12
Table 1.1 presents stylised information on the quarterly export and GDP figures. It can be
seen that export shares per unit of GDP are practically the same for all three countries. While
Germany’s exports of goods and services make up more than 30% of GDP, with 28% the
export share of France is not significantly lower, with Italy taking an intermediate position.
The relative importance of exports alone therefore does not explain any possible difference in
the sensitivity of growth to export performance.
Table 1.1: Exports and GDP growth
Exports as % of
GDP (average for
1999/2000)
Exports (standard
deviation
of quarterly
growth rate)
GDP (standard deviation
of quarterly growth rate)
Cumulative growth
response due to one-time
export growth increase
(by 1 standard
deviation)*
Germany
30.6%
1.57%
0.58%
0.41%
France
28.0%
1.28%
0.37%
0.42%
Italy
29.1%
2.02%
0.42%
0.25%
* Own Estimation
A more significant aspect is the degree to which exports fluctuate. Here it can be seen that
France apparently benefits from a relatively stable situation in its export markets. The
standard deviation of quarterly export growth lies at 1.3%, while that of Germany is
noticeably higher with some 1.6%, with Italy suffering from even higher export volatility
(2%). France also benefits from a very constant degree of GDP growth, with a standard
deviation of 0.37% in quarterly growth rates. With a value of 0.42%, growth volatility of Italy
is only slightly higher than that of France.
By contrast, the growth volatility in Germany is 1½
times higher than that of France with a value of 0.58%. Ceteris paribus, a high volatility of
GDP growth should lower the potential medium term growth path, because it increases the
risk of sub-optimal factor allocation in an economy. Noticeably, volatility renders investment
decisions more difficult.
An econometric analysis should pick up three potential links
between exports and GDP
growth. First, the two are directly linked, because exports by definition are a component of
GDP. Second, export growth can induce additional GDP growth via economic multipliers,
notably investment. Third, any correlation could simply reflect joint influences. In particular,
exports are a strong indicator of global economic growth, which
should also go in parallel
with rising stock markets. Insofar as stock markets are increasingly linked and, therefore,
economic sentiment moves increasingly parallel across the world, an econometric analysis
might therefore pick up a spurious correlation. Nevertheless, it indicates
the impact of the
global economy on the domestic market.
Clearly, therefore, exports should exert a positive influence on growth and investment. By
contrast, as the three countries are small compared to the rest of the world, they are largely
unable to pull the global growth rates. Export growth should therefore lead GPD growth, but
not the other way round. This is confirmed by a Granger causality test undertaken for the
three countries.
A dynamic means of testing the impact exports have on GDP consists in a vector
autoregression (VAR). This econometric technique implies regressing each variable of the
studied system on lagged values of itself and the other variables. Here we present only the
findings of a bivariate VAR that includes quarterly GDP and export growth. The
presented
results are very robust. This means neither the inclusion of other variables, variation of the
time lag or even the inclusion of a dummy variable to distinguish the pre- and post-unification
situation alter the result in a qualitative way. The estimated parameters of the VAR allow a
dynamic analysis of what happens to each variable if export growth is changed.
13
Figure 1.5 shows the impulse response of GDP over time to a one-time
increase in the export
growth rate by one standard deviation. The response differs clearly among the three countries.
Despite Italy’s higher export volatility, and hence relatively large initial export increase, its
GDP response is rather flat and peters out after 7 quarters. The cumulative impact on GDP is
0.25% (Table 1). By contrast, France’s response is bell shaped with additional growth rates of
over 0.10% in the second
to fourth quarter, after which it drops to zero in the sixth quarter.
The cumulative impact is accordingly higher than the Italian one with 0.42 %.
Do'stlaringiz bilan baham: 
