Content uploaded by Jean-Paul Faguet
Author content
All content in this area was uploaded by Jean-Paul Faguet
Content may be subject to copyright.
Decentralization of Health and Education in Developing
Countries: A Quality-Adjusted Review of the Empirical
Literature
Anila Channa
Jean-Paul Faguet
This Version: July 11, 2012
Abstract
We review empirical evidence on the ability of decentralization to enhance preference
matching and technical efficiency in the provision of health and education in developing
countries. Many influential surveys have found that the empirical evidence of
decentralization’s effects on service delivery is weak, incomplete and often contradictory.
Our own unweighted reading of the literature concurs. But when we organize the
evidence first by substantive theme, and then – crucially – by empirical quality and the
credibility of its identification strategy, clear patterns emerge. Higher quality evidence
indicates that decentralization increases technical efficiency across a variety of public
services, from student test scores to infant mortality rates. Decentralization also improves
preference matching in education, and can do so in health under certain conditions,
although there is less evidence for both. We discuss individual studies in some detail.
Weighting by quality is especially important when evidence informs policy-making.
Firmer conclusions will require an increased focus on research design, and a deeper
examination into the prerequisites and mechanisms of successful reforms.
Keywords: Decentralization, School-Based Management, Education, Health, Service
Delivery, Developing Countries, Preference Matching, Technical Efficiency
2 of 46
1. Introduction
In the late 1990s it was estimated that 80 percent of the world’s countries were
experimenting with one form or another of decentralization (Manor 1999). Since then,
new or deepening reforms have been announced in nations as diverse as Bolivia,
Cambodia, Egypt, Ethiopia, France, Indonesia, Japan, Mexico, South Korea, Turkey, and
many others. By now it is safe to say that experiments with, and enthusiasm for,
decentralization are essentially ubiquitous across the globe.
Theories underpinning such enthusiasm are compelling and argue that by taking the
government “closer to the people”, decentralization can improve the responsiveness and
accountability of the state, decrease corruption, increase the political voice and
participation of ordinary citizens, and also reduce bureaucracy and lower the unit costs of
government expenditure. The slogan “closer to the people” can be decomposed into
three underlying analytical advantages that local governments have over central
government: (a) superior information on local conditions and needs, (b) greater
participation of citizens in decision making and the production of local services, and (c)
greater accountability of public officials to voters. The local governments possessing
such advantages preside over jurisdictions that are smaller and more homogeneous than
those of national government. Local governments’ decision making will thus be
facilitated by not having to cater to a more diverse set of needs and wants. With superior
information, participation, accountability, and policy challenges that are less onerous, it
follows logically that decentralization should improve public services.
Yet the many surveys of the literature overwhelmingly agree that empirical evidence
is inconclusive. In one of the earliest reviews, for instance, Rondinelli et al. (1983) note
that decentralization seldom, if ever, lived up to its promise. Shah et al. (2004) concur in
a review of 56 studies published since the late 1990’s, chronicling that decentralization
in some cases improved, and in others worsened, service delivery, corruption and growth
across a large range of countries. Treisman’s (2007) more recent survey is bleaker still.
“To date,” he says, “there are almost no solidly established, general empirical findings
about the consequences of decentralization” (p. 250). The lack of consensus on
decentralization’s effects over 25 years and literally hundreds of studies is striking.
One of the main challenges faced by such review efforts is the sheer size and
diversity of scholarship. The empirical literature on decentralization originates from a
variety of disciplines, including public economics, development studies and comparative
3 of 46
politics, to name just a few. Evaluations of reforms are done in markedly different ways
and focus on very different outcomes, ranging from service delivery to corruption to
macro-economic stability and happiness. Any attempt to review these results as a whole
quickly loses the forest for the trees in a confusion of particular findings that may appear
contradictory, but are more often simply different. To draw firmer conclusions from this
vast literature, we argue, a clearer organizing principle is required – a principle that
allows students of decentralization to neatly compare the causal effects of a similar kind
of reform on similar predefined outcomes.
In this review we apply such an organizing principle to assess decentralization’s
ability to enhance service delivery in developing nations. Decentralization, defined here
as “the transfer of authority for decision-making, finance, and management to quasi-
autonomous units of local governments” (Litvack and Seddon 1999: pp.3), is probably
the single most advocated measure for improving the provision of health and education
in the developing world. This popularity is not surprising. Of the many arguments in
favor of decentralization, the most important is that devolving power and resources to
local governments can increase the accountability of public servants, and hence the
responsiveness of public services to citizens’ needs (Faguet 2012 and 2008). While
decentralizers’ motives have no doubt differed across different countries, improved
delivery of public goods has been at least an implicit goal of most reforms, and usually
an explicit one.
To ensure that our conclusions on decentralized public provision of health and
education are not influenced by an arbitrarily selected group of studies, we use
predefined criteria to identify papers for inclusion in this survey. Specifically, we focus
on empirical evidence in the economics literature from the last 20 years that evaluates
the causal effect of decentralization on service delivery in developing nations. We group
these studies according to the main substantive themes they address as follows: 1)
Preference matching1, defined as the extent to which public goods provided by local
governments match citizens’ preferences or demands, and/or 2) Technical efficiency,
meaning the production of more or better public goods by a decentralized government
for a given set of inputs. In the latter theme, we further segregate the evidence into sub-
categories based on whether it concerns the provision of (i) health, (ii) education to
1 In the decentralization policy literature, the term allocative efficiency is often used to refer to this same
concept.
4 of 46
lower tiers of the government, or (iii) education to schools in what is commonly referred
to as a School-Based Management (SBM) reform.
Within this thematic classification, we further classify studies according to the self-
reported quality of their data and credibility of their identification strategies, and place
greater weight on what high-quality evidence has to say. Distinguishing between studies
that are able to tease out the causal effects of decentralization more plausibly than others
is the crucial step that allows us to identify patterns in the findings. Earlier empirical
contributions on decentralization were commonly plagued by problems of attribution –
surveys based on such evidence therefore had similar challenges in isolating the effects
of reform. In recent years a deeper appreciation of the pitfalls associated with causal
inference has pushed empiricists to find more credible identification strategies that use
observational data to construct valid counterfactuals, and thus approximate the ‘gold
standard’ of randomized experiments2. This is the higher quality literature we focus on in
our review.
The following sections lay out our methodology and results, and then discuss and
contrast individual studies in some detail. But it may be useful to first summarize our
findings.
The overall evidence base is thin, although this varies by category
We find that the overall evidence base on decentralized health and service
delivery in developing countries is thin. Only 35 studies meet the selection
criteria detailed in section 2 below.
We also find that the distribution of scholarship is skewed by theme (preference
matching vs. technical efficiency) and sub-category (health, education, SBM).
For example, many more studies focus on how decentralization affects technical
efficiency than preference matching. Likewise, education and SBM have been
the subject of examination much more often than health. The papers reviewed
are summarized by theme and sub-category in Table 1.
2 See Angrist and Pischke 2010 for a good discussion on identification strategies
5 of 46
The econometric techniques employed by studies are less sophisticated than we
would prefer
Fewer than a third of the papers reviewed can be classified as having a highly
credible identification strategy. Our categorization hinges on the ability of the
methodologies employed to mitigate endogeneity concerns, in accordance with
the established hierarchy of econometric techniques. Hence, for example,
randomized and quasi-randomized evidence are considered to have stronger
identification strategies than cross-sectional work. Table 2 describes how we
categorize the credibility of a study’s empirical design in detail.
In this particular sense, the “quality” of the studies reviewed also varies
substantially by theme and sub-category. Papers in the technical efficiency
theme, and specifically studies investigating school decentralization reforms,
appear to have a greater number of high quality contributions. By contrast,
contributions in the preference matching category are not only fewer but also less
rigorous, making the task of drawing conclusions from this group difficult.
Externalities in health drive pessimism in the preference matching theme
Our review indicates that pessimism in the small preference matching literature
is due primarily to the externalities that characterize the health sector.
Decentralized local governments often match local preferences more efficiently
while ignoring spillover effects on neighboring regions, as some of the classic
public economics literature predicts (Oates 1972; Rubinfeld 1987), thus reducing
overall social welfare. The evidence of preference matching in education
delivery, on the other hand, appears to be somewhat positive. But the small size
of this body of work limits firmer conclusions.
Higher quality work on technical efficiency appears to be favorable
Importantly, evidence on technical efficiency is on the whole optimistic. This
optimism rises with the quality of the evidence. The highest quality empirics
show that decentralization can enhance a variety of service delivery outcomes,
from student test scores to infant mortality rates.
6 of 46
Although such results are not conclusive, they do demonstrate the potential of
decentralization to enhance service delivery in developing countries. Stronger
conclusions are not possible until the field sees a more general shift towards
better research design, and the development of a deeper understanding of the
prerequisites and mechanisms of successful reforms.
The rest of this paper proceeds as follows. Section 2 describes the criteria used for
including and classifying studies in this review. Section 3 discusses the papers included
by theme. We conclude by comparing our findings to broad surveys of the literature and
suggesting priorities for future research. The papers reviewed are summarized in Tables
1 to 4. Our aim is to provide insights into patterns of findings on one piece of the larger
decentralization puzzle. What follows hopefully helps to answer some important
question surrounding decentralized service delivery of health and education in
developing countries. Even so, we do not pretend that this survey can cover more than a
fraction of a huge literature.
2. Scope and Methodology of Review
This section describes the steps we undertook to identify, organize and classify
studies from this vast literature for our review. Our strategy was to conduct a wide search
and then systematically filter papers that met predefined criteria for relevance. Following
this, we arranged the literature first by substantive theme, and then – crucially – by
quality of the evidence. The latter step allows us to see empirical patterns in the
scholarship that broad decentralization reviews have not previously uncovered.
2.1 Identifying and Organizing the Literature
We began our search by focusing on published and unpublished working papers
found on EconLit, an established bibliography of economics literature. Using the key
words “decentralization” or “devolution” in conjunction with either “education”,
“health” or “service delivery”, we first conducted our search in October 2010, and then
subsequently updated it in December 2011.
The EconLit database yielded approximately 350 studies on decentralization and
health, and some 350 more on decentralization and education. To ensure thorough
coverage we then cast a wider net, consulting other key resources such as Google
7 of 46
Scholar, JSTOR, SCOPUS and Web of Knowledge. We also referred to major
publications by international organizations such as the World Bank, used citation
indices, and reviewed reference lists in identified papers to confirm that no critical
contributions were omitted from our review. This pushed the total number of studies
above 1,000.
Based on the information contained in their abstracts, we selected those studies that
a) were of an empirical nature, b) dealt specifically with decentralization of service
delivery of health and/or education in developing nations3, and c) date from 1992 or
later. This reduced the sample dramatically to 35 contributions. We then organized this
short-listed body of evidence into our two themes: preference-matching, and technical
efficiency. The body of scholarship on technical efficiency is larger, and therefore we
further segregated the papers examined into sub-categories based on whether they
address: (1) decentralization of health, (2) decentralization of education to lower tiers of
governments, or (3) decentralization of education to schools or School-Based
Management.
Table 1 summarizes the evidence by theme, sub-category, author, publication type,
and countries covered.
3 For a good review of empirical evidence from OECD and other developed nations, refer to Ahmad,
Brosio and Tanzi (2008) and Ahmad and Brosio (2009).
