Abstract

     Despite the unambiguous predictions of the canonical model of a competitive labor market, empirical studies of the labor market effects of payroll taxation provide conflicting evidence. We estimate the labor market impacts of payroll taxation in Singapore, the country with the most competitive and flexible labor market among the countries investigated in the literature. By exploiting the sharp reduction in payroll tax rate when workers turn 60, we find that the reduced payroll tax rate in Singapore has a large effect on wages without changes in employment. Our meta-analysis shows consistent evidence that varying degrees of labor market competitiveness across places and time could explain the mixed results in the literature. Our findings corroborate the prediction of the canonical model that the welfare costs of social insurance programs financed by payroll taxes can be small in a competitive labor market.

Introduction

     Most social insurance programs, such as old-age pension and health, disability, and unemployment insurance, are financed by payroll taxes (or social security contributions). These payroll taxes often account for more than 30% of workers’ labor income and about a quarter of the total tax revenue of OECD member countries (Saez et al., 2019). Although social insurance programs play an important role in improving social welfare, the payroll taxes levied to finance those programs can distort the labor market.

     In the canonical model of a competitive labor market, workers have an incentive to increase (decrease) their labor supply depending on how they value their social insurance benefits when the payroll tax rate increases (decreases). Thus, employers’ additional labor costs due to higher payroll taxes can be shifted to workers through reduced wage payments. If workers regard the payroll taxes borne by their employers as their social insurance benefits, the incidence of payroll taxes will fall solely on workers’ wages. This leaves employment or work hours unchanged, thereby minimizing deadweight loss in the labor market (Summers, 1989).

     Several empirical studies have tested Summers’ (۱۹۸۹) insights into the labor market impacts of payroll taxes to finance worker benefits across various countries and time periods and found mixed results. One possible explanation to reconcile the mixed findings in the literature is discrepancies in labor market institutions between the real world and the theoretical model. Although the canonical payroll tax incidence model assumes a perfectly competitive labor market, competitiveness in an actual labor market varies across countries, demographic groups, and time periods.

Conclusion

     Despite clear predictions under the canonical model of a competitive labor market, existing empirical studies have provided mixed evidence on how much a payroll tax change is shifted to workers’ wages. One possible explanation for this discrepancy is the heterogeneity in labor market flexibility across countries and time. To better understand the role of market flexibility or competitiveness, we investigate the labor market impact of payroll taxation in Singapore, where the labor market is considered more competitive and flexible than in any other country investigated in the literature. We document that saved labor costs via the reduced payroll tax rate largely shifts to workers’ wages. The estimated pass-through rate in Singapore is 77.5%, which is similar to the weighted average pass-through rate of the studied countries where wage bargaining predominantly takes place at the local or firm level, as in Singapore. These findings indicate that labor market competitiveness can play a significant role in determining the labor market impact of payroll taxation. A policy implication is that labor market distortion due to payroll taxation would be small when the labor market is competitive.