Geoscience Reference
In-Depth Information
TABLE 8.6.
Interregional Redistribution in Political Unions
PCSE (AR1) TSIV(2)
FEVD
Mobility
1.40***
.89**
1.6***
(predicted)
(.49)
(.29)
(.30)
Representation
−
.62**
−
.33
−
1.16***
(.29)
(.48)
(.07)
Interregional
.09
−
.39
.18**
Inequality
(.11)
(.26)
(.08)
Country Size
.083
.28*
−
.16***
(.11)
(.16)
(.07)
GDP per
−
.011
−
.009
−
.001
capita (lag)
(.021)
(.039)
(.002)
Ethno-linguisticn
1.86***
2.16***
1.75***
Fractionalization
(.57)
(.82)
(.06)
Total
−
.84
−
1.16***
−
.44***
Taxation
(.14)
(.20)
(.07)
Democracy(Polity)
.006
.021
0.036***
(.025)
(.039)
(.003)
Intercept
4.68****
3.64*
6.02***
(1.3)
(1.99)
(.12)
R-squared
.86
.60
.75
N
197
197
197
taxation on interpersonal and interregional redistribution reveals an interesting
connection between the two dimensions of redistribution: political unions that
collect larger revenues as a share of their economy tend to both decentralize
the administration of taxes and transfers and limit the flow of discretionary
resources between regions. To the extent that the size of tax revenues is asso-
ciated with the amount of redistribution in society, this points to an interest-
ing trade-off between interpersonal and interregional redistribution. Using a
sample with both developing and developed unions, the negative association
between total taxation and interregional transfers suggests that those countries
that redistribute more through interpersonal taxes and transfers manage to
limit the ability of local lords to extract resources via the system of interregional
transfers.
The estimated impact of representation on interregional transfers also gener-
ates implications consistent with this view. The left panel in
Figure 8.8
displays
the predicted value of interregional transfers at different points of the repre-
sentation scale.
Ceteris paribus
, as local elites lose power relative to national
elites (as the system of representation moves from centrifugal to centripetal),
the scope of interregional transfers declines significantly. The effect is pretty
noticeable, albeit not overwhelming: covering the entire range of the repre-
sentation scale reduces the predicted level of interregional transfers by about
2% of GDP. Putting together this finding with the ones in the previous section