Table 1: Summary of Evidence Reviewed
Region Author
PM TE Health Educ. SBM Journal Book Working
Paper Other
Unpublished Description
Africa
1Kenya Duflo et al (2007) X X X Unpublished manus cript
2Madagascar Glewwe and Maiga (2011) X X X Journal of Development Effecti veness
3Uganda Akin et al (2005) X X X Journal of Development Studi es
Asia
4China Uchimura and Jutting (2009) X X X World Development
5India Asfaw et al (2007) X X X Journal of Developing Areas
6Indonesia Skoufias et al (2011) X X X X World B ank Poli cy Research Paper S eries
7Pakis tan Hasnain (2008) X X X X W orld Bank Polic y Rese arch Paper Series
8Pakis tan Aslam and Yilmaz (2011) X X X Public Adminis trati on and Development
9Philippi nes Schwartz et al (2002) X X X MEAS URE Evaluation, University of North Carolina at Chapel Hill
10 Philippines Jim enez and Paqueo (1996) X X X Ec onomics of Educat ion Review
11 Philippines Loc kheed and Zhao (1993) X X X International Journal of Educat ional Development
12 Philippi nes Khatt ri et al (2010) X X X World B ank Poli cy Research Paper S eries
13 Russia Frienkman and Pl ekhanov (2009) X X X X W orld Development
Latin & Central America
14 Argentina Habibi et al (2003) X X X X Journal of Human Development
15 Argentina Galiani et al (2008) X X X Journal of Public Ec onomic s
16 Argentina Es keland and Fi lmer (2007) X X X World B ank Poli cy Research Paper S eries
17 Bolivia F aguet (2012) X X X X Book
18 Bolivia F aguet (2004) X X X X Journal of Public E conomic s
19 Bolivia Inc hauste (2009) X X X X Book Chapter
20 Brazil Paes de B arros and Mendonca (1998) X X X Organizati on matters : agency problems i n health and educat ion in Lati n Americ a
21 Chile Di Gropello (2002) X X X X World Bank Ec onomist s' Forum
22 Colombia F aguet and Sanchez (2008) X X X W orld Development
23 El Salvador Jimenez and Sawada (1999) X X X The World B ank Ec onomic Review
24 El Salvador Sawada and Ragatz (2005) X X X Book Chapter in World B ank Publ icati on
25 Honduras Di Gropello and Marshall (2005) X X X Book Chapter in World B ank Publi cati on
26 Mexic o Gertler et al (2011) X X X Journal of Devel opment Ec onomic s
27 Mexic o Skoufias and Shapiro (2006) X X X W orld Bank P olic y Researc h Paper Series
28 Nicaragua K ing and Ozl er (2000) X X X World Bank P olicy Research P aper Series
29 Nicaragua P arker (2005) X X X Book Chapter in World B ank P ublicat ion
Cross-country
30 Various Arze del Granado et al (2005) X X X X Georgia St ate University
31 Various Gunnarsson et al (2009) X X X Economi c Development and Cultural Change
32 Various Hanushek et al (2011) X X X NBER W orking Paper S eries
33 Various Khaleghian (2004) X X X S ocial S cienc e and Medici ne
34 Various Robalino et al (2001) X X X World Bank Poli cy Res earch Paper S eries
35 Various Treisman (2002) X X X X Unpublished manus cript
Total 7 28 121614 18 4 9 3
Notes
PM: P reference Matc hing
TE: Technical Efficiency
SBM: Sc hool-based Management
By Theme By Type of PublicationBy Sub-Category
2.3 Quality of the Evidence
Next we evaluated the quality of the evidence. We do so in the knowledge that
researchers attempting to assess the effects of decentralization on education and health
services face a number of difficult challenges. These include the difficulty of
disaggregating decentralization’s effects from (a) the other reforms that often accompany
it, and – more importantly – (b) from country politics, which necessarily plays a crucial
role in both decentralization’s impetus as well as its eventual effects. Together these
impose sizeable data demands on researchers. Add to these problems the time it takes for
service delivery outcomes to change, and the difficulty of conducting randomized
decentralization experiments, and the varying quality of the evidence seems fully
justified.
To classify the persuasiveness of each paper’s identification strategy, we use a four
point scale of Very Strongly Credible, Strongly Credible, Somewhat Credible, and Less
Credible. In our categorization, the primary consideration is the nature of the data
available and the identification strategy this permits. In effect, we rank papers’ empirical
methodologies according to their widely accepted abilities to mitigate endogeneity
concerns and identify causal effects. For this we rely on the established hierarchy of
identification strategies in economics as widely taught in graduate programs today. We
supplement our quality distinctions by reviewing the covariates included in analysis, the
measures used for decentralization4, the self-reported quality of data, and the nature of
robustness checks performed in the paper. This scale, along with a snapshot of how
papers in this survey have been classified, is presented in Table 2. The categorization is
adapted from a similar typology by Santibañez (2006).
4 For a good discussion on challenges of measuring decentralization, see Ebel and Yilmaz (2002)
Table 2: Quality Distinctions
Scale Criteria
Generally Positive
Findings Generally Insignificant or
Negative Findings Generally Positive
Findings Generally Insignificant or
Negative Findings
Health
-▪ Hasnain (2008)
▪ Schwar tz et al (2002)
-▪ Treism an (2002)
Educatio
n
-▪ Hasnain (2008)
▪ Freinkman and Plekhanov
(
2009
)
▪ Di Gropello (2002)
▪ Jimenez and P aqueo (1996)
▪ Lockheed and Zhao (1993)
▪ Treism an (2002)
SBM
-- ▪ Eskeland and Film er (2007)
Health
▪ Arze del Granado et al
(2005)
▪ Akin et al (2005) ▪ Asfaw et al (2007)
▪ Robalino et al (2001)
▪ Habibi et al (2003)
▪ Inchauste (2009)
▪ Khaleghian (2003)
Education
▪ Arze del Granado et al
(2005)
-▪ Aslam and Yilmaz (2011)
▪ Freinkman and Plekhanov
(2009)
▪ Inchauste (2009)
SBM
-- ▪ Paes de Barros and
Mendonca (1998)
▪ King and Ozler (2000)
▪ Khattri et al (2010)
▪ Parker (2005)
▪ Di Gropello and Marshall
(2005)
▪ Gunnarsson et al (2009)
▪ Jimenez and S awada (1999)
▪ Sawada and Ragataz (2005)
Preference Matching Technical efficiency
Less Credible
Identification
Strategy
▪ Researc h that bas es findings on s elf-s elected populations, and
makes little to no effort to produce a valid com parison group.
▪ Work that is likely to suffer from serious omitted variable bias and
other endogeneity issues suc h as those related to meas urem ent
error due to s elf-reported poor quality of data.
▪ Most cross-s ectional work that does not us e any other
sophis ticated methodology to address endogeneity will fall in this
category.
Somewhat
Credible
Identification
Strategy
▪ Research that attempts to c onstruct a valid c ompar ison gr oup but
does so with limited success.
▪ Work that is likely to c ontinue to s uffer from s ome endogeneity
biases in spite of efforts at mitigation.
▪ Cross -s ectional work that uses matc hing tec hniques, for instance,
falls in this group. O ther studies us ing panel estimation m ay also fall
in this group if they use random effec ts or between effects . Papers
using difference in differ ences but without providing any support of its
key identifying assumption of parallel trends , papers using IV that are
not cons idered particularly s trong and papers using fixed effects but
with very lim ited covariates als o fall here.
▪ Finally, preferenc e matc hing studies that only establis h a change in
allocation patterns but not any enhanced alignm ent to citizen
preferenc es als o are in this category.
▪ Compared to previous c ategory, w ork in this group generally
attempts to validate the m easure of decentralization through use of
multiple m easures or qualitative validation.
11 of 46
Health
-▪ Faguet (2004)
▪ Faguet (2012)
▪ Skoufias et al ( 2011)
▪ Uchimura and Jutting (2009)
Education
▪ Faguet (2004)
▪ Faguet (2012)
▪ Skoufias et al ( 2011)
-▪ Galiani et al (2008)
▪ Faguet and Sanc hez (2008)
SBM
-- ▪ Skoufias and Shapiro
(2006)
▪ Gertler et al (2011)
▪
Hanushek et al (2011)
Health
-- - -
Educatio
n
-- - -
SBM
-- ▪ Duflo et al (2007) ▪ Glewwe and Maiga (2011)
▪ Studies with very strong research design such as randomized
control trials that have a valid contr ol group fall here.
▪ Work that is likely to have lim ited endogeneity c oncerns.
Very Strongly
Credible
Identification
Strategy
Strongly Credible
Identification
Strategy
▪ Research that is able to cons truc t a reas onable comparis on group.
▪ Work that spec ifically attempts to address sources of endogeneity
and is mostlys succ ess ful in its attempt.
▪ Most of the studies in this category quasi-experimental designs
suc h as difference in differenc es and instrumental var iables. Papers
providing panel estimates in a fixed effec ts model w hile contr olling for
more than one soc io-ec onomic c ovariate and m ore than one
covariate from the health/education production function also fall here.
Our top category, Very Strongly Credible, consists of randomized control trials
(RCTs), the ‘gold standard’ for identifying causal effects. At the other end of the
spectrum, work that relies on simpler quantitative methods such as ordinary least squares
(OLS), and fails to employ any more sophisticated methodology to control for
endogeneity bias, is categorized as having a design that is Less Credible in drawing
causal inferences. The papers we place in this category are mostly cross-sectional OLS
analyses of observational data. Because studies here attempt to draw findings from self-
selected populations without being able to construct a valid comparison group, their
ability to make causal claims is limited when compared to studies that use more
sophisticated econometric methods.
How do we place work into the two middle groupings? The Strongly Credible
category consists of research that is reasonably successful in producing a valid
comparison group. Much of this literature uses quasi-experimental techniques such as
instrumental variables (IV) or difference in differences (DID) approaches. The key issue
for being classed as Strongly Credible is how persuasive studies are in communicating a
thorough understanding of the institutional environment and then – importantly – using
this understanding to design their empirical strategy.
So for instance, Strongly Credible papers using IV techniques can make plausible
claims for the relevance and exogeneity of their instruments. Studies using difference in
differences can persuade that the treatment is what is responsible for altering a trend
between treatment and control groups. For this reason, the category also contains some
panel data estimations using fixed effects and a set of relevant covariates, but only where
the case for limited endogeneity based on knowledge of confounding factors is
particularly convincing.
The remaining studies are classed as Somewhat Credible. In our view, studies in this
category are less persuasive in addressing endogeneity than those that are Strongly
Credible, but more convincing than the Less Credible set due to their use of various
kinds of comparison groups. This category thus houses diverse econometric methods,
from matching to instrumental variables.
It is worth underlining what this survey does not seek to do. We recognize that the
identification strategies employed by researchers are largely determined by a
combination of the data available, the nature of the reforms implemented, and the
nuanced questions they seek to answer. Hence we make no attempt to rank papers’
broader quality as pieces of research, nor comment on the analytical skills of their
13 of 46
authors. What we do seek to do, rather, is recognize that there is an established hierarchy
of rigor in econometric identification, and apply that hierarchy to the evidence that the
literature provides. This allows us to roughly categorize how convincing studies’ results
are, where credibility is principally determined by what data is available, and hence how
we should weight evidence when making policy.
Henceforth we focus on the results of studies falling into the three stronger quality
categories. Studies in the Less Credible category are occasionally highlighted when they
present specific policy ideas that relate to studies in the more credible groups.
