Regulation by (Bad) Proxy: How Selective Application of Transaction Cost Economics Tainted the FTC’s Proposed Ban of Employee Noncompete Agreements
Abstract
Agencies have imperfect information about conduct they regulate. This problem is particularly acute when identical conduct has differing effects in various markets. Determining the economy-wide impact of such conduct can be difficult or impossible.
The FTC faces such a challenge. The Commission has proposed a rule banning the nation’s 30 million employee noncompete agreements (“NCAs”) as unfair methods of competition under Section 5 of the FTC Act. The Commission determined that NCAs reduce aggregate wages, establishing a presumptive violation. The Commission also found that nearly all NCAs are both procedurally coercive — because employers use overwhelming bargaining power to impose them — and substantively coercive — because they restrict employees from starting new firms or accepting offers from rival employers. The Commission also implied that procedural coercion was a necessary condition for substantive coercion.
The Commission then assessed possible business justifications. Echoing Transaction Cost Economics (“TCE”), the Commission concluded that NCAs sometimes produce cognizable benefits, increasing productivity and product quality. The Commission framed the inquiry as assessing whether, “overall,” NCAs’ harms exceed benefits. The Commission subjected justifications to a “high bar,” given its finding that nearly all NCAs are doubly coercive.
Determining the overall impact of 30 million contracts is a daunting task. The Commission employed a creative proxy, however. The Commission hypothesized that employers would share benefits of NCAs by paying premium wages to employees with such agreements. However, most studies find a negative correlation between state-level enforceability of NCAs and wages, implying that harms exceed benefits. The Commission therefore rejected justifications and indiscriminately condemned all NCAs.
This proxy seems sound and consistent with TCE. Wages impound vast data generated by innumerable decisions. Resulting wages should reflect benefits employers expect from NCAs as well as the harms resulting from their restrictive impact. This proxy would seemingly generate an economical assessment of the net impact of NCAs.
This essay critiques this proxy and rejection of business justifications. The essay contends that the proxy may produce misleading results reflecting selective application of TCE’s model of contract formation. For instance, the proxy could produce false negatives. A positive correlation between enforceability and wages is consistent with two conclusions: (1) NCAs produce net benefits or (2) most raise rivals’ costs and injure consumers. Absent additional information, both hypotheses would be equally plausible. Immunizing conduct because of a positive correlation between enforceability and wages risks entrenching harmful agreements.
Invocation of a negative correlation between enforceability and wages risks false positives. Markets are not pre-legal entities that generate immutable results. Background rules impact transaction costs and resulting market activity. The Commission implied that transaction costs prevent employees from learning of NCAs before accepting offers and assumed that courts enforce NCAs regardless of precontractual knowledge. These two aspects of the institutional framework will prevent employees from receiving wage premia to compensate for NCAs. Wages will not impound the benefits of NCAs, and condemnation because of a negative correlation between enforceability and wages may produce a false positive.
The prediction that a wage-based proxy can produce false positives assumes that Section 5 is indifferent between whether employers or employees capture the benefits of NCAs. If, however, benefits must offset harms that NCAs impose on employees, the wage-based proxy will produce no false positives. Unfortunately, the Commission did not address whether Section 5 requires such sharing.
Even if business justifications fail, blunt condemnation of NCAs is not the only remedy that can enhance employee welfare. The Commission rejected an alternative derived from TCE, namely, banning NCAs not disclosed in advance. TCE teaches that this change to the institutional framework would reduce transaction costs and induce employers to pay premium wages to employees bound by NCAs, thereby sharing the benefits of such agreements. These higher wages would also force employers to internalize the impact of NCAs on employees. Some employers would abandon NCAs, and some others would narrow their scope. Both effects would reduce the restrictive impact of NCAs, mitigate NCAs’ negative impact on wages, weakening the Commission’s prima facie case against such agreements.
Moreover, fully-disclosed NCAs that produce net benefits or raise rivals’ costs are voluntary. Thus, neither category of agreement is procedurally or substantively coercive. Mandatory disclosure would thus reduce the proportion of NCAs that are coercive in either sense, further weakening the Commission’s prima facie case.
Despite mandatory disclosure, NCAs’ harms may still exceed the benefits that employers share with employees. However, the Commission could no longer apply a “high bar” to efforts to justify all nonexecutive NCAs and would instead have to estimate how many such agreements are voluntary, lowering the bar for those that are. The combination of a weakened prima facie case, lower bar for some NCAs and increased sharing of benefits could well alter the outcome of a comparison of NCAs’ costs and benefits.
Indeed, the Commission need not guess about the impact of changing the institutional framework. Such a change could itself inform empirical tests that would determine whether the benefits of NCAs realized by employees exceed harms in well-functioning labor markets. In such markets, wages would be a more accurate proxy for the overall impact of NCAs. The result could be a conclusion that, “overall,” NCAs produce more benefits than harms and that employers share a sufficient portion of such benefits with employees such that NCAs improve employee welfare compared to a regime that indiscriminately bans such agreements.