We study the employment and wage effects of age-dependent payroll tax subsidies for both younger (under 25) and older (over 55) workers using a policy change in Hungary. Using rich administrative data and a difference-in-differences strategy, we find that employment increases in both age groups as a result of the payroll tax cut, but the effects differ by job and firm characteristics. Employment of older workers mostly increases in worse-paid, blue collar jobs and at less productive and lower-quality firms. On the other hand, employment of older workers at more productive and higher-quality firms is unaffected, but their wages increase significantly in response to the tax cut. We explain these findings using an equilibrium job search model with heterogeneous firms and workers. In the model, more productive firms hire skilled workers by poaching them from other firms. Tax reduction for these firms mainly increase worker's wages, while employment is not affected. On the other hand, low-productivity firms, which tend to hire from unemployment, mainly react to the tax subsidy by increasing employment and the effect on wages is limited. These results point to important heterogeneity in the incidence of payroll tax subsides by firm and worker types: the incidence mainly falls on low-productivity firms, and higher-skilled workers.