14 of 46
3. The Effects of Decentralization on Education and Health
We now move on to the heart of our review, and describe the papers included in this
survey in some detail. We first discuss preference matching and then turn to technical
efficiency. Throughout this section, we make use of our quality distinctions when
describing key papers in order to allow readers to understand how our conclusions are
drawn.
3.1 Preference Matching
Although preference matching is one of the classic arguments posited in favor of
decentralization (see Oates 1972), the empirical evidence devoted exclusively to testing
this proposition is surprisingly small. It also produces somewhat contradictory results for
the service delivery of education on one hand, and the provision of health on the other.
The book and two papers we review with Strongly Credible identification strategies
in the theme of preference matching yield somewhat contradictory findings. In one of the
first papers to employ a before and after estimation strategy to examine preference
matching, Faguet (2004; see also Faguet 2012 for a substantial update) finds favorable
evidence from Bolivia. Bolivia undertook devolution in 1994, and as part of the reform
moved responsibility of key public services to local governments. The shift in
responsibility was accompanied by two other critical changes – the doubling of funds
available to these devolved units during this period, and the establishment of oversight
bodies to monitor local spending.
Faguet examines the patterns of investment in public investment projects in a total of
ten categories, including education and health, from 7 years before and 3 years after this
reform. By doing so first for municipal averages, and then one by one for all 311
municipalities examined, he finds a statistically significant increase in investment in
education overall, as well as a statistically significant increase in 71% of individual
municipalities in just three years after devolution.
This shift in investment patterns was especially evident in poorer regions. As
devolution increased funding to previously neglected regions, this finding is not
necessarily an indication of greater preference matching. But Faguet then offers further
evidence to support his responsiveness argument – he demonstrates that regions with
high illiteracy levels, or where there seemed to be a greater need, invested more heavily
in education. Regions with strong education indicators, on the other hand, prioritized
15 of 46
other sectors. This, he contends, “implies that local government is more sensitive to local
need than central government” (p. 24). The author’s optimism is supported by similar
findings in the sectors of water management and urban development, but noticeably not
in our second sector of interest here – health.
However, greater spending on socially oriented sectors does not necessarily imply
that preference matching has improved. This is the main contrasting finding in Skoufias
et al.’s (2011) recent working paper on Indonesia. The paper exploits an arguably
exogenous phasing of local direct elections to conduct a difference in differences
analysis of the effect of political decentralization on the pattern of public spending.
Although fiscal and administrative devolution commenced in 1999 in the country, in
2005 Indonesia implemented electoral reforms to enhance accountability in service
delivery. The date local elections were held in a particular municipality depended on
when the previously appointed head of government completed their tenure. Deviations in
this timing were a result of illness, death, no confidence votes and the creation of new
districts and thus, claim the authors, exogenous to spending allocation.
Skoufias et al. compare changes in expenditure patterns in districts that held local
elections in 2005 to patterns in districts that did not hold elections until 2008, after
providing proof that pre-implementation trends in spending in both groups were similar.
They find that political decentralization was associated with greater overall public
spending – when disaggregated, however, they demonstrate that while there was an
increase in the education sector, there was no significant difference in health spending.
Skoufias et al. then follow Faguet (2004), attempting to use his methodology to establish
whether these shifting patterns were based on local needs. In contrast to Faguet,
however, they find no evidence to suggest an improvement in preference matching at all.
The two contributions from the Somewhat Credible identification group are also
contrary. In the only cross-country study concerned with preference matching that we
review here, Arze del Granado et al. (2005) seek to establish that Faguet’s (2004)
findings on the change in functional composition post decentralization in Bolivia are
“not a unique experience of a specific country” (p. 4). Employing a before and after
strategy similar to Faguet, but using data for 45 developed and developing countries over
28 years, Arze del Granado et al. analyze the relationship between the ratio of local
expenditure to total government expenditure as the measure of decentralization, and the
ratio of health and education spending to overall spending.
16 of 46
The authors employ a total of five econometric models to ensure robustness of their
results, including OLS with fixed and random effects. They control for per capita
income, budget balance and population and use time effects in all, and country effects in
some, of their specifications. They find a statistically significant relationship between
decentralization and expenditure ratios. Because, they assert, “implicit in the argument
that decentralization can increase allocative efficiency is the implication that
decentralization is likely to alter the composition of public expenditures” (Arze del
Granado et al. 2005: p. 2), they concur with Faguet (2004) and conclude in favor of the
potential of decentralization to enhance preference matching.
Akin et al. (2005) take a slightly different tack. They attempt to provide a deeper
understanding of spending allocations within the health sector after a decentralization
reform occurred in Uganda. The authors postulate a model in which users undervalue
public-type health goods such as family planning, health education, immunization and
infectious disease control. Because local governments will be more responsive to the
preferences revealed by their residents for private-type health goods, the authors posit
that districts will under-provide public-type health care and ignore spillover effects on
neighboring regions if they are not under the same jurisdiction.
Akin et al.’s theory is borne out in the district-level data they examine from Uganda.
They find, after controlling for per capita income in a fixed effects model, that
decentralization is associated with higher budgeting of private-type health goods.
Because their main empirical strategy compares districts that decentralized earlier to
those that decentralized later, the authors attempt to provide validation that the groups
did not differ systematically. Moreover, Akin et al. also cite evidence in favor of
crowding-out effects – districts whose neighbors budget higher amounts on public-type
goods budget less on such goods themselves. On the basis of this evidence, Akin et al.
(2005: p.3) pessimistically call for “A reappraisal of the central government’s role in
providing public goods in developing countries”.
What is interesting, however, is that their argument is not one against the preference
matching effects of decentralization per se, as they assume local governments are indeed
responding to local inclinations in Uganda. Rather, their pessimism arises from the result
of the responsiveness versus spillover effects trade-off. The Oates Decentralization
Theorem (1972) suggests that devolution is superior only so long as there are no
spillover effects. In the presence of spillover effects, the theoretical prediction for
17 of 46
preference matching of decentralization is ambiguous or even negative (Besley and
Coate 2003; Bardhan and Mookherjee 1998).
The papers with Less Credible empirical designs take Akin et al.’s pessimism even
further. Schwartz et al. (2002), for instance, examine the trends in spending composition
of health services in 1600 regions in the Philippines to show, like Akin et al., a shift in
local spending composition from public-type health services to private-type curative
health care. Along a similar vein, Hasnain (2008) considers budget allocation trends in
Pakistan’s province of Punjab and reports that decentralized local governments are
prioritizing allocations for infrastructure over those for health and education. And in
sharp contrast to all of the studies above, Frienkman and Plekhanov (2009) do not find a
change in allocation patterns after decentralization in Russia at all. The authors use a
between effects model on cross-sectional data to conclude that fiscal decentralization is
not significantly associated with an investment in education inputs.
So what, if anything, can we take away from this short review of the evidence of
decentralization’s ability to enhance preference matching? The literature in this theme is
small, and the number of high quality contributions is even smaller. But studies across
the quality distinctions appear to mostly concur that decentralization changes the patterns
of local spending. On the other hand, whether or not these changes are responsive to
local needs is an area where there is less agreement. While the evidence appears
somewhat encouraging for enhanced preference matching in education, contributions in
the area of health are decidedly pessimistic due both to a lack of visible change in
allocation patterns and the possibility of externalities in the area.
Table 3: Selected Empirical Evidence of Preference Matching
Source: Cited arti cles
No. Author (Date) Country of
Study Date
Implemented Programme
Description Method of
Analysis Sample Measure/s of
Decentralization Results Identification
Strategy No.
Health
1Faguet (2004);
Faguet (2012);
Faguet and
Sanchez (2008)
Bolivia 1994 Increase in devolved funds to
LG, responsibility for public
services, es tabli shment of
oversight committees
OLS using a fixed
effects model
Universe of 311
regions over 1987 -
2007
▪ Binary measure of before
and after D implementation
▪ Investment in health did not change
significantly post D
Strongly Credible 1
2Skoufias et al
(2011)
Indonesia 1999 Increase in devolved funds to
LG, responsibility for public
services. In 2005, direct
election of local government
Difference in
diffe rences
200 out of 400 districts
during 2001 to 2006
▪ Binary measure of election
date
▪ Overall public expenditure increased
post D
▪ No significant change in health
spending post D
Strongly Credible 2
3Akin et al (2005) Uganda 1993 - 1994 Fiscal decentralization, with
rule-based unconditional
grants given to regions
OLS using a fixed
effects model
30 out of 45 regions
during 1995 - 1998
▪ Years since unconditional
grant given to LG
▪ Proportion of LG
expenditure financed by LG
reve nues
▪ Share of allocation to health decreased
in local budgets, esp. in preventive and
primary health care areas
▪ Some evidence of spillover, where
neighbours of high health spending
districts have lower health spending
Somewhat Credible 3
4Arze del Granado
et al (2005)
Cr os s-c ountry Va ri ous Var io us Vari ous i nclud ing
OLS using fixed
and random, also
QMLE models
45 developed and
developing countries -
Unbalanced panel over
1973 - 2000
▪ Share of LG expenditure in
total government
expenditure
▪ Higher D associated with higher
proportion of spending on health and
education
Somewhat Credible 4
5Hasnain (2008) Paki stan 2001 Limited financial devolution
but implementation of rule-
based transfers,
respo nsib ility for public
services, political devolution
Alloca tion trend
analysis
33 out of 35 districts in
Punjab over 2006-
2007
▪ Budget allocations to
sectors
▪ Relative size of LG spend
on sector
▪ LG spe nd focuse d on infra structure and
away from health
▪ Provinces driving health interventions,
providing incentives to shift away
budgets
Less Credible 5
6Schwartz et al
(2002)
Philippines 1991 Increase in devolved funds to
LG, responsibility for public
services
Spending trend
analysis
1600 LG - period
includes 4 yrs before
and 6 yrs after
▪ Binary measure of before
and after D implementation
▪ Greater % spend allocated to health
post D
▪ Higher spend on private health goods,
explained by expenses arising from
devolution of hospitals
Less Credible 6
19 of 46
Education
7Faguet (2004);
Faguet (2012);
Faguet and
Sanchez (2008)
Bolivia 1994 Increase in devolved funds to
LG, responsibility for public
services, establishment of
oversight committees
OLS using a fixed
effects model
Universe of 311
regions over 1987 -
2007
▪ Binary measure of before
and after D implementation
▪ Investment in education increases
significantly post D
▪ Investment increases are associated
with illi terac y levels
Strong ly Cred ible 7
8Skoufias et al
(2011)
Indonesia 1999 Increase in devolved funds to
LG, responsibility for public
services. In 2005, direct
election of local government
Di fference in
diffe rences
200 out of 400 districts
during 2001 to 2006
▪ Binary measure of election
date
▪ Overall public expenditure increased
post D
▪ Increase in spending on education post
D
Strong ly Cred ible 8
9Arze del Granado
et al (2005)
Cross-country Various Various Various including
OLS using fi xed
and random, also
QMLE models
45 developed and
developing countries -
Unbalanced panel over
1973 - 2000
▪ Share of LG expenditure in
total government
expenditure
▪ Higher D associated with higher
proportion of spending on health and
education
Somewhat Credible 9
10 Fre inkma n and
Plekhanov (2009)
Russ ia Phase d b eginni ng in
1994
Increased fiscal powers with
rule-based transfers,
responsibility for public
services
OLS using a
betwee n effects
model
73 out of 83 regions,
with data colle ction in
2003
▪ Share of LG education
exp end iture f inanc ed by own
reve nue
▪ Type of de centrali zatio n
arrangement
▪ No significant impact on inputs Less Credible 10
11 Hasnain (2008) Paki stan 2001 Limited financial devolution
but implementation of rule-
based transfers,
responsibility for public
services, political devolution
Allocation trend
analysis
33 out of 35 districts in
Punjab over 2006-
2007
▪ Budget allocations to
sectors
▪ Relati ve size o f LG sp end
on sector
▪ LG spe nd focuse d on i nfrastructure a nd
away from education
▪ Provinces driving education
interventio ns, provi ding incenti ves to shift
away budgets
Less Credible 11
Notes
PM: Allocative Efficiency
TE: Technical E fficiency
D: Decent ralization
LG: Local G overnment
3.2 Technical Efficiency
The body of work on the ability of decentralization to enhance technical efficiency in
the delivery of education and health fortunately is much larger than that found in the
previous theme. Strikingly, it is also more rigorous, and fairly optimistic of the potential
of decentralization to improve service delivery.
