Hansen (1983) studies the effect of a number of variables on tax innovation, using
a cross-sectional framework on the US states. Three types of taxation are regarded:
general sales tax, personal income tax and corporate income tax. Her analysis focuses
on bivariate relationships between the variables examined. Three important conclusions
can be drawn from her study. Firstly, tax innovations are different from other policy
innovations. Using a state’s dinnovation scoreT, it is shown that the bgeneral
propensity to innovate is not much help in explaining broad-based tax adoptionsQ
(Hansen, 1983, p.147). The innovation scores are based on the analysis of 88 different
policy programs adopted by at least 20 states prior to 1965 (Walker, 1969). The
earlier a state adopts new policies on average, the higher is its innovation score.
However, contrary to Hansen’s expectation, the correlation between this innovation
score and the introduction of taxation is consistently negative for all three types of
taxation.3 Secondly, economic considerations affect tax innovation. More precisely,
economic crises reduce the political risks of introducing new taxes. Hansen (1983) shows
that the introduction of (3 types of) taxation by US states cluster in the Depression years
(early 1930s) and occur only infrequently in the prosperous 1920s and 1950s. Thirdly, the
dpolitical opportunityT to innovate is a very important intermediary factor. Hansen (1983,
p.153) argues that the existence of unified versus divided party control of the legislative
and executive body in a state diminishes at least part of the dinstitutional roadblocksT
against policy implementation. As such, unified governments should be more likely to
introduce taxes than divided governments. The data show some support for this
hypothesis.