Business Networks and Performance: A Spatial Approach

firm that accesses a horizontal network for inputs has 50.69% more use of locally
produced inputs from a firm that accesses either vertical networks or exercises spot
trade for inputs. Between two firms with all their characteristics equal at sample’s
means, the firm that accesses vertical networks for customers has 44.44% more
exported output from a firm that accesses either horizontal networks or exercises spot
trade for customers. Table 4 shows the estimated coefficients and asymptotic t-values
for the four logit models of business performance. The percentage of material input
from local firms exerts a statistically significant and positive effect on employment
change and a statistically significant but negative effect on investment change. The
percentage of exported product affects significantly and negatively the performance in
terms of profit margins. The location of the firm affects all dimensions of performance
with firms located in Evrytania having higher probabilities of performing better than
firms located in Kalavryta. Firms in the manufacturing and tourism sector are less
probable to have increased total sales while firms in the trade sector are less probable to
have increased investments. Table 5 shows the marginal effects of independent
variables on the probability of having a positive change in each one of the four
dimensions of performance.

Table 2. Coefficient estimates of tobit models for PMIR and PSE.

Note: Two asterisks indicate significance at the 5%.

18

<<   <   15   16   17   18   19   20   21   >   >>