3.2.1 Health
The lone paper with a Strongly Credible empirical strategy in this sub-category, for
instance, provides the first piece of evidence strongly in favor of decentralization’s
ability to enhance technical efficiency in health delivery.
Uchimura and Jutting (2009) examine the interesting case of China, a country that
has had consistently high levels of spending decentralization, but a growing
recentralization of revenue decisions since 1994. Improving on previous studies that use
only province-level data, Uchimura and Jutting employ data from counties in 26
provinces over a seven year period. Counties in China have responsibility for
implementing health programs. However, local government officials are elected through
parties, not the adult franchise, which limits political accountability of officials to
citizens.
The authors determine the statistical relationship between two measures of county-
level fiscal decentralization and the outcome of provincial infant mortality rates (IMR),
while controlling for key health production function elements such as illiteracy rates,
fertility rates and per capita income. The main specification includes fixed effects.
Finding statistically significant and negative coefficients in most of the models tested,
the authors conclude that counties in more fiscally decentralized provinces have lower
IMR. Interactions between their two measures of decentralization – own expenditure
financed and proportion of provincial expenditure - are also positive. This suggests to the
authors that IMRs are lower in provinces not only where fiscal capacity is strengthened,
but also where counties and provinces have a functional transfer system in place.
Two contributions from the group with Somewhat Credible evidence are also
positive. In a study quite similar to the above, Asfaw et al. (2007) consider empirical
evidence on rural infant mortality rates from India. Decentralization in India has a long
history. But, it took its current form with the passing of the 1989 Panchayat Raj bill and
later constitutional amendments in the early 1990s that devolved power to the traditional
21 of 46
village organizations or Panchayats. Panchayats now form a part of the local
government, hold elections, and bear responsibility for health and education delivery.
Evidence suggests, however, that different states have followed differing models of
devolution, making comparative analysis of the reform difficult.
Nonetheless, Asfaw et al. attempt to estimate the role of devolution in affecting the
outcome of rural infant mortality rates using data from 14 states over seven years. In
order to improve the measure of fiscal decentralization, the authors use factor analysis of
three different indicators to build a decentralization index, including a Panchayat’s share
in total expenditure, Panchayat expenditure per rural population and share of
Panychayat’s own revenue in Panychayat’s expenditure. Controls are included for per
capita income of state, percentage share of literate women, and an index of political
decentralization (constructed from data on voter turnout, women’s participation in polls
and number of polling stations). Notably absent though are important health production
function components such as fertility.
The authors demonstrate a statistically significant and negative relationship between
decentralization and IMR in both their random and fixed effects models, but not in the
between-effects model. Asfaw et al. (2007) conclude that having an above average
decentralization index is associated with a 17.16% reduction as compared to states with
below average fiscal decentralization scores. The results hold when the measure of
decentralization is altered, when indices are made continuous measures and also when
two year averages of IMR are used.
The final positive single country study we review here is due to Habibi et al. (2003)
who consider devolution of basic health and education (see also next section) services in
Argentina. In their paper, Habibi et al. use nationwide data from over a 25 year period to
examine the relationship between two measures of fiscal decentralization and the infant
mortality rate, while controlling for per capita income, per capita expenditure and the
number of public sector employees. The authors present results from three models,
including OLS, GLS and OLS with fixed effects, finding a significant and negative
relationship between the parameters of interest. On the basis of these findings, the
authors conclude that devolution can have positive effects on human development,
especially when there is greater tax accountability in a province.
Less optimistically, Inchauste (2009) reports Bolivian evidence from the first half of
the 2000s in the context of the Highly Indebted Poor Countries (HIPC) initiative, which
directed resources saved from repayment of debt to local governments based on poverty
22 of 46
levels. Although she shows that there has been increased investment in both health and
education, she does not find a significant association between the number of poor in a
municipality and HIPC transfers, and argues that HIPC funds have not been targeted
well.
Using a random effects model, Inchauste also examines the relationship between the
change in health spending and (1) the change in share of unattended illnesses and (2) the
unattended cases of respiratory diseases, finding a significant decline in the former and
no significant change in the latter. The author controls for several socio-economic factors
such as family size and illiteracy. But the short time period studied presents two
important shortcomings: (1) the entire sample is in the post-decentralization period, and
(2) it rules out the use of municipal fixed effects, which implies that her results may
suffer from endogeneity despite the controls employed. Inchauste argues that there has
been a lack of improvement in social indicators based on this mixed result, as well as on
the results on education which we discuss in the next section. But the lack of pre-reform
data imply that her results may say more about administration of the HIPC initiative than
about decentralization per se.
What cross-country evidence do we have of associations of decentralization and
health service delivery? Two notable studies over the past decade investigate the impact
of decentralization on health service delivery, finding somewhat mixed results that
appear to depend crucially on the level of development of a nation.
In an oft-cited paper, Robalino et al. (2001) perform a cross-country data regression,
using IMR as the dependent variable, fiscal decentralization as the independent variable
and a set of controls for GDP per capita and institutional capacity variables such as
corruption, ethno-linguistic fractionalization and political rights. Like several other
papers in this section, however, the authors miss out on controlling for fertility, which
scholars argue is a key determinant of IMR (Treisman 2007). Their sample, though not
expressly given in their paper, comprises between 45-70 low and high income countries
and they rely on data from GFS.
Robalino et al.’s fixed effects model yields a significant and negative relationship
between the key measures of interest. In their basic model, if a country with a GDP per
capita of USD 2000 increases its share of expenditures managed by local governments
by 10%, this would be associated with a 3.6% decrease in mortality rates. Robalino et al.
also find evidence to show that the benefits associated with fiscal decentralization may
have a U shaped curve with respect to GDP per capita, implying that countries with low
23 of 46
and high incomes are more likely to benefit from the reform than middle income
countries. They conclude that decentralization benefits are “particularly important for
poor countries” (Robalino et al. 2001: p. 11).
According to Khalegian (2004), on other hand, the benefit curve is L shaped for
immunization. Using data on 140 low and middle income countries over 18 years, he
conducts a cross-country regression of a measure of decentralization against
immunization rates against measles and diphtheria. Unlike other papers in this area,
Khalegian uses a political, not a fiscal measure of decentralization, sourced primarily
from the Database of Political Institutions. Here, decentralization is measured as a binary
variable, indicating whether or not local governments have authority for taxing, spending
and regulation although measures of fiscal decentralization from GFS are also employed
in the regression. The typical socio-economic controls for GDP per capita, population
density, and illiteracy are included, in addition to measures of institutional quality, ethnic
conflict and fractionalization, and income inequality.
The author notes positive and significant coefficients for lower income decentralized
countries in his main specification which uses between effects and time dummies –
decentralization is associated with 8.8% and 8.3% increase in diphtheria and measles
coverage, respectively. In middle income countries, however, this reverses and
decentralization is associated with a decrease in diphtheria and measles coverage of
4.9% and 5.5% respectively. Analysis indicates that the turning point is per capita GDP
of 1400 (in 1995 USD), after which a negative relationship stabilizes.
Khalegian’s outlook based on these results is mixed with regards to decentralization
– after exploring some channels, he proposes that the difference between the results seen
in lower and middle income countries can be attributed to the possibility that poorer
countries decentralize less fully than middle income ones. This implies that were
complete devolution to occur, we would see a negative impact on immunization. The
author uses this to encourage continued central government support of health initiatives.
In the Less Credible category, Treisman (2002) uses OLS regression on cross-
sectional multi-country data to show that decentralization’s effects may be sensitive to
the income level of a country. Using data on 166 countries, Treisman explores how
having constitutional sub-national authority relates to two indicators of health care
performance – the share of infants inoculated for diphtheria, tetanus and pertussis, and
the share of population for which 20 essential drugs are available and affordable. Once
an extensive set of socio-economic controls are added, the paper’s findings are not
24 of 46
significant for sub-national authority, although greater electoral accountability is
correlated with better access to medicines. In his analysis, nations with per capita GNP
greater than USD 5000 have worse service delivery performance than their counterparts
with lower per capita GNP.
In sum, while the body of scholarship in this sub-category is larger than that in
preference matching, it is still rather thin. Moreover, high credibility contributions are
also rare. That said, the three country studies of Strongly Credible or Somewhat Credible
empirical strategies all demonstrate the ability of health decentralization to have a
positive influence on infant mortality rates. The same is not necessarily true for
immunization, although we are forced to draw this conclusion on the basis of two cross-
country studies with less convincing methodologies.
3.2.2 Education
Amongst studies of education, two papers of high quality set the stage for prevalent
optimism. In the first, Galiani et al. (2008) examine decentralization in Argentina by
comparing changes in student test scores in secondary schools that have always been
under provincial control to changes in schools that were under federal control until the
1991 reform. Like many other Latin American countries, Argentina undertook
devolution to provinces as part of a broader structural reform, first devolving
responsibility for pre-schools and primary schools, and then undertaking the same
reform for secondary schools. Provinces now have authority over personnel and
budgeting decisions, while schools are largely responsible for textbook selection and
teaching methods.
Galiani et al. perform a difference in differences estimation while controlling for
provincial per capita, unemployment and fiscal deficits. Using average school test scores
from a sample of students tested in almost 99% of the secondary school universe, they
compare the change in outcomes in those schools that were decentralized to changes in
those schools that were always provincial. Because the impact on scores is unlikely to be
immediate, the authors estimate the impact of exposure to decentralized schools for up to
five years by cohort. Their results show a positive association between decentralization
and Mathematics and Spanish scores - after five years, a 4.9% and 6.9% increase
compared to the mean, respectively. The analysis in the paper does not include any
validation of parallel trends, which is the key identifying assumption of a difference in
differences estimation - however, the authors do use matching techniques to demonstrate
25 of 46
that the results do not change when looking at treatment and control groups with similar
characteristics.
In a comparable paper with a solid methodological design, Faguet and Sánchez
(2008) use changes in enrolment rates in state schools as the measure of student
achievement in order to evaluate the impact of decentralization on service delivery in
Colombia. They analyze the impact of a phased decentralization reform in the country,
which not only left local governments responsible for provision of public services but
also provided them increased fiscal powers to fulfill this responsibility.
