As part of their efforts to reduce fiscal deficits, many governments have allowed public sector salaires to erode, often on the assumption that government workers are overpaid vis-a-vis those in the private sector. We test that assumption by analyzing public-private pay differentials in the Ivory Coast and Peru. Switching regressions models are estimated using full information maximum likelihood (FIML), and the results are compared to those obtained useing ordinary least squares (OLS) techniques. The OLS yields seriously biased estimates of the pay structure, suggesting that public wages are higher than private wages; the FIML estimates show the opposite. Our probit analysis also shows that the wage disadvantage of civil servants is a determinant of the greater prevalence of moonlighting among public than private employees. The evidence suggests that reductions in employment rather than pay, while being less palatable in the short term, will be more effective in the long run. Copyright 1989 by Oxford University Press.