DO HOME TOWNS OF MINISTERS RECEIVE MORE CONDITIONAL GRANTS PER CAPITA Sample Clauses

DO HOME TOWNS OF MINISTERS RECEIVE MORE CONDITIONAL GRANTS PER CAPITA. A linear regression was calculated to assess the ability of the main independent variable, the home towns of Flemish ministers, to predict the dependent variable, the conditional grants per capita. Each linear regression will be executed twice, one for 2012 and another for 2013. First, in step 1, Model 1 was created in which only the variable ‘home town minister’ and the dummy variables that represent the political parties are represented. Step 1 was executed to assess the impact of the independent variable on the dependent variable without the impact of the control variables. The dummy variable for the home town of a N-VA minister (N=2) was not included and will count as the reference category. In step 2, Model 2 was created in which the control variables were added. This includes the socio-demographic variables and the political variables. The actual SPSS output tables for each segment of the statistical results, can be found in the Attachments. Results 2012 In 2012, Model 1 explained 1,6% of the variance in conditional grants, as can be seen in Table

DO HOME TOWNS OF MINISTERS RECEIVE MORE CONDITIONAL GRANTS PER CAPITA. The general research question is how the presence of a Flemish minister in a municipality impact the conditional grants per capita that the municipality will receive since the scientific literature has shown that politicians tend to favour their home towns. Based on the descriptive results, the assumption could be made that the home towns of Flemish ministers overall receive more conditional grants per capita. However descriptive results do not tell what the predictors are so in order to assess the impact of the variables, a linear regression was calculated for both 2012 as 2013. In 2012, Model 1, with only the home town variable and dummy variables for the political parties, explained a total 1,6% of the variance. Model 2, after adding the control variables explained 11,9% of the total variance in the dependent variable. Based on the R² Change value, the control variables thus added another 12,8% to the model. In the estimated model coefficients model the variable home town sp.a. appeared to be significant without the control variables. However, after adding the control variables, the dummy variable does not significantly predict the dependent variable anymore. Similar results were found in 2013 since the first model explained 1,6% of the variance and Model 2 8,6%. However, unlike the previous linear regression, the home town variables did not appear to be significantly associated with the dependent variable, even without the addition of the control variables in Model 2. The conclusion can thus be made that in 2012 and 2013, the home bias was not very present in Flanders since no significant association was found between the presence of a Flemish minister and the conditional grants per capita received by a home town of a minister. Nonetheless, differences could be noticed between the political parties in the Flemish government. For instance, N-VA did not show any effect while sp.a. appeared to show the strongest association, remarked by a significant association in Model 1 of the linear regression of 2012. Furthermore, the home towns of CD&V ministers do not seem to be favoured in the allocation of conditional grants. Hence, this master’s thesis is not able to confirm the results from existing literature and cannot confirm hypothesis 4 (cf. Xxxxxxx & Repetto, 2016; Fiva & Xxxxx, 2016). Concerning the control variables, the unemployment rate appeared to be the most significant predictor throughout 2012 and 2013. The positive association between the unemplo...