COMBATTING CORRUPTION AS A WAY TO INCREASE
ENROLLMENT AND STUDENT LEARNING: EVIDENCE FROM A
NEWSPAPER CAMPAIGN IN UGANDA
RITVA REINIKKA
THE WORLD BANK
JAKOB SVENSSON
IIES, STOCKHOLM UNIVERSITY,
DEV. RESEARCH GROUP, THE
WORLD BANK
what are the most effective ways to increase primary school
enrollment and student learning?
•
•
•
•
•
many would argue that the large literature on schooling in developing countries
provide few reliable insights about what governments in developing countries
should prioritize to raise educational attainment.
this lack of knowledge does not reflect a lack of interest in the economic
profession but that most of the contributions suffer from methodological
shortcomings (Glewwe 2002, Duflo, Glennerster, Kremer 2004)
recently, progress have been made in one important dimension: assessing the
effects and costs of specific school inputs and programs through the use of quasior randomized experiments.
similar to the conventional literature, the new literature studies schooling through a
somewhat narrow (and benevolent social planner’s) lens, i.e. they disregard
“political-economy” issues.
while it is important to know what specific school programs and inputs that work,
this information alone is not enough to guide government policy in developing
countries.
understanding the service delivery chain
•
when scaling-up a specific program that works in a controlled experiment run by a
specific organization, it is crucial also to have an understanding of the whole social
service delivery chain;
– institutional constrains that forms central government policy decisions
– incentive constraints that influences different layers of government agencies and
officials that implement a given policy
– actions and incentives of the end producers (schools) and beneficiaries (students
and parents).
•
not taking into account the workings of the service delivery system may yield results
that are very different from what a simple extrapolation of the estimates of the
controlled experiment would suggest.
governance of social services
•
as the total impact (when scaling up a program to the national level) depends on all the
elements in the service delivery chain, and as social service delivery in developing
countries is often plagued by inefficiencies and corruption, it follows that interventions
that focus on improving governance in general and governance of social services in
particular may be the most cost-effective way to increase school enrollment and student
learning.
•
as traditional approaches to improve governance has produced weak results in most
developing countries, experimentation and evaluation of new tools to enhance
accountability should be at the forefront of research into improving outcomes of social
services.
an unusual policy experiment
•
newspaper campaign in Uganda aimed at reducing capture of public funds by providing
schools (parents) with information to monitor local officials’ handling of a large
education grant program.
•
a capitation grant to cover primary schools’ nonwage expenditures.
•
program that appeared to work well, based on central government budget data. Funds
were disbursed by the Ministry in charge on regular basis.
•
but the situation on the ground was completely different from the official statistics. A
public expenditure tracking survey (PETS) to gauge the extent to which public resources
actually filtered down to the schools showed that in the mid-1990s, on average, schools
received only around 20 percent of central government spending on the program
(Reinikka and Svensson, 2004a). Most schools received nothing and schools in poor
areas suffered most.
•
the bulk of the grants was captured by local government officials (and politicians) in
charge of disbursing the grant to the schools.
newspaper campaign
•
government initiated a newspaper campaign in response to the problems identified in the
first PETS.
•
as an anticorruption program it differs from the standard package.
•
traditionally, anticorruption programs target corruption in service delivery primarily
through capacity building of the legal and financial institutions - judiciary, police, and
financial auditors - in charge of enforcing accountability in the public sector.
•
a top-down approach where some government agencies are assigned to monitor and
control others.
•
problem is that in many poor countries the legal and financial institutions are weak and
among the most corrupt. More resources to these agencies may therefore not be the best
solution.
•
for these reasons, the government decided to take the route of citizen engagement and
began to publish data on monthly transfers of capitation grants to districts in the national
newspapers.
why would the newspaper campaign matter?
•
Government:
– by using newspaper adds to inform the users of their entitlement, the government
signaled that it attached a high weight on primary education.
– it also signaled strengthened oversight.
•
Communities:
– by giving users access to information on the grant program, head teachers and
parents could themselves monitor the local administration and voice complaints if
funds did not reach the schools.
why would the newspaper campaign matter?
•
Hirschman (1970) identifies a set of conditions under which voice is the preferred
choice of action of dissatisfied customers:
– individuals must be able to compare outcomes of available choices (voice, exit, and
loyalty).
