Noncompete agreements, bargaining, and wages: evidence from the National Longitudinal Survey of Youth 1997

Abstract

We examine the use of noncompete agreements (NCAs) and their relationship with wage bargaining and wage outcomes using new data from the National Longitudinal Survey of Youth 1997. NCAs cover 18 percent of the workers in our sample, and adoption patterns are broadly consistent with prior research. The NCA–wage correlation is positive and highly sensitive to controls for demographics and job characteristics, suggesting selection into NCAs causes positive bias in the estimates. While it is not obvious what the baseline level of the NCA–wage differential is, some heterogeneous effects are more stable: the NCA–wage differential is lower for workers who do not bargain over wages, have less education, have lower ability, or live in a state that enforces NCAs. Notably, wage bargaining—which is only marginally more likely with NCAs in our most saturated model—does not explain the heterogeneous effects across subgroups. We discuss these findings in light of competing theories of the social value of NCAs.