South Africa has very high unemployment, yet few adults work informally in small firms. One explanation is that unions extend arbitration decisions to non-unionized firms, raising wages. These agreements are enforced in a spatially discontinuous way; employment effects of these bargaining councils are identi?ed through spatial fixed effects. This approach represents a methodological improvement over sample restrictions used in other spatial discontinuity studies. Bargaining councils are found to decrease employment by 6-11%, with larger decreases among small Firms. These effects are not explained by resettlement to uncovered areas, and are robust to a wide variety of forms for average spatial heterogeneity.