27.12.2012 Views

industrial policy, its spatial aspects and cluster development in ...

industrial policy, its spatial aspects and cluster development in ...

industrial policy, its spatial aspects and cluster development in ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Prov<strong>in</strong>ce× year fixed effects <strong>in</strong>cluded Yes Yes<br />

Pseudo R 2 0.130 0.131<br />

Number of observations 349295 349295<br />

Notes: All regressions are estimated by probit maximum likelihood. Numbers <strong>in</strong> parenthesis are<br />

asymptotic t-values obta<strong>in</strong>ed from robust st<strong>and</strong>ard errors adjusted for <strong>cluster</strong><strong>in</strong>g at the district level.<br />

The regressions <strong>in</strong>clude <strong>in</strong>tercept terms, but they are not reported. **, <strong>and</strong> *** denote statistical<br />

significance at the 5 <strong>and</strong> 1% levels, respectively. For def<strong>in</strong>ition of control variables, see Table 5.5.<br />

stronger pro poor effects <strong>in</strong> low <strong>and</strong> medium <strong>in</strong>equality regions. The magnitude of the<br />

effect suggests that a one SD <strong>in</strong>crease <strong>in</strong> post-primary school system <strong>in</strong>dex leads to a<br />

0.26% reduction <strong>in</strong> rural poverty <strong>in</strong> low-<strong>in</strong>equality districts as compared with 0.08%<br />

poverty reduction <strong>in</strong> medium-<strong>in</strong>equality districts. The impact of present levels of G<strong>in</strong>i<br />

<strong>in</strong>equality on poverty is relatively weak <strong>in</strong> our sample where low <strong>in</strong>equality districts do<br />

not immediately benefit from these <strong>in</strong>vestments while the impact on medium-<strong>in</strong>equality<br />

districts roughly rema<strong>in</strong>s unchanged.<br />

5.3.4 Road <strong>in</strong>frastructure <strong>and</strong> poverty: Does <strong>in</strong>equality matter?<br />

How <strong>in</strong>vestment <strong>in</strong> regional road <strong>in</strong>frastructure affects rural poverty under different<br />

<strong>in</strong>equality regimes? We <strong>in</strong>vestigate the effects of changes <strong>in</strong> regional road <strong>in</strong>frastructure<br />

on poverty by classify<strong>in</strong>g districts <strong>in</strong>to three categories of <strong>in</strong>equality. Low <strong>in</strong>equality<br />

districts are those where G<strong>in</strong>i <strong>in</strong>dex is 0.5 st<strong>and</strong>ard deviations below the mean; high<br />

<strong>in</strong>equality refer to districts where G<strong>in</strong>i <strong>in</strong>dex is 0.5 st<strong>and</strong>ard deviations above the mean<br />

<strong>and</strong> medium <strong>in</strong>equality districts refer to all other districts. We <strong>in</strong>corporate road density<br />

<strong>in</strong> respective districts <strong>in</strong>to the model by <strong>in</strong>teraction of these three dummy variables with<br />

the road density to capture the slope differentials of road density. Thus the effect of a<br />

change <strong>in</strong> the district level <strong>in</strong>dex of road density on the probability of poverty under<br />

alternative <strong>in</strong>equality regimes is captured by these <strong>in</strong>teraction terms.<br />

Regression results shown <strong>in</strong> column (1) of Table 5.10 reveal that hold<strong>in</strong>g all else as<br />

constant <strong>in</strong>crease <strong>in</strong> road density significantly decreases the probability of poverty <strong>in</strong> all<br />

the low, medium <strong>and</strong> high <strong>in</strong>equality districts. For example, a one st<strong>and</strong>ard deviation<br />

<strong>in</strong>crease <strong>in</strong> road density (i.e., 0.125) demonstrates highest levels of poverty reduction <strong>in</strong><br />

the low-<strong>in</strong>equality districts (5.8%), followed by the medium-<strong>in</strong>equality districts (3.9%)<br />

<strong>and</strong> then the high-<strong>in</strong>equality districts (3.5%). Column (2) presents the results of <strong>in</strong>crease<br />

<strong>in</strong> road density with start<strong>in</strong>g levels of <strong>in</strong>equality <strong>and</strong> shows that a one st<strong>and</strong>ard deviation<br />

370

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!