Their empirical strategy is deemed Strongly Credible for a number of reasons. First,
the authors, like some of the other papers we review in this theme, overcome the
challenges of measuring decentralization by using three different indicators. This reduces
the possibility that results are driven primarily by choice of measure. Second, they
include relevant controls for resources that may influence school attendance rates
independently, in addition to using many of the standard local government level controls
for wealth and unemployment, limiting the possibility of omitted variable bias. Finally,
in order to address the possibility of reverse causality, i.e. an increase in enrolment rates
causing the increase in local government spending, the authors instrument for
decentralization using t-1 and t-2 lagged local per capita tax revenues. These instruments
should be correlated to the decentralization ratio of own resources to total expenditure,
but are unlikely to influence increases in student attendance directly.
Using both OLS and IV, they find that measures of decentralization have a
significant and positive correlation with changes in student enrolment, with the effects
being larger for smaller municipalities.
The larger Somewhat Credible category fuels further enthusiasm for decentralized
education delivery. In a paper discussed earlier under Health Technical Efficiency,
Habibi et al. (2003) report the empirical relationship between fiscal decentralization and
the ratio of students enrolled in secondary school per 1000 primary students. Using
Argentine data from 1970-1994 in a fixed effects model, the authors find that their
measure of decentralization - own resources to total resources - has a positive and
significant association with their measure of education output.
Freinkman and Plenakanov’s (2009) examination of the impact of fiscal
decentralization on student scores in Russia presents the only comparison of preference
matching and technical efficiency of decentralization that we have come across in this
review. The authors evaluate the statistical relationship between test scores of students
26 of 46
from 73 regions in Russia tested in 2004 and 2005 and fiscal decentralization of a region.
The 1994 reforms passed responsibility of key public services to local governments,
giving them control over 80% of social spending on health and education. The authors
exploit regional variances to estimate the impact of decentralization in a between effects
model, employing many of the conventional controls, but also including one for initial
stock of inputs. Their data on household factors, however, is fairly limited.
Despite this, their findings are rather interesting – they find that the change in
spending on education was marginal with no significant impact of decentralization
variables on computers, pre-school years or student teacher ratios. However, the
relationship between student outcomes, as measured by an average of Language and
Mathematics test scores, and decentralization is consistently positive in all of their
specifications. A 10 percentage point increase in own revenues of municipalities is
significantly correlated with 30 percent of one standard deviation improvement in
secondary school exam scores. Taken together, they propose that their results are
consistent with a technical efficiency argument arising from accountability and local
official incentives, rather than allocative efficiency of increased inputs into the education
production process.
Aslam and Yilmaz (2011) are similarly positive about decentralization, and support
their arguments with analysis conducted on a unique dataset collected from 183
randomly selected villages in 5 purposively chosen districts in Pakistan. Pakistan
embarked on an ambitious decentralization program in 2001, which left local
governments responsible for basic service delivery, although many scholars note that
devolution over fiscal and personnel management was limited.
The authors construct a measure of education service delivery by collecting
retrospective data from villagers on changes in capital improvements, school
maintenance and education services during the period 1995 to 2007. They validate this
data on historical changes through on-site documentation. They then regress a composite
measure of these indicators on a dummy variable for decentralization using a fixed
effects model, to find that provision of education increased dramatically after the
introduction of the decentralization reform. Of some worry, however, are data limitations
due to which the only control they can include is an estimate of village population.
Can we conclude that decentralization enhances technical efficiency in education
based on the above? Inchauste (2009) would disagree. Her examination of the
relationship between changes in education allocations, and children not attending school
27 of 46
and un-enrolled children in Bolivia, shows limited support for devolution. Using data
over a slightly longer period than her analysis of health indicators (see above), her results
for funds made available to local governments through the HIPC initiative are again
mixed. She finds that increases in education transfers were associated with a decrease in
children not attending school in the 1999-2002 period, but an increase in unenrolled
children in 2002-2005. The impact of education spending in both periods, and on other
intermediate education indicators, is not significant.
The lower quality evidence, primarily simple regression analysis on cross-sectional
data, is likewise mixed. Some of the contributions, nonetheless, do present noteworthy
findings. One example is Di Gropello (2002), which shows conflicting results on the
impact of municipality level and school level devolution on student test scores in Chile.
Using an education production function design for testing conducted in 1996, Di
Gropello regresses the outcome of student test scores on a measure of fiscal
decentralization, school level parameters of autonomy and participation, and some
controls commonly found in the education economics literature. Both devolved wage
incentives and training expenditure at the municipal level are associated with higher test
scores, while greater financial autonomy as measured by municipal own funds spent on
education to total funds spent on education is not. At the school level, Di Gropello finds
that coefficients of involvement in financial and pedagogical decision-making are
significant and positively associated with student test scores.
By contrast, Lockheed and Zhao (1993)’s review from the Philippines is decisively
negative. By comparing national, private and municipal or baranguay-run schools and
controlling for socio-economic background, they find no significant difference in
attitudes or achievement in science or mathematics. They argue that this is due to little
actual control and resources being devolved to local schools, presenting the “empty
opportunity of decentralization” in the country. Treisman (2002) is equally pessimistic in
the only cross-country evidence we review in this sub-category. He finds that the
presence of constitutional autonomy and electoral accountability at the local level were
both associated with a higher level of youth illiteracy in data from up to 166 nations. In
line with his findings in the area of health, the negative associations he reports are
stronger for countries with GNP per capita greater than USD 5,000.
In summary, this sub-category is not only larger than the previous ones, but also
significantly more positive about decentralization’s effects, especially when adjusted for
quality of evidence. Although the marginal increase in number of contributions comes
28 of 46
from studies falling in the middle quality distinction, they are almost unanimous in their
support of decentralization’s ability to enhance both the quality and quantity of
education.
3.2.3 School-Based Management (SBM)
As indicated by a number of recent reviews, the literature in this category has made
considerable progress. We provide a basic examination of the most important pieces of
works in this literature below, but also refer readers to Galiani and Perez-Truglia (2011);
Bruns et al (2011); and Barrera-Osorio et al. (2009) for more comprehensive reviews5.
What does ‘gold standard’ evidence tell us about the effectiveness of SBM reforms?
Two recent experimental contributions, and the only ones to be categorized as having
Very Strongly Credible research designs in our review, investigate SBM’s potential in
enhancing student attainment. They yield contradictory findings.
The more optimistic evidence comes from Duflo et al.’s (2007) paper on a
randomized control trial in Western Kenya. The trial tested a number of interventions on
a total of 210 primary schools, one of which involved an SBM component that
empowered school councils to hire and monitor contract teachers. Duflo et al. compare
the SBM groups to their counterparts in the control group, to show that students in the
treatment cell scored 0.18 and 0.24 standard deviations higher in Mathematics and
Language than their non-treated counterparts two years following the intervention.
On the other hand, Glewwe and Maiga (2011) present less optimistic experimental
results. They examine a randomized trial in Madagascar, which involved management
reforms at three levels – district, sub-district and school. In a sample of 30 districts, sub-
districts and schools were randomly sorted into treatment and control groups. Glewwe
and Maiga document some school improvements in the first six months, but by the end
of two years find no discernible impact on aggregated test scores. They conclude not
against the reform per se, arguing instead that results may be driven by the short time
since intervention. Their conclusion is consistent with suggestions in the SBM literature
from the US that reforms may take up to five years to affect student test scores (see
Borman et a. 2003).
5 For good reviews of this literature from developed countries, see Summers and Johnson (1994) and
Borman et al. (2003). See also a related and relatively more rigorous literature on charter and grant-
maintained schools from the US and UK respectively e.g. Abdulkadiroglu et al. (2011) and Clark (2009)
29 of 46
The four studies that rely on the quasi-experimental technique of difference in
differences, on the other hand, are unanimously favorable. The former two we classify
under the Strongly Credible quality distinction, while the latter two are deemed to have
Somewhat Credible identification strategies primarily as a result of challenges they face
with the key parallel trend assumption required in a through DID analysis.
The strongest paper of this type is due to Gertler et al. (2011), who consider the AGE
(Apoyo a la Gestión Escolar) intervention from Mexico. AGE is an SBM reform that
provides training and small grants to parent associations in disadvantaged schools to
invest in infrastructure and materials. Gertler et al. exploit the phased implementation of
the AGE program to achieve identification, comparing schools that adopted AGE earlier
to those that adopted it later.
The key concern is if these two groups differed systematically from each other, and
these differences were correlated with school performance. To address this, the authors
first provide proof that trends in both early and late adopters were parallel. They contend
that because many of the unobservable factors that confound identification tend to be
fixed over time, difference in differences is an appropriate strategy that allows the
elimination of not only pre-treatment differences, but also all other time invariant
differences between the treated and control groups. To further limit chances of
endogeneity, the authors use a vector of relevant time-varying school characteristics.
Their analysis suggests that participation in the AGE program is associated with a
0.6 and 0.4 percentage point reduction in failure and repetition rates, respectively. This
translates to a 4% and 5.4% decrease in these respective indicators. The authors find no
significant association between AGE and intra-year drop-out rates. Through qualitative
research, the paper also suggests that the channel for improvement is the increased
participation of parents in decision-making.
In the second study of this type, Skoufias and Shapiro (2006) also use a difference in
differences method but combine it with a matching technique to examine a different
intervention from Mexico. They consider the PEC (Programa Escuelas de Calidad)
program, another SBM type reform which provides annual grants to disadvantaged
schools to improve education quality. The program gave up to a five-year USD 15k
grant to the 20,000 schools, or 10% of the schooling system, that volunteered to
participate. Like other SBM reforms, school councils participated in the design,
implementation and monitoring of the improvement plans that the grants financed.
30 of 46
The authors use data for approximately 75,000 schools to first conduct a simpler
OLS, and then a much improved difference in differences with matching estimation, to
address concerns of lack of counterfactual data, and self-selection. Their use of
propensity score matching attempts to correct for self-selection by allowing comparison
of PEC and non-PEC schools that are as similar as possible. This similarity is captured
by an index or propensity score that is based on observable characteristics of schools. In
addition, the authors control for a variety of school and municipal characteristics.
Employing this mix of methods, they find that PEC participation is significantly
associated with a 0.24, 0.24 and 0.31 reduction in dropout, failure and repetition rate,
respectively. The impact, it should be noted, is marginal and represents a 6% to 8%
reduction relative to the baseline means. Due to data limitations, the authors do not
provide pre-intervention trend validation.
Paes de Barros and Mendonca (1998)’s study of the three key SBM changes in
Brazil of financial autonomy of schools, head teacher election and establishment of
school councils has similar limitations where lack of pre-implementation trend validation
is concerned. Using one of the first difference in differences methodologies seen in the
field, Paes de Barros and Mendonca examine changes in a series of outputs by using data
from education censuses during the 1981-1993 period. Their methodology relies on
estimating, over this period of time, the change in states that received an innovation
against the change in groups that did not receive an innovation. To address time-varying
characteristics, the authors include a set of controls for per capita income and indicators
of teacher quality.
Their findings suggest that financial autonomy is associated with a significant drop
in repetition rates. Similarly positive trends are evident when the authors examine mean
level lags in grades attended and share of children with lag – the former is negatively and
significantly associated with school financial autonomy and the presence of school
councils, while the latter has a negative and significant correlation with school council
presence only. To perform their analysis, however, Paes de Barros and Mendonca use
state-level averages, which may mask important variation at lower levels of government.
