The African Financial Review July-August 2014 | Page 64

is not only possible for the present time period but can also be
generalized for other time periods. In addition, this means that
the results obtained from this study are not spurious.

Equation (3.7) was estimated to obtain the results presented
in Table 2 (see overleaf ). The results show that the adjusted R2
is 0.244. This implies that all the independent variables explain
about 24.4 percent variation in the dependent variable. The
R2 for panel data is usually low; this explains why the R2 is
0.280. The F-stat probability is 0.0000, that is, significant at 1
percent. In conclusion, based on these results, political rights
(proxy for political institutions) have a higher significant impact
on economic growth than trade liberalization, economic and
cultural institutions examined in this study. The LSDV results
in Table 2 reveal that Gkap (gross fixed capital formation) and
political rights (proxy for political institutions) are statistically
significant at 1 percent, while Ssenr (secondary school enrolment
– proxy for human capital), employment to population ratio (Lab),
degree of openness (Open), repudiation risk (proxy for economic
institutions), taxes and natural resource endowment (Nare) are
statistically significant at 10 percent, on the other hand, Psenr
(primary school enrolment – proxy for human capital) and ethnic
tensions (proxy for cultural institutions) are statistically significant
at 5 percent. In addition, the coefficients of elasticity are less than
one for the entire variables.

As regards the coefficient estimates, the coefficients of Open,
Reprisk and natural resource endowment are small (13.3 percent,
29.2 percent and 14.3 percent respectively). This implies that trade
openness and economic institutions do not have a noticeable
impact on economic growth. But labour, ethnic tensions, gross
fixed capital formation and political rights have a high impact
on economic growth; this is evident from the high coefficient
estimates of 63.1 percent, |75.4| percent 48.6 percent and 36.2
percent respectively. This implies that political and cultural
institutions have statistically significant influence on economic
growth in the selected SSA countries. This supports the empirical
findings of Alonso and Garcimartin (2009) who opined that taxes
and strong economic and political institutions exert a positive
impact on economic growth. But this may not be totally true for
the sampled SSA countries as some of them are not experiencing
the growth they are supposed to, due to many militating factors
such as economic and political insecurity, high inflation rate and
so on.

The second aspect of the estimation process involved the
Generalized Method of Moments (GMM) regression analysis.
Equation (3.9) was estimated to obtain the results presented in
Table 3. The system GMM estimator is categorized into the
one-step and two-step options, these are reported in columns 2
and 3 respectively. The pooled Ordinary Least Square (OLS) and
the Least Square Dummy Variable (LSDV) results are reported
in columns 1 and 4 respectively. The results in Table 2 begin
with some diagnostic tests. The starting point is based on the
assumption that, the individual errors are serially uncorrelated
for the system GMM estimators for consistent estimations. The
presence of autocorrelation will indicate that lags of the dependent
variable (and any other variables used as instruments that are not
strictly exogenous), are in fact endogenous, hence bad instruments.
Arellano and Bond (2001) develop a test for this phenomenon
that would potentially render some lags invalid as instruments. Of
course, the full disturbance is presumed autocorrelated because it
contains fixed effects, and the estimators are designed to eliminate
this source of trouble. The next diagnostic test is a test of over-

64 | The African Financial Review

identifying restrictions of whether the instruments, as a group,
appear exogenous. This test of instrument validity has to do with
a comparison of the number of instruments used in each case
and the related number of parameters. It is implemented by the
Sargan and Hansen J tests. The Sargan and Hansen J tests are
used to test if the instruments as a group are exogenous. The test
is out to accept or reject the null hypothesis that states that the
instruments as a group are exogenous. The higher the p-value
of the Sargan statistic, the better.

The results in Table 3 (see overleaf ) reveal that for onestep, non-robust estimation, the Sargan statistic which is the
minimized value of the one-step GMM criterion function, is
applicable. The Sargan statistic in this case is, however, not robust
to autocorrelation. So for one-step, robust estimation (and for
all two-step estimation), the xtabond2 (STATA command) also
reports the Hansen J statistic, which is the minimized value
of the two-step GMM criterion function, and is robust to
autocorrelation. In addition, xtabond2 still reports the Sargan
statistic in these cases because the Hansen J test has its own
problem: it can be greatly weakened by instrument proliferation.
Only the respective p-values are reported for this test results in
the lower part of Table 2. Here, the null hypothesis that the
population moment condition is valid is not rejected if p > 0.05.
The summary statistics indicate that the one-step and two-step
system GMM dynamic panel models of the selected 30 SSA
countries have 30 instruments and 11 parameters each. This
represents a total of 19 over-identifying restrictions in each case.
The number of instruments satisfies the rule that says that the
number of instruments should be less or equal to the number
of groups. In this study, we have thirty sampled countries.
In both specifications, the Hansen–J statistic does not reject
the over-identifying restrictions (OIR), thus confirming that
the instrument set can be considered valid. The Sargan test is
significant at 5 percent.

With respect to the results of the proxies for capital
and labour (gross fixed capital formation and employment to
population ratio); they are satisfactory, the coefficient estimates
are consistent with theoretical expectations. The Blundell–Bond
(system-GMM) robust estimates indicate that the lagged growth
value (first lag) is statistically significant across the sampled
SSA countries. In other words, past realizations of economic
growth do produce some significant impact on the current level
of economic growth. Secondary and primary school enrolments
– proxies for education produced some very interesting results in
the Blundell–Bond robust estimates. One striking observation
here is that education produced a positive impact on economic
growth across the sampled countries over the study period. This
variable is also statistically significant at the 5 percent level in the
one-step and two– step system GMM options. In more definitive
terms, a one percent change in secondary and primary school
enrolment under the two–step system GMM estimates, brings
about a greater proportionate change in economic growth across
the study group respectively; and under o ne-step system GMM
estimates also brings about a greater proportionate change in
economic growth respectively. Theoretically, the implication of
this result is education has a great impact on economic growth
in the selected SSA economies. The more educated the citizens
of the countries are, the better growth these countries experience,
ceteris paribus. A one percent change in the employment to
population ratio (Lab) brings about a less than one percent
change in economic growth. The implication of this result is