The African Business Review May-Jun 2014 | Page 32

Table 1. Panel least squares dummy variables
estimates for heterogeneous sample of
countries. Dependent variable: INF
Variable

C
DEM
POLSTAB
MONgrowth
LnRES
Growth
Trade

Coefficient

t-stat

7.8
-3.7
-2.32
0.35
-0.36
-0.35
0.03

0.41
-1.89**
-2.3**
3.57***
-0.45
-3.3***
0.84

*,**,*** indicate statistical significance at 10%, 5% and 1% level, respectively

Table 2. Panel least squares dummy variables
estimates for developing countries. Dependent
variable: INF
Variable

C
DEM
POLSTAB
MONgrowth
LnRES
Growth
Trade

Coefficient

t-stat

3.04
-3.63
-2.25
0.37
-0.18
-0.38
0.03

0.13
-2.19**
-1.77**
2.64***
-0.19
-2.81***
0.38

R-squared = 0.27
F(6,1744) = 6.84
Prob > F = 0.0000

and transparency’s criteria’s, constitutes a significant factor in
reducing inflation.

Moreover, the results suggest that in terms of impact,
improving democratic standards has the highest marginal effect
among all other economic and institutional variables. Indeed, a
1% improvement in democracy index, as established by Kaufmann
et al. (2012) reduces inflation by 3.7%. In other words, a country
that in the specified period have improved its democracy index
by one percentage point have experienced, on average, a level of
inflation that is 3.7 percentage point less than a country that
haven’t insured the same institutional effort.

The estimated coefficient of the variable democracy is, for
example, higher than the estimated coefficient of the variable M2
growth, which is a very representative proxy of monetary policy
and represents an important tool in handling inflation strategies.

13

The selection of countries as developing and emerging ones has been made according to
the classification of the International Monetary Fund, World Economic Outlook,
April 2011.

32 | The African Business Review

That being, we have restricted our sample of countries to only
developing and emergent countries[13] in order to demonstrate
our theoretical position with respect to the thesis of the median
voter. The estimation results for 82 developing countries and for
the period 1996-2012 are found overleaf (Table 2)

Estimation results confirm our critiques of the median voter
theory, which finds its purpose especially in developing countries
and who implies the existence of a positive correlation between
political participation and inflation. The size of the coefficients
is indeed very similar to those of the previous model which is
constituted by a heterogeneous sample of countries and the overall
model is significant according to the overall significance test of
Fisher.

In addition, the variable (DEM) continues to have a negative
impact on inflation and the highest marginal impact compared
to all other variables retained in the model. The negative impact
of democracy on inflation is, therefore, independent of the level
of development of countries.

Conclusion

In this paper, we have demonstrated that democracy; when
apprehended by including concepts of participation, contestation
and transparency, constitutes an important instrument in

Fiscal discipline has been empirically linked
to a number of institutional factors, including
budgetary centralization, transparency, fiscal
rules, and parliamentary procedures.
reducing inflation. The economic literature has often explained
the evolution of inflation by referring to economic policies. Our
study differs from this issue and attributes to socio-political factors
a fundamental role in explaining differences in macroeconomic
performances, showing that by providing a theoretical and an
empirical framework.

In the theoretical framework, we have argued that monetary
and fiscal policies are redistributive tools under the control
of policymakers and that democracy provides institutional
arrangements that constrain these policymakers to use these
policies in favour of general interest and avoiding inefficient
macroeconomic policies. In our empirical framework, we have
initially made estimations based on a heterogeneous sample,
w X