Also supportive is the only investigation from Asia that we review in this sub-
category. In a recent World Bank working paper, Khattri et al (2010) evaluate the impact
of school-based management reforms implemented in 2003 in 23 districts in the
Philippines. The program, sponsored by the World Bank, involved providing training to
principals and parents in designing School Improvement Plans, in addition to direct
31 of 46
funding for the improvements planned. Implementation of the program was in three
phases, with early selection based primarily on a perception of school capability.
Using data from the period 2003 to 2005, Khattri et al. compare the test score
performance of students in schools that implemented the intervention in the first phase to
those that implemented the intervention in a later phase. Due to the nature of the
implementation strategy, however, the authors note chances of bias. These chances are
highlighted by their analysis of pre-intervention trends, which shows that attainment
trends in the treatment and comparison groups were not similar. To mitigate this
concern, Khattri et al. combine their DID strategy with matching. However, they
concede that their analysis may be more useful for noting trends as they “cannot make
any claims of causality” (p. 11).
In comparison to the control group, the treatment group showed a 1.45 percentage
point improvement in overall student attainment. For the subjects of Science, English
and Mathematics, the improvements were 1.45, 1.32 and 1.88 percentage points,
respectively.
In contrast to the above papers, the rest of the Somewhat Credible identification
strategy category has mixed findings. The two cross-country studies on SBM are
negative. Gunnarsson et al. (2009) evaluate the effects of School-Based Management
reforms in 10 Latin American countries. School autonomy and participation, two
indicators that have been popularly used by many other authors working on SBM, are
measured for the various countries using survey data from 1997 and then quantified
using factor analysis. The authors support their OLS analysis by instrumenting for
autonomy and participation using principal attributes and legal structure.
After first noting the variation in autonomy and participation across countries, they
find a negative and significant association between school autonomy and test scores.
They do, however, find a positive relationship of test scores with parental participation.
That said, Barrera Osorio et al. (2009) indicate that Gunnarsson et al.’s instruments are
not necessarily solid. Good instruments are correlated with the endogenous variable, in
this instance autonomy, but have no effect on the dependent variable except through the
endogenous variable. However, of concern is the fact that both principal attributes and
legal structures can plausibly have an independent impact on attainment.
The second cross-country study comes from a recent contribution by Hanushek et al.
(2011) who use data from four waves of PISA test scores to establish the relationship
between student achievement and autonomy in curricular, personnel and budgeting
32 of 46
areas. Their dataset contains test scores and background data on 1 million students from
42 countries, of which 25 are classified as high income nations.
Using a two way fixed effects model and controlling for a variety of family and
school characteristics, Hanushek et al. find the relationship between the parameters of
interest to be negative, albeit heterogeneous across countries based on income levels. A
disaggregated analysis suggests that school autonomy is related to positive outcomes in
developed and high-performing nations, but to negative ones in developing and low-
performing nations. The authors perform a number of robustness checks, including one
to address the concern of reverse causality – they test whether changes in autonomy in
nations are predicted by previous PISA scores, and find they are not.
Hanushek et al.’s study suggests that understanding when SBM can be effective is
critical. King and Ozler (2000)’s older paper on Nicaragua’s reform provides an
interesting answer to this question by arguing that it is de facto and not de jure school
autonomy that improves student performance. Nicaragua’s school autonomy intervention
was implemented in 1991 and allowed Nicaraguan schools to sign contracts with the
Ministry of Education to become autonomous. Autonomous schools were meant to work
through school councils, which had de jure control to hire and fire teachers, manage
school budgets and maintain infrastructure.
King and Ozler use a number of models to estimate the impact of both de jure
autonomy measured by the signing of a contract, and de facto autonomy measured by
factor analysis of teacher’s responses. In their main specification, the authors rely on an
extended production function on cross-sectional data. They attempt to mitigate
challenges arising from both high student attrition in their sample and endogenous
implementation of the reform, the former through estimation of a Heckman selection
model and the latter through the use of a matching technique and analysis of panel data
available for a smaller sample base. Their matched comparison design strategy compares
similar treated and non-treated schools to show that de facto autonomy is associated with
higher test scores in Mathematics and Spanish, but de jure autonomy has no significant
impact. The school fixed effect strategy on the smaller sample yields consistent findings.
Parker (2005) provides more support for the case of the Nicaraguan autonomy
reform, using more nationally representative data that she contends is less prone to
student attrition bias than King and Ozler’s (2000) analysis. To mitigate the self-
selection bias that exists in her analysis, Parker also uses propensity score matching to
compare autonomous and centralized schools. Her results are mixed – after controlling
33 of 46
for the standard components of an education production function, she finds that third
graders in autonomous schools scored significantly higher than their counterparts in
centralized schools in Mathematics. But for sixth graders the effect is negative for
Mathematics, and in neither case are results significant for Spanish.
Also relying on cross-sectional analysis, Jimenez and Sawada (1999) study EDUCO,
perhaps the most celebrated case of SBM, and find no significant difference in test
scores. EDUCO (Educación con Participación de la Comunidad) was first implemented
in 1991 and has served as a model for many of the community-run schools in the Latin
American region. The program established community schools to enhance access in
rural areas in El Salvador following the end of civil war. EDUCO schools are run by
councils consisting of elected community members called Associations for Community
Education (ACE), which have considerable authority in hiring/firing teachers, setting
school curriculum and monitoring school performance.
The authors employ one of the first education production functions in the field to
assess the impact of decentralization on student test scores, albeit for a fairly small
sample of 600 students tested in 1996. To address endogenous program placement,
Jimenez and Sawada use a Heckman two-stage procedure, exploiting the government
priorization formula as an instrument. They find no significant difference in test scores
between traditional and EDUCO students. Given that EDUCO students come from
disadvantaged backgrounds, the authors consider this a positive result. They also find
evidence that student absenteeism is lower in EDUCO schools.
In a more recent paper, Sawada and Ragatz (2005) use propensity score matching on
the same dataset, and still find no impact on student test scores. They do, however, report
evidence of significantly lower teacher absenteeism in EDUCO schools. The authors
propose that lower absenteeism arises as the result of improved community monitoring,
and the authority of councils to hire/ fire teachers. Evidence shows that ACEs of
EDUCO schools use incentives for renewable contracts to motivate this outcome among
teachers. Their finding is supported by other studies that compare absenteeism rates of
permanent and contract teachers, especially in India, to show that the community
monitoring aspect is critical in enhancing outcomes (see for example Ramachandran et
al. 2005; Banerjee and Duflo 2006).
Di Gropello and Marshall (2005) employ a methodology similar to Jimenez and
Sawada (1999) to assess the impact of participating in a PROHECO (Proyecto
Hondureño de Educación Comunitaria) community school in Honduras. PROHECO
34 of 46
schools were first established in 1999 in order to enhance primary school access in rural
areas. Unlike more traditional schools, they are run almost entirely by school councils
which are legal entities that set budgets, maintain school infrastructure and perform key
personnel management functions including hiring, monitoring and paying.
The authors use a Heckman selection correction model to attempt to address
endogeneity in placement of PROHECO schools into areas where parental demand is
low and other community-level problems exist which would otherwise underestimate the
benefits of participation. The two-step Heckit model works by first estimating a probit
equation to predict participation in a PROHECO school based on exogenous factors, and
then by using this probability factor in regressions against the output of drop-out rates to
mitigate self-selection bias. As instruments, Di Gropello and Marshall (2005) use the
presence of potable water and other community services. The authors also control for the
usual education production function factors. Their results point to marginally lower
dropout and repetition rates in SBM schools. Validity of their findings, however, may
have been compromised due to inconsistent data collection as pointed out by the authors.
Moreover, as Gertler et al (2007: p. 20) note, “because services that affect access might
also affect learning, these instruments are not particularly convincing.”
The only study in this category with a Less Credible identification method is one
from Argentina, which nonetheless demonstrates findings of interest. Eskeland and
Filmer (2007) perform a simple OLS regression using an expanded education production
function to investigate the impact of an education devolution reform that left many key
education decisions decentralized to the municipal and school level. Exploiting cross-
sectional data containing test scores of over 24,000 6th and 7th grade students across the
nation, the authors present one of the first attempts to explore the interdependent nature
of autonomy and participation by using an interaction model. Eskeland and Filmer find
that autonomy is significantly associated with student test scores in Mathematics, but not
in Language. They contend that participation has no independent effect on scores, but
that its interaction with autonomy is positive and significant.
In summary, studies with credible identification strategies appear to consistently
support school decentralization’s ability to improve repetition, failure and even drop-out
rates. Results on student test scores, however, are mixed in both higher quality and
medium quality evidence. Interestingly, authors have as a consequence tried to consider
when SBM may be successful. This effort has yielded diverse results suggesting efficacy
particularly in developed countries, or in schools with de facto autonomy, or even in the
35 of 46
presence of both autonomy and participation together. The scholarship on this latter
aspect, however, is too small to draw firmer conclusions.
Table 4: Selected Empirical Evidence of Technical Efficiency
Source: Ci ted articles
No. Author (Date) Country of
Study Date
Implemented Programme
Description Method of
Analysis Sample Measure/s of Decentralization Results Identification
Strategy No.
Health
1Uchimura a nd
Jutting (2009)
China 1970s Devolved expenditure but
growingly centralized revenue
authority, respo nsibi lity for
delivery of health services, no
poli tical de volution
OLS using a fixed
effects model
26 provinces over period 1995 -
2001
▪ LG expenditure / LG own revenue
▪ LG expenditure / provincial expenditure
▪ Higher LG expenditure is associated with lower IMR Strongly Credible 1
2Asfaw et al (2007) India 1980s Rule-based fiscal transfers to
LG, responsibility for public
services, political devolution
OLS us ing
between, fixed and
random effects
models
14 States over period 1990 - 1997 ▪ Index determined by factor analysis - share
of LG expenditure in state, total LG
expenditure per person, share of LG own
revenue in LG expenditure
▪ D sig nifica ntly associa ted with lower IMR in fixed and
random e ffects mode l but not in the betwee n effects
model
Somewhat Credible 2
3Habibi et al (2003) Argentina 1991 Decentralized financing, staff
management and budgeting
to LG.
OLS using a fixed
effects model, GLS
23 of 23 provinces over 1970 to
1994
▪ Share of resources in provincial control to
total re sources
▪ Share of locally generated resources to
loca lly contro lled reso urces
▪ Share of locally generated to locally controlled
resources is associ ated with lower infant mortali ty rates
Somewhat Credible 3
4Khaleghian (2004) Cross-country V arious Various OLS using between
effects model with
time fixed effects
140 Low and Middle Income
countries over period 1980 - 1997
▪ Presence of taxing, spending or regulatory
authority by LGs
▪ D associated with higher coverage in lower income
countries, and lower coverage in higher income countries
- L shape suggested
Somewhat Credible 4
5Inchauste (2009) Bolivia 1994 Increase in transfers to LGs
followi ng HIPC i nitia tive -
2000 onwards
OLS using a
random effects
model
300 municipalities over period
1999 - 2002
▪ Change in spending in health ▪ Decrease in share of unattended illness during period
▪ No significant change in cases of respiratory diseases
Somewhat Credible 5
6Robalino et al
(2001)
Cross-country Various Various OLS using a fixed
effects model
Low and High Income countries
over period 1970 - 1995
▪ Expenditures managed by LG / managed by
CG
▪ 10% increase in share of locally managed expenditures
is associated with 3.6% reduction in IMR for USD2k per
capita country
▪ Benefits U shaped - higher for high and low income
countries and low for middle income
Somewhat Credible 6
7Treisman (2002) Cross-country Various Various OLS Up to 166 countries with cross-
secti onal data c ollected from mid-
90s
▪ Presence of LG exclusive authority on any
one item
▪ Some other measures of fiscal and political
decentralization also used
▪ No significant association of sub-national autonomy
with either ind icator
▪ Electoral accountability associated with greater access
to med icatio n
Less Credible 7
37 of 46
Education
8Faguet and
Sa nchez (2 008)
Co lombi a Pha sed b egi nning
in 1970s - key
reform in 1991
Increase in devolved funds to
LG, greater responsibility for
public services, political
devolution
OLS and 2SLS 90% of municipalities universe
over period 1994 to 2004
▪ LG o wn reve nue sour ces / LG exp endit ure
▪ Binary factor of Municipal Certification
▪ Share of transfers to education expenditure
▪ D positively associated with higher enrolment in public
school
Strongly Credible 8
9Galiani et al (2008) Argentina 1991 Decentralized financing, staff
management and budgeting
to LG. Schoo ls choose
textbooks and teaching
methods.
