On the background of 2007-08 financial crisis, the American Recovery and
Reinvestment Act (ARRA) of 2009 proposed fiscal stimulus packages in
response to a large economic downturn. Supply-side measures included in ARRA
involve distortionary taxation reform such as labour income, or capital
income tax. Some of them were implemented retroactively, while others will
be engineered beyond 2009. However, the existing public finance literature
evaluates and ranks various distortionary tax reforms according to their
welfare consequences under rational expectation.
Building on the contribution of Evans et al. (2009) and Mitra et al. (2013),
we aim to generalize their analysis of anticipated fiscal policy under
learning by studying an economy featuring distortionary taxes. When agents
use adaptive learning rules to forecast factor prices, our model predicts
oscillatory dynamic responses to pre-announced permanent tax changes.
Moreover, the case of distortionary taxation is particularly different with
regard to the effects on impact and the volatility throughout the transition
period. Confronted with this result, we then investigate the welfare
implications of those striking differences in the dynamics of a
pre-announced tax reform at the presence of several tax instruments. We find
that tax reforms designed to improve welfare, do so to a much lower extent
under learning compared to perfect foresight. Thus, the learning perspective
on tax reforms provides fundamental different insights for benevolent policy
makers.
Researchers: Shoujian Zhang, George Evans and Kaushik Mitra
Related CDMA Working Papers
1301 Emanuel
Gasteiger and Shoujian Zhang "Anticipation, Learning and Welfare: the Case
of Distortionary Taxation"