– return of voice must be higher than the alternative - to exit or not act.
– the relative returns depend on both the availability of an exit option and the
community’s ability to combine for collective action.
– the propensity to resort to the voice option depends also on the general readiness of
a population to complain and the existence of institutions and mechanism that can
communicate complaints cheaply and effectively.
outline
•
the newspaper campaign and its effect on local capture (details in Reinikka and
Svensson (2004b)
•
effects of the newspaper campaign on enrollment and student learning (preliminary
evidence from Reinikka and Svensson (2004c), Reinikka and Svensson (2005), and
Björkman 2004)
capitation grant program
•
national program that prescribes a set amount to each student
•
local (district) offices are used as distribution channels
– opportunity to capture the funds
•
unlike other government programs, the capitation grant was a rare liquid money infusion
into a local administrative and political system
– facilitates the capture of funds
•
school-specific measure of capture sjt is grants received (by school j in year t) as a share
of the grants disbursed by the central government to that school.
•
data from central ministry (disbursments), the local government/districts, and schools
(grant receipts).
post campaign improvement
All schools
1995
2001
Mean
Median
St. dev.
23.9
81.8
0
82.3
35.1
24.6
Obs
229
217
estimating the impact of improved access to information
•
in Reinikka and Svensson (2004b) we use two complementary approaches
– exploit the differences in access to newspapers across schools.
– measure exposure to the information campaign directly by testing head teachers’
knowledge of the grant program and then instruments for it by using distance to the
nearest newspaper outlet as instrument.
valid instrument?
•
must affect the school’s (HT and community) exposure to new information about the grant
program but have no direct effect on its ability to claim funds from the district
•
distance to a NP outlet captures the cost and ease of accessing a newspaper
– correlated with both the school’s and the community’s likelihood of being exposed.
•
NP penetration is partly determined by logistic factors outside the school/community’s control
•
explain the diff. in sj  underestimate the effect if j and distancej are correlated.
•
“standard instrument” [Card (1993), McClellan et al (1994)]
•
exogenous and time invariant
valid instrument I: distance is correlated with newspaper access of HT
Specification
(1)
Dep. var.
(2)
newspaper
Distance to nearest
newspaper outlet
-0.100***
(.018)
-0.098***
(.020)
Income as control
no
yes
Schools
388
388
valid instrument II: distance is uncorrelated with test scores of local
knowledge and ”general ability”
Spec.
Dep. var.
(1)
(2)
Knowledge Knowledge
about grant
about
formula
timing
(3)
(4)
(5)
(6)
Info
Knowledge
about news
events
Knowledge
about local
affairs
General
political
knowledge
-0.063***
(.021)
-0.040**
(.020)
-0.103***
(.029)
-0.039***
(.010)
-0.001
(.004)
-0.013
(.010)
Range of
scores
[0,1]
[0,1]
[0,1,2]
[0,1]
[0,1]
[0,1]
Average test
0.65
0.24
0.89
0.65
0.75
0.57
Schools
388
388
388
388
388
388
Distance to
nearest NP
outlet
valid instrument III: distance does not proxy for remotness and
community effect seems important
Specification
(5)
(6)
Dep. var.
info
info
Distance to nearest
newspaper outlet
-0.080**
(.038)
-0.096***
(.033)
Distance to district
headquarters
-0.065
(.060)
Distance to nearest
bank branch
0.021
(.060)
0.148**
(.075)
Newspaper
Income as control
yes
yes
Schools
388
388
valid instrument IV: uncorrelated with changes in capture in the 5year period prior to the NP campaign
Specification
(1)
(2)
Panel B: Control experiment (1991-1995)
Distance to nearest
newspaper outlet
23.7***
(5.4)
0.64
(2.24)
18.6**
(7.70)
0.62
(2.55)
Income as control
Adj. R2
Schools
no
0.01
147
yes
0.01
147
1995
newspaper campaign and its effect on local capture
Specification
(1)
(2)
Regression
Dep. variable
1st stage
info
2nd stage
s
Distance to nearest
newspaper outlet
-0.103***
(.029)
2001
2.30
(21.1)
65.9***
(23.5)
[23.6]
Info
Income as control
Yes
Yes
Schools
388
199
a one standard deviation increase in info results in a 1.1 standard deviation increase in
spending reaching the schools (i.e., 44.2 percentage points increase in funding reaching the
school between 1995 and 2001)
effects of the newspaper campaign on enrollment and student
learning
•
•
•
use distance to the nearest newspaper outlet (distance) as instrument.