Differ ence in
differences
Almost all secondary schools over
period 1994 to 1999
▪ Actual transfer from province to LG ▪ D associated with higher Math and Spanish scores Strongly Credible 9
10 Asla m and Yilmaz
(2011)
Pakistan 2001 Limited financial devolution
but implementation of rule-
based transfers,
responsibility for public
servic es, politi cal devolution
OLS using a fixed
effects model
183 villages from 5 districts out of
120+. D ata collected for 1 995 to
2007
▪ Binary measure of before and after D
implementation
▪ Magnitude of education services increases after
introducti on of D
Somewhat Credible 10
11 Freinkman and
Plekhanov (2009)
Russ ia Phas ed be ginni ng
in 1994
Increased fis cal powers with
rule-based transfers,
responsibility for public
services
OLS using a
between effects
model
Secondary school results from 73
out of 83 regions, tested in 2004
and 2005
▪ Share of LG education expenditure financed
by own revenue
▪ Type of decentralization arrangement
▪ D associated with higher test scores Somewhat Credible 11
12 Habibi et al (2003) Argentina 1991 Decentralized financing, staff
management and budgeting
to LG.
OLS using a fixed
effects model, GLS
23 of 23 provinces over 1970 to
1994
▪ Share of resources in provincial control to
total re sources
▪ Share of locally generated resources to
locally controlled resources
▪ Share o f locally generate d to locally co ntrolled
resources is associated with higher enrolment
Somewhat Credible 12
13 Inchauste (2009) Bolivia 1994 Increase in transfers to LGs
following HIPC initiative -
2000 onwards
OLS using a
random effects
model
300 municipalities over period
1999 - 2005
▪ Change in transfers for education
▪ Change in spending in education
▪ Increase in share of unenrolled children associated with
increase in transfers
▪ No significant impact of change in spending
Somewhat Credible 13
14 Di Gropello (2002) Chile Early 1980s Some increase in devolved
funds to LG, grea ter
responsibility for public
services. In 1990s, greater
pedagogical devolution to
schools
OLS 50 municipalities (out of 355) -
Student tests conducted in 1996
▪ LG own funds / Total funds spent on
education
▪ School level parame ters on partic ipation,
autonomy
▪ Muni cipa l financ ial a utonomy no t sig nific ant
▪ Municipal training spend and wage incentives positively
associated with test scores
▪ School involvement in financial decision-making
positively associated with test scores
▪ School pedagogical and curricular autonomy positively
associated with test scores
Less Credible 14
15 Jimenez and
Paqueo (1996)
Philippines 1987 School councils raise funds,
while LGs earmark taxes for
supplemental school
compensation
OLS 600 elementary schools over
period 1982 to 1983
▪ Proportion of school revenues from local
sources
▪ Lo cally fi nance d scho ols ha ve low er fina ncia l
expenditure, indicating cost efficiency
Less Credible 15
16 Lockheed and
Zhao (1993)
Philippines 1987 School councils raise funds,
while LGs earmark taxes for
supplemental school
compensation
OLS comparison of
private, national
and local
government
schools/ HLM
8k 9th grade students in 214
schools
▪ Lo cal go vernme nt scho ol sta tus ▪ No significant impact Less Credible 16
17 Treisman (2002) Cross-country Various Various OLS Up to 166 countries with cross-
sectional data collected from mid-
90s
▪ Presence of LG exclusive authority on any
one item
▪ Some other measures of fiscal and political
decentralization also used
▪ Negative and significant relationships between
measures of decentralization and illiteracy
Less Credible 17
38 of 46
School-based Management (SBM)
18 Duflo et al (2007) Kenya 2005 Randomized trial which gave
school councils money and
autonomy to hi re extra
teachers and monito r their
performance
OLS comparison of
treatment and
control groups.
Alternate
specification uses
covariates
21k students from 210 schools ▪ Dummy variable indicating treatment group SBM associated with
▪ Increase in Mathematics scores of 0.24 standard
deviations
▪ Increase in Language scores of 0.18 standard
deviations
Very Strongly
Credible
18
19 Glewwe and Maiga
(2011)
Madagascar 2005 Randomized trial in which
materia ls, training and
greater accountability is given
to three levels of districts, sub-
districts and schools
OLS comparison of
treatment and
control groups.
20k students from 30 districts over
period 2006 to 2007
▪ Dummy variable indicating treatment group ▪ No significant association with test scores Very Strongly
Credible
19
20 Gertler et al (2011) Mexico 1996 Small grants to parent
councils and parental training
targeted at disadvantaged
areas - AGE
Di fference in
differe nces
30,000 students from 6,000
schools over 1997 - 2001
▪ Dummy variable indicating whether school
received AGE intervention
SBM associated with
▪ Reduction in failure rates by 4%
▪ Reduction in repetition rates by 5.4%
▪ No imp act on drop -out rates
Strongly Credible 20
21 Hanushek et al
(2011)
Cross -country Various Various OLS with country
fixed effec ts
1mn students from 42 countries - 4
waves of PISA from 2000 to 2009
▪ Autonomy over curriculum and pedagogy
▪ Autonomy over personnel management
▪ Autonomy over budgeting decision
▪ Overall negative association between autonomy and
scores
Somewhat Credible 21
22 Skoufias and
Shapiro (2006)
Mexico 2001 Annual grants of up to USD
15k given to schools/SMCs to
improve education quality
OLS and
Difference-in-
differe nces with
matching
75000 schools over period 2001 -
2004
▪ School received PEC grant in all three
years
▪ School received PEC grant in any one year
SBM associated with
▪ 0.24 lower dropout rates
▪ 0.24 lower failure rates
▪ 0.31 lower rep etition rates
Strongly Credible 22
23 Di Gropello and
Marshall (2005)
Honduras 1999 Community schools where
SMCs can hire/ fire tea chers,
mana ge sc hool f unds and
maintain i nfrastructure
OLS with heckma n
correction model
200 rural schools tested in 2002
and 2003
▪ Probit model predicting participation in
PROHECO school
SBM associated with
▪ Higher science scores but with no change in Math or
Language test scores
▪ Marginally lower dropo ut rates
Somewhat Credible 23
24 Gunnar sson e t al
(2009)
Cross-country Various Various OLS and 2SLS 17k students from 10 Latam
countries - 1997 survey data
▪ Autonomy of school in decisions regarding
hiring, budg et allocation, curriculum desi gn,
etc.
▪ Participation of parents and communities in
school
▪ Autonomy associ ated with lower test sco res
▪ Participation associated with higher test scores
Somewhat Credible 24
25 Jimenez and
Sawada (1999)
El Salvador 1991 Community schools where
SMCs can hire/ fire tea chers,
mana ge sc hool f unds and
maintain i nfrastructure
OLS with heckma n
correction model
605 3rd grade students from 162
municipalities - data from 1996
▪ Binary indi cating whether it is an EDUC O
school or not
▪ No ass ociati on with Math or E nglis h test s cores
▪ Students in EDUCO schools have lower absenteeism
Somewhat Credible 25
39 of 46
26 Khattri et al (2010) Philippines 2003 Training and direct funding for
school improvement
Difference in
differe nces with
matching
5k schools from 23 districts over
2003 to 2005
▪ Dummy variable indicating whether school
recei ved intervention i n first year
SBM associated with
▪ 1.45 percentage points overall improvement
▪ 1.82 percentage points improvement in Science
▪ 1.32 percentage points improvement in English
▪ 1.88 percentage points improvement in Mathematics
Somewhat Credible 26
27 King and Ozler
(2000)
Nicaragua 1991 Autonomous schools with
SMCs that can hire/ fi re
teachers, manage school
funds and ma intai n
infrastructure
Matching +
validation using
fixed effects and IV
3000 students from primary and
secondary schools over period
1995 - 1997
▪ De jure autonomy - binary variable of
whether the school is autonomous by law or
not
▪ De facto autonomy - actual school autonomy
in various areas
▪ No impact of de jure Autonomy
▪ Posi tive associa tion of de facto Autonomy with Math
and Sp anish, no asso cia tion wi th Lang uage
Somewhat Credible 27
28 Paes de Barros
and Mendonca
(1998)
Brazil 1982 SBM with three key
inno vatio ns:
▪ Fi nancia l autono my of
schools
▪ Ability to elect principals
▪ Presence of school councils
Difference in
differences - state-
level
18 states over period 1981 - 1993 ▪ F inanc ial a utonomy o f schoo ls
▪ Ability to elect principals
▪ Pre sence of scho ol co uncils
▪ Lower repetition rates associated with financial
autonomy
▪ Lower mean grade level lag associated with financial
autonomy and school councils
▪ Lower proportion of students with lag associated with
school council presence.
▪ Proportion of students with lag negatively associated
with principal election
Somewhat Credible 28
29 Parker (2005) Nicaragua 1991 Autonomous schools with
SMCs that can hire/ fi re
teachers, manage school
funds and ma intai n
infrastructure
Matching 1000 3rd and 6th grade students -
tested in 2002
School autonomy in various areas SBM associated with
▪ Higher third grade Math scores
▪ Lower sixth grade Math scores
▪ No association with Spanish scores
Somewhat Credible 29
30 Sawada and
Ragatz (2005)
El Salvador 1991 Co mmunity schools where
SMCs c an hire/ fire teachers,
mana ge sc hool fund s and
maintai n infrastructure
Matching 605 3rd grade students from 162
municipalities - data from 1996
▪ Bi nary ind icat ing whet her it i s an ED UCO
school or not
▪ No association with scores
▪ Lower teacher absenteeism in EDUCO schools
Somewhat Credible 30
31 Eskeland and
Filmer (2007)
Argentina 1978 Decentralized financing, staff
management and budgeting
to LG. Schools choose
textbooks and teaching
methods .
OLS with province
fixed effects
24000 6th and 7th grade students
from urban schools
▪ Autonomy of school in various decisions
▪ Participation of parents
SBM associated with
▪ Higher Math but no change in Language scores
▪ Effect is stronger for poorer households
Less Credible 31
Notes
PM: Allocative Efficiency
TE: Technical E fficiency
D: Decentral ization
LG: Local Government
4. Conclusions
How do the conclusions of our review compare with those of other broad surveys
that, at various points over the past 25 years, have assessed the state of the field,
attempting to reach concrete conclusions about decentralization’s empirical effects? One
of the broadest and most widely cited international surveys is Rondinelli, Cheema and
Nellis (1983), who argue that reformers hopes for decentralization usually outrun the
reality. Serious administrative problems bedeviled implementation in most of the
developing countries they survey. The few comprehensive evaluations of the costs and
benefits of decentralization that had been conducted then indicated limited success in
some countries and failure in others.
