My latest paper (with the excellent Brandon Pizzola) is on occupational licensing in the funeral services industry. Almost all of the previous work on occupational licensing has used cross-sectional data, comparing outcomes in states that license an occupation with outcomes in states that do not. Since many factors vary between states it’s difficult to be sure whether those studies are identifying causal effects. Pizzola and I take advantage of a unusual change, Colorado delicensed its funeral service industry in 1983. The time-series variation combined with the cross-sectional variation lets us examine and test the data in many ways.
In 1983, Colorado delicensed funeral services….the results from difference-in-differences, difference-in-difference-in-differences, and synthetic control specifications suggest occupational licensing causes a wage premium of 11-12 percent.
Importantly, we also do a cross-sectional test similar to those that have been done before in other industries and that test is also consistent with a wage premium of 11-12 percent. In other words, our paper makes all the previous papers on occupational licensing that use cross-sectional data more credible.
We find similar results from a standard cross-sectional wage regression using data on individuals in 1990. Thus, this suggests that cross-sectional regressions of wages on occupational licensing in other industries are a good baseline estimate of a causal effect.
Finally, consistent with an earlier paper by David Harrington and Kathy Krynski that used cross-sectional data, we find some evidence that licensing, which requires training in embalming, increases prices even more than the wage premium alone would suggest because under licensing consumers appear to be pushed away from cremation and towards more expensive burial.
Consistent with Colorado’s decision to delicense in 1983, we find no evidence that delicensing reduced quality in the funeral services industry.