instead of estimating the whole chain of events from distance through improved
information to more funds, we first link distance to funding.
this regression then forms the first stage of a two-stage least squares procedure of the
impact of reduce capture on school enrollment.
sj = constant + 1  xj + 2distancej +j
(FIRST STAGE)
studentsj = constant + β1  xj + β2  sj +j
(SECOND STAGE)
effects of the NP campaign on enrollment
Specification
Dep. variable
Period
Constant
(1)
students
95-01
(2)
s
95-01
(3)
students
95-01
(4)
students
95-01
(5)
students
91-95
450***
(20.3)
74.1***
(6.77)
-5.74**
(2.45)
29.7
(286)
574***
(49.3)
-37.7**
(17.9)
68.4**
(33.6)
-4.7
(12.1)
Distance to nearest
newspaper outlet
Share of funding
reaching school
7.55*
(4.62)
Income (control)
Yes
Yes
Yes
Yes
Yes
Schools
202
188
188
202
153
a one standard deviation increase in Δs results in a 0.66 standard deviation increase in school
enrollment (i.e., 297 more students in school), but the IV estimate (reg. 3) is marginally
insignificant (at 10 %) when reg 3. is estimated with robust standard errors.
effects of the NP campaign on student learning (PLE grade 7)
•
replace students with test score
•
utilize more data to estimate
test scoresjt = γ1xjt +γ2distancejσt + μj + δt + jt
effects of the NP campaign on student learning (PLE grade 7)
Specification
Years
(1)
Test
scores
95, 01
(2)
Test
scores
95, 01
Method
OLS
IV
Distance*post-campaign periods
-0.509
(.329)
Dep. variable
(3)
Test
scores
91, 93,
94, 95,
01, 02
OLS
(4)
Test
scores
91, 93,
94, 95,
01, 02
OLS
(5)
Test
scores
91, 93,
94, 95,
01, 02
OLS
(6)
Test
scores
91, 93,
94, 95,
01
OLS
-0.545***
(.176)
-0.545**
(.257)
-0.604**
(.258)
-0.612*
(.351)
Yes
Yes
0.084*
(.050)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
362
633
Yes
Yes
213
374
Yes
Yes
377
1829
Yes
Yes
377
1829
Yes
Yes
374
1811
Yes
Yes
Yes
Yes
Yes
Yes
213
822
Share of funding reaching school
Income
Grade 7 students
Students-teacher ratio
Students (log)
Time effects
School effects
Schools
Observations
complementary evidence based on district averages (Björkman 2004)
•
Björkman’s study is based on data on all Grade 7 students in Uganda, combined with
district specific data on newspaper circulation.
•
by looking at district averages, she can to a large extent get around problems of sorting
and selection.
•
Björkman exploits the fact that exposure to information about the per-student capitation
grant, and thus funding according to the results reported above, varied by district.
•
she uses the variation in newspaper circulation per district as an instrument of exposure
and a differences-in-differences approach
evidence based on district averages: results
•
•
•
student enrollment increased significantly faster in districts with high newspaper
circulation
using a similar approach, she finds that the newspaper campaign had a large effect on
student performance.
on average, pupils in districts which were highly exposed to the per-capitation grant
program scored 0.42 standard deviations (1.46 points) better in the Primary Leaving
Exam than pupils in districts that were less exposed to the program.
conclusion
•
•
•
•
in this paper we have argued that innovations in governance of social services may yield
the highest return since social service delivery in developing countries is often plagued
by inefficiencies and corruption.
we illustrate this by using data from an unusual policy experiment: a newspaper
campaign in Uganda.
while preliminary, the evidence suggests that both enrollment and test scores increased
significantly more in schools in areas with higher newspaper penetration, and thus in
schools with more funds.
since schools were free to spend their grants on whatever nonwage items they needed,
be it textbooks, school meals, school uniforms, or flipcharts (or even to boost wages
since funds generated by the schools themselves are fungible), these findings suggest
that experimentation and evaluation of processes and institutions that improve voice and
accountability, and thus indirectly enhance the learning environment in schools by
ensuring that entitlements actually reach their beneficiaries, should be high on the policy
and research agenda.