The empirical evidence on the incidence of payroll taxation is primarily based on the wage bill of firms. This paper applies matched employer-employee register data on individual wages for all private sector workers in Norway. Exploiting a payroll tax reform and using the difference-in-difference approach, I find that 1% reduction in labor costs generates 0.5% wage increase. Among low educated workers the degree of tax shifting equals 50%, while the wage response for highly educated is insignificant. Lower payroll taxes have limited effects on employment. The findings imply that the absolute value of the labor demand elasticity decreases with the level of education.