Effects for Michigan Based on Estimation Resultsfrom Spatial Autoregressive Model, 2018
The Indiana-specific analysis compares county-level data in the Hoosier State to Michigan, Illinois, Kentucky and Ohio. Of these states, neither Ohio nor Illinois have right-to-work laws, and Kentucky passed such legislation just in 2017. In the industries studied, right-to-work increased employment share in Indiana, but it decreased employment in non-right-to-work counties.
In manufacturing, employment as a percentage of total private employment was 27.3% higher in the typical Indiana county. The economic significance in Indiana is very similar to that of Michigan, but unlike the results from the Wolverine State, this finding is statistically significant. The net impact on manufacturing employment in the five-state region is to increase manufacturing employment by 19.6%. This finding, too, was statistically significant.
The impact in the construction industry of right-to-work in the five-state region was to lift employment share by 19.5% in Indiana counties. This effect swamps industry employment losses in non-right-to-work counties in the region, where employment share was 3.0% lower. These findings were not statistically significant.
In transportation and warehousing all effects were statistically insignificant. Employment share was 6.7% higher in Indiana counties, but significantly lower in non-right-to-work counties: 33.1%, specifically. The total effect across the region was also negative, with total industry employment for the typical county in the five-state region down by 26.4%, nearly identical to the estimate for the Michigan sample.
Full results for Indiana are shown in table below.
Graphic 3: Marginal Effects for Indiana Based on Estimation Resultsfrom Spatial Autoregressive Model, 2018
Right-to-work laws have shown themselves to be useful economic development tools. Our new study combines natural controls of cross-border analysis with other controls to help isolate the jobs impact that right-to-work laws have on 18 industrial sectors, the most notable of which is manufacturing. The results are largely positive, even, in some cases, for non-right-to-work counties just across the border from states with right-to-work protections.
[*] Some scholars suggest that it may take 10 years’ worth of data to detect meaningful differences in statistical outcomes between right-to-work and non-right-to-work states.