of either the tax rate or the tax base, i.e. the data points are located on the axes.
Among the tax reforms which changed the tax rate and the tax base at the same
time, only the Canadian tax reform of 1991 follows the way predicted by theory;
however, it just reversed the reform of 1990 to the same extent and may therefore
be interpreted as a mere correction. The only country5 to implement a revenue
decreasing tax reform of both the tax rate and base is Portugal in 1988, whereas the
United States (1992), Finland (1995), France (1996) and Ireland (2002) implement
revenue increasing tax reforms (quadrant IV).
Most tax reforms which changed tax rate and base simultaneously were of the
tax rate cut cum base broadening kind, which consists of lowering the statutory tax
rates on business profits and reducing the present value of depreciation allowances
at the same time. Among those are tax reforms in Great Britain, Germany and
Japan, and - probably known best - the US tax reform of 1986. It is striking
that even the large countries which could be expected to be relatively autonomous
in their tax policy pursued this kind of strategy. The question arises how this
development can be explained.
Compared to the literature on the efficiency and welfare enhancing effects of
consumption tax systems, the literature explaining the obvious deviation from this
ideal is relatively scarce. There are basically two approaches to explain this trend.
A first approach is based on the idea of ’policy learning’, which is extensively
discussed in the political science literature (see e.g. Steinmo (2003) and Swank
& Steinmo (2002)): Inspired by the fundamental reforms in Great Britain and
the US, policymakers around the world followed their example and adjusted their
tax system to the new model (e.g. see Whalley (1990) and Gordon (1992)). The
underlying assumption is that policymakers do not have an explicit model of the
economy in mind and no clear efficiency goals, but they do observe other policy-
makers and try to copy their strategies when they observe successful ones.6 The
5The recent tax policy reforms in the U.S. are not included in the diagram. As Gordon et al.
(2004) show, these reforms narrowed the tax base by improving depreciation rules for investment
goods, and should therefore be depicted in quadrant II. Combined with various opportunities
of tax exempt savings and the continued deductability of interest payments, this policy ends up
subsidizing the marginal investment.
6Another aspect here is that the US was an important supplier of foreign direct investment
at the time. The foreign tax credit system enables the host country to increase tax rates on US
multinationals up to the US statutory rate without increasing the effective tax rate for these