Desktop version

Home arrow Management arrow Improving Access and Quality of Public Services in Latin America: To Govern and To Serve

Determinants of PB

PB intensity varies significantly across municipalities in Peru. Since PB has not been randomly assigned, socio-demographic and political factors at the local level may be driving demand for PB use. In this section we present an analysis of potential determinants of PB intensity, running linear regression of PB intensity measures against those variables.

Our dependent variable includes two measures of PB results: percentage of municipality’s investment budget prioritized through PB, and percentage of investment projects prioritized through PB. As explained above, we use both a strict and a lax definition of matching projects in the PB and SIAF databases. In addition, we also present regression results for two measures of participation in PB: number of participating organizations, and number of individual participants as a percentage of the district’s population. We consider four socio-demographic variables: percentage of urban population, poverty rate, Gini coefficient, and human development index.

Our electoral results database allows us to capture several features of the political environment. These include measures of political participation (number of political groups participating in the last election—2006—and percentage of votes cast over voting population), and authorities’ political backing (percentage of votes obtained by the mayor and whether she was immediately re-elected). We also include the percentage of invalid votes, which may be a measure of political dissent (if votes are consciously invalidated) or political culture (i.e. people do not know how to cast a valid vote). Another measure of political culture that we also include is the percentage of undocumented adults. Finally, we have a measure of women’s participation (percentage of women among voters). Results are presented in Table 4.5.

There are three main findings from this analysis. One, we do not find any strong systematic correlation between socio-demographic or political variables and PB results measures. Particularly, our preferred measure of PB intensity (percentage of investment budget prioritized by PB) is not significantly correlated with any of the independent variables included. Two, among political variables, we have some evidence of a relationship between political participation and PB use (not PB intensity). In effect, the number of political groups correlates positively with the number of organizations participating in PB. However, it correlates negatively with the number of individuals participating. This is consistent with a view of political groups as entities

PB intensity

PB intensity

Projects

Projects

Plumber of

People participating

lux definition

strict

prioritized by

prioritized by PB

participating

on the PB (% of total

definition

PB (%), lax

(%), strict

organizations

district population)

definition

definition

Urban population

-0.0514

0.116 (0.112)

-0.00398

0.00678

31.25 (51.14)

-0.00407 (0.00301)

(%)

(0.194)

(0.0155)

(0.0127)

Poverty incidence

-0.162

0.0262

0.00663

0.0151 (0.0187)

198.7***

-0.00131 (0.00441)

(0.286)

(0.165)

(0.0228)

(74.84)

Gini coefficient

-1.321

-0.0974

-0.0479

-0.00290

-151.4 (218.2)

-0.0117 (0.0128)

(0.832)

(0.480)

(0.0665)

(0.0544)

Human

1.490 (1.285)

0.536 (0.742)

0.243**

0.152* (0.0839)

540.5 (334.1)

-0.0222 (0.0196)

development

index

(0.103)

Number of

-0.00344

0.00636

0.000247

0.000660

8.023***

-0.000432***

political groups

(0.00930)

(0.00537)

(0.000743)

(0.000607)

(2.515)

(0.000148)

Votes cast over

-1.471

-0.230

0.0604

0.0820 (0.0651)

435.5 (264.4)

0.0272* (0.0155)

voting population

(0.996)

(0.575)

(0.0796)

% of the votes of

-0.631

0.0371

0.00190

0.0144 (0.0258)

67.89 (109.1)

-0.00224 (0.00643)

winner over the total number of voters in the last

(0.395)

(0.228)

(0.0315)

election process Invalid votes over

0.188 (0.618)

0.167 (0.357)

-0.0101

-0.0221

354.1**

0.00590 (0.00936)

votes cast

(0.0494)

(0.0404)

(158.9)

Mayor

0.0564

-0.00444

0.00694

0.00617

13.96 (17.50)

-0.000509 (0.00102)

immediately re-elected (yes =

i)

(0.0646)

(0.0373)

(0.00516)

(0.00422)

(continued)

Table 4.5 (continued)

PB intensity, lax definition

PB intensity,

strict

definition

Projects prioritized by PB (%), lax definition

Projects

prioritized by PB (%>), strict definition

Plumber of

participating

organizations

People participating on the PB (% of total district population)

% of voting women over the total voting population

0.675 (1.229 )

  • -1.095
  • (0.709)
  • 0.0227
  • (0.0982)
  • -0.0230
  • (0.0802)

-426.0 (328.5)

0.00373 (0.0192)

Population over

0.00142

0.00154

0.00380**

0.00370***

-1.652 (5.506)

-0.000301

18 with no ID

(0.0212)

(0.0122)

(0.00169)

(0.00139)

(0.000324)

Constant

0.949 (1.540)

0.357 (0.888)

-0.194 (0.123)

-0.172* (0.101)

-613.6 (403.7)

0.00490 (0.0238)

Observations

195

195

195

195

184

185

R-squared

0.091

0.045

0.120

0.107

0.143

0.198

Standard errors in parentheses.

* ^<0.1; ** ^<0.05; *** p<0.01

that represent different sectors of the population. In this context, a smaller number of political groups will leave more room for greater participation of groups representing narrower interests (neighborhood associations, parents groups, among others). Percentage of votes cast also correlates positively with one PB participation measure, though only weakly, but not with PB results (PB intensity). The percentage of invalid votes also correlates positively with one measure of PB participation, suggesting that political dissatisfaction may be associated with PB participation. Women’s political participation does not seem to have a systematic correlation with our PB intensity measures. Three, a possible measure of social exclusion, the percentage of undocumented adults, is positively correlated only with percentage of projects prioritized by PB, both under the strict and lax match.

In sum, we do not find any strong correlates of our PB results measures among the socio-demographic and political variables analyzed. This is particularly so for the case of our preferred measure of PB intensity. However, we have some evidence of a positive correlation between PB participation and broader political participation.

 
Source
< Prev   CONTENTS   Source   Next >

Related topics