Pierre F. Landry: Decentralization and Regime Transformation
28
model in which one endogenous variable is dichotomous
14
(REGIME) while the other is
continuous (ECONOMIC WELL BEING).
Impact of Decentralization on Economic Well-Being
Economic well-being is measured as the gross domestic product per capita in 1995
constant dollars, using the PPP method. I hypothesize that current levels of wealth are the
product of past economic performance. The specific causes of growth are captured only
indirectly by the coefficients that account for recent growth performance and a time
trend.
The equation also incorporates two institutional measures of fiscal
decentralization that are central to analysis: the (lagged) percentage of state and local
expenditures among all government expenditures, and the corresponding variable on the
revenue side. Finally, we account for the impact of the political regime (lagged) on
economic performance, including the impact of constitutional federalism.
GDP
t
= +
β
11
*
∆GDP
(t/t-1)
+
β
12
*
∆GDP
(t-1/t-2)
+β
13
*
FEDERALISM
(t)
+β
14
*
LOCAL EXPENDITURES
(t-1)
+β
15
*
LOCAL REVENUE
(t-1)
+β
16
*
REGIME
(t-1)
+β
17
*
YEAR +
α
1
+ε
1
=
X
β
I
+ε
1
(eq1)
14
Here, the existence (REGIME) is observed as a dichotomous variable, but underlying
hidden process is assumed to be continuous, as in standard presentations of the probit
model.
Convention