4.1. Results
The models in Table 1 test the magnitude and signifi-
cance of different interactions between economic development,
change in the quality of governance, and
economic growth that are part of the three-way interaction
in model 5. The results in model 1 e with no interaction
terms e indicate that incumbents benefit out of
improvements in good governance and the economy. Both
indicators are positively associated with vote share, and
their coefficients are statistically significant.9 Thus, the
prime minister's party receives on average 1.13% extra
votes for a 0.1 point increase in governance effectiveness
and 0.6% votes for 1% economic growth.10 The impact of
the quality of governance in elections is comparable to the
role of the economy (i.e. one standard deviation increase
in economic growth or the quality of governance is associated
with 1.5% or 1.8% increase in the vote share, with
their 95% confidence intervals overlapping). These results
reflect the averaged effects of these two independent
variables across the observed range of economic wealth
level, economic growth, or change in good governance.
Models 2-5 show that these effects are, in fact, moderated
by the other two predictors. Before discussing these
models, a brief presentation of the control variables is
necessary.
Incumbents with a large base - a high absolute vote in
the previous elections - lose electoral support easier in the
next elections. The positive statistically significant coeffi-
cient of economic wealth shows that, all else being equal,
incumbents have an advantage in prosperous democracies.
The other control variables do not have statistically
significant coefficients, but if they had, their
effect would have been as expected. In a coalition, the
prime minister's party is more vulnerable and loses more
of its previous support than if it was a single-party cabinet.
As countries get more democratic, the electorate is