Geoscience Reference
In-Depth Information
TABLE 8.5.
Does Inequality Reproduce Itself? Determinants
of Overall Economic Inequality in OECD Countries (1980-2000)
FEVD
Decentralization of Interpersonal
.08***
-
Redistribution
(.000016)
Geographical Disparities in the
-
.04***
Incidence of Inequality
(.00006)
Representation
-
.10***
(.00001)
Geographical Disparities in the Incidence
-
−
.09***
of Inequality*Representation
(.0001)
Market Income
.193***
.232***
(.0007)
(.0001)
Economic Coordination
−
.028***
−
.068***
(.0003)
(.0005)
Partisanship
−
.003***
−
.003***
(.00006)
(.0005)
N
53
53
Adj. R-sq
.924
.920
Key: * p
<
.10,**p
<
.05, *** p
<
.01
these variables to be major determinants of redistribution (Beramendi and
Cusack
2009
, Huber and Stephens
2001
, Kenworthy and Pontusson
2005
).
14
In turn, controlling for market income inequality provides a straightforward
solution for (b) and (c). Inequalities of disposable income are a function of mar-
ket income inequalities and the direct effect of redistributive policies. What is
to be assessed in the different specifications is the role of one or more variables
that determine the level of redistribution and, through that channel, the distri-
bution of disposable income. This goal makes an indicator of market income
inequality a very useful tool. It aggregates the impact of economic, demo-
graphic, and labor market related variables. More importantly, it also includes
the feedback effects of previous redistributive policies on people's labor mar-
ket behavior and, subsequently, on the distribution of wages and additional
sources other than earnings and welfare state transfers. Finally, the indicators
for the decentralization of interpersonal redistribution, the geographic spread
in the incidence of income inequality, and political representation are the same
as the ones used in the first section.
Because the indicator of economic coordination is time invariant, I report
a panel regression with vector decomposition following the procedure devel-
oped by Pl umper and Troeger (
2007
).
15
Table 8.5
reports the estimations.
14
For a critical review of this literature, see Stasavage and Scheve (
2009
).
15
The estimation also introduces a Prais-Winsten correction for serial correlation of order one
and cluster the standard errors by cross-sectional unit.The results reported are robust to other
estimation approaches such as panel corrected standard errors with serial correlation of order
