The Timing of Redistribution

We investigate whether late redistribution programs that can be targeted towards low income families can “dominate” early redistribution programs that cannot be targeted due to information constraints. We use simple two- period OLG models with heterogenous agents under six policy regimes: A model calibrated to the U.S. economy (benchmark), two early redistribution (lump sum) regimes, two (targeted) late redistribution regimes, and finally a model without taxes and redistribution. Redistribution programs are financed by a labor tax on the young and a capital tax on the old generation. We argue that late redistribution, if the programs are small in size, can dominate early redistribution in terms of welfare but not in terms of real output. Better targeting of low income households cannot offset savings distortions. In addition we find that optimal tax policy includes a positive capital tax rate.