The National Labor Relations Board’s (NLRB’s) decision in the Browning Ferris Industries (BFI) case was a bad idea. Recall that since 1984, a firm was defined as a “joint employer” only if it exercised “direct” control of the employees in another business. In the 2015 BFI decision, the NLRB ruled in favor of a “direct or indirect” control standard that left a lot of ambiguity and potential for litigation, and could be applied to a broader array of business arrangements. (There were further developments that supposedly got rid of the BFI ruling until, alas, they didn’t. As of this writing, “direct or indirect” is back as the standard.)

Among the businesses most affected by the BFI decision were those organized as franchises. After BFI, a firm would be less likely to sell its license and be liable for litigation costs and less likely to provide broad infrastructure support for franchisees that could be interpreted as part of indirect control. Thus, one would expect that the BFI decision would be bad for franchises. That is exactly what AAF’s Ben Gitis found in earlier research.

Gitis, along with Emma Pettit, revisit this issue in a new paper that also provides a fascinating lesson in economic research. There is an old saying in economics that there are two kinds of results: (1) those that are apparent in the raw data and hold up to statistical torture, and (2) those that cannot be discerned in the raw data but reveal themselves after the aforementioned statistical torture. Gitis and Pettit have a wonderful example of (1), which can be seen in Table 1 (below and replicated from their paper).

As the table displays, employment in hotels grew at an average annual rate of 1.9 percent before the BFI decision and 1.5 percent after. But here is where it gets fun. If one splits the data, hotels that are not organized as franchises grew at a rate of 1.9 percent before, and 2.1 percent after, the BFI decision. Of course, the BFI decision did nothing to their business model.

In contrast, employment among franchises grew at a nearly identical 1.8 percent before the BFI decision, but dropped dramatically to only 0.4 percent afterwards. The impact of the BFI decision itself is the difference between the acceleration in growth in non-franchises (0.2 percentage points) and the slowdown in franchises (1.4 percentage points) or a negative impact of 1.6 percentage points.

Table 1: Job Growth in Hotels Overall and by Franchise Status

That’s it! No need to read the entire paper. No need to wonder if the BFI decision hurt job growth. The only real need is legislation that permanently reverses the “direct or indirect” standard.