A decade and a half later, surveys by Manor (1999), Piriou-Sall (1998), and Smoke
(2001) are only somewhat more positive, with many caveats about weakness in the
evidence favoring reform. Manor notes that the evidence, though extensive, is
nevertheless incomplete, while Smoke finds the evidence mixed and anecdotal, and asks
whether there is empirical justification for pursuing the reform at all. Litvack et al.
(1999) have a darker view: “It is not much of an exaggeration to say that one can prove,
or, disprove, almost any proposition about decentralization by throwing together some
set of cases or data” (p.30). Treisman’s (2007) more recent review of the literature is
bleakest of all. He finds results on the effects of decentralization mixed at best, and for
the most part weak and incomplete. On the consequences of decentralization, he
concludes “Almost nothing that is robust or general has emerged” (p.268).
Our own survey of the scholarship at first glance corroborates such conclusions. The
evidence for decentralization’s effects do appear to be weak, incomplete and at many
times inconclusive. But when we organize the empirical literature first by theme and
then – crucially – by quality of evidence provided, we are able to identify patterns of
empirical results that previous surveys – including our own6 – have missed. Admittedly,
these patterns are not conclusive across all areas of interest. But in many instances they
show that reforms can have clear, positive consequences – in some cases remarkably so,
as we have attempted to outline above. This is very different from the general
6 See for example Faguet 2004, Faguet 2008, and Faguet and Sánchez 2008.
41 of 46
indeterminacy that previous surveys find, and particularly important to the extent that
evidence informs real world policy-making.
Still, many important questions remain unanswered. The variation in research
findings in even the higher quality literature, for instance, implies that decentralization
may work well in one context but not in another. Thus, one critical avenue for further
research is the factors or prerequisites that enable improved outcomes after the
intervention is made. A handful of studies from our review attempt to shed light on this
aspect – Asfaw et al. (2007), for instance, contend that the presence of political rights
enhances the consequences of decentralized health delivery, while Duflo et al. (2007)
demonstrate that training and capacity building prior to implementation can deliver even
better student outcomes. But a lack of focus on this crucial facet7 means that to date the
conditions that facilitate enhanced service delivery post reform have remained elusive.
A related area where the literature appears wanting is the channel through which
improved service provision is achieved. In other words, when decentralization works,
how exactly is it that it works? While there are often suggestions of better monitoring or
improved incentives in the favorable evidence base, few studies provide systematic
evidence to substantiate these claims. This deficiency restricts the policy implications
scholarship can have for developing countries trying to implement, or even improve,
their own reform.
Finally, it is difficult to step away from our review without reiterating the
methodological challenges faced by empiricists in this arena. Undoubtedly, scarce data,
big bang implementation and all-encompassing reform packages make it difficult to
disentangle the causal effects of decentralized service delivery. The weaknesses of
evidence pointed out above are not shortcomings of logic, less still of skill on the part of
researchers. Rather, they are direct consequences of a combination of the data
limitations that we all work under, plus the complexity of the questions we seek to
answer.
The last decade has seen great improvements in identification as better data has
become available, and more recently years have brought us the first randomized trials in
the field. Understandably, implementing RCTs in the broader education and health
decentralization reform is a much more challenging task. Yet there are other possibilities
7 For good recent exceptions focusing on enablers, see Enikolopov and Zhuravskaya (2007) and Loayza et
al. (2011)
42 of 46
that can add to the credibility of the research base. Early involvement of researchers
together with cooperation with government agencies, for example, can allow for quasi-
randomized design. And the use of panel data can help mitigate many of the econometric
challenges associated with cross-sectional work. As this body of work grows, it will
become possible to draw firmer conclusions on the effects of decentralization on service
provision.
43 of 46
Bibliography
Abdulkadiroglu, A., J. Angrist, S. Dynarski, T.J. Kane, and P. Pathak. 2009.
“Accountability and Flexibility in Public Schools: Evidence from Boston's Charters
and Pilots”. Cambridge, MA: National Bureau of Economic Research.
Ahmad, E., Brosio, G., and V. Tanzi. 2007. “Local service provision in selected OECD
countries: do decentralized operations work better?” Paper presented at the
International Conference on Land Policies and Fiscal Decentralization, Cambridge,
Massachusetts. June 4–5.
Ahmad, Ehtisham and Giorgio Brosio. 2009. Does decentralization enhance service
delivery and poverty reduction? Cheltenham, UK ; Northampton, MA: Edward
Elgar.
Akin, J., P. Hutchinson and K. Strumpf. 2005. "Decentralisation and government
provision of public goods: The public health sector in Uganda." Journal of
Development Studies 41(8):1417-1443.
Arze del Granado, FJA, J Martinez-Vazquez and R McNab. 2005. "Fiscal
decentralization and the functional composition of public expenditures."
International Studies Program Working Paper Series, at AYSPS, GSU.
Asfaw, A., K. Frohberg, K. S. James and J. Jutting. 2007. "Fiscal decentralization and
infant mortality: empirical evidence from rural India." Journal of Developing Areas
41(1):17.
Aslam, G. and S. Yilmaz. 2011. “Impact of Decentralization Reforms in Pakistan on
Service Delivery— An Empirical Study.” Public Administration and Development.
31: 159–171.
Angrist, J. and J. Pischke. 2010. “The Credibility Revolution in Empirical Economics:
How Better Research Design is Taking the Con out of Econometrics”. NBER
Working Paper 15794. Cambridge, MA: National Bureau of Economic Research.
Bardhan, P. and D. Mookherjee. 1998. “Expenditure Decentralization and the Delivery
of Public Services in Developing Countries.” Working Paper No. C98-104.
Berkeley: Center for International and Development Economics Research.
Barrera-Osorio, F., H. A. Patrinos and T. Fasih. 2009. Decentralized Decision-Making in
Schools. Washington D.C.: World Bank.
Besley, T. and S. Coate. 2003. "Centralized versus decentralized provision of local
public goods: a political economy approach." Journal of public economics
87(12):2611-2637.
Borman, G. D., G. M. Hewes, L. T. Overman and S. Brown. 2003. "Comprehensive
school reform and achievement: A meta-analysis." Review of Educational Research
73(2):125.
Bruns, B., D. Filmer and H. Patrinos. 2011. Making schools work: New evidence on
accountability reforms. Washington, DC: The World Bank
Cheema, G. S. and D. A. Rondinelli. (eds.). 1983. Decentralization and Development:
Policy Implementation in Developing Countries. Beverly Hills: Sage Publications.
Clark, D. 2009. “The performance and competitive effects of school autonomy”. Journal
of Political Economy 117 (4):745-783.
Di Gropello, E. 2002. An assessment of the impact of decentralization on the quality of
education in Chile. Washington D.C.: World Bank
Di Gropello, E. and J. Marshall. 2005. "Teacher effort and schooling outcomes in rural
Honduras." In Incentives to Improve Teaching, ed. E. Vegas. Washington D.C.:
World Bank.
44 of 46
Duflo, E., P. Dupas, and M. Kremer. 2007. “Peer effects, pupil-teacher ratios, and
teacher incentives: Evidence from a randomized evaluation in Kenya”. Unpublished
manuscript.
Ebel, Robert D. and Serdar Yilmaz. 2002. On the measurement and impact of fiscal
decentralization. Washington, D.C.: World Bank Institute, Economic Policy and
Poverty Reduction Division.
Enikolopov, R and E Zhuravskaya. 2007. "Decentralization and political institutions."
Journal of public economics 91(11-12):2261-2290.
Eskeland, G and D. Filmer. 2007. "Autonomy, Participation and Learning: Findings
from Argentine Schools, and Implications for Decentralization" Education
Economics Vol. 15, No. 1, March pp. 103-127.
Faguet, JP. 2012. Decentralization and Popular Democracy: Governance from Below in
Bolivia. Ann Arbor: University of Michigan Press.
Faguet, JP. 2008. “Decentralization’s Effects on Public Investment: Evidence and Policy
Lessons from Bolivia and Colombia.” Journal of Development Studies, 44: 1100–
1121.
Faguet, J.P. 2004. “Does Decentralization Increase Responsiveness to Local Needs?
Evidence from Bolivia.” Journal of Public Economics, 88(4): 867-894.
Faguet, J.P. and F. Sánchez. 2008. “Decentralization’s Effects on Educational Outcomes
in Bolivia and Colombia.” World Development, 36(7): 1294–1316.
Freinkman, L. and A. Plekhanov. 2009. "Fiscal decentralization in rentier regions:
Evidence from Russia." World Development 37(2):503-512.
Galiani, S., P. Gertler and E. Schargrodsky. 2008. "School decentralization: Helping the
good get better, but leaving the poor behind." Journal of public economics 92(10-
11):2106-2120.
Galiani, S. and Perez Truglia. 2011. “School Management in Developing Countries”.
Unpublished manuscript.
Gertler, P., M. Rubio-Codina, and H. Patrinos. 2011. “Empowering Parents to Improve
Education: Evidence from Rural Mexico.” Journal of Development Studies. In
Press.
Gertler, P., H. Patrinos and M. Rubio-Codina. 2007. "Methodological Issues in the
Evaluation of School-Based Management Reforms." Unpublished manuscript,
World Bank, Washington, DC.
Glewwe, P., and E. Maïga. 2011. “The Impacts of School Management Reforms in
Madagascar: Do the Impacts Vary by Teacher Type?” Available at
http://www.povertyactionlab.org/publication/impacts-school-management-reforms-
madagascar-do-impacts-vary-teacher-type
Gunnarsson, V, PF Orazem, MA Sánchez and A Verdisco. 2009. "Does Local School
Control Raise Student Outcomes? Evidence on the Roles of School Autonomy and
Parental Participation." Economic Development and Cultural Change 58(1):25-52.
Habibi, N., C. Huang, D. Miranda, V. Murillo, G. Ranis, M. Sarkar and F. Stewart. 2003.
“Decentralization and Human Development in Argentina”. Journal of Human
Development. Vol. 4, Iss. 1.
Hanushek, E., S. Link and L. Woessmann. 2011. “Does School Autonomy Make Sense
Everywhere? Panel Estimates from PISA.” NBER Working Paper 7084.
Cambridge, MA: National Bureau of Economic Research.
Hasnain, Z. 2008. "Devolution, accountability, and service delivery: some insights from
Pakistan." WP 4610.
45 of 46
Inchauste, G. 2009. "Decentralization in Bolivia: has it made a difference?" In Does
Decentralization Enhance Service Delivery and Poverty Reduction?, ed. E. and
Brosio Ahmad, G. Chelthenham: Edward Edgar Publishing.
Jimenez, E and V Paqueo. 1996. "Do local contributions affect the efficiency of public
primary schools?" Economics of Education Review 15(4):377-386.
Jimenez, E. and Y. Sawada. 1999. "Do community-managed schools work? An
evaluation of El Salvador's EDUCO program." The World Bank economic review
13(3):415.
Khaleghian, P. 2004. "Decentralization and public services: the case of immunization."
Social Science & Medicine 59(1):163-183.
Khattri, N., C. Ling, and S. Jha. 2010. “The effects of school-based management in the
Philippines: An initial assessment using administrative data.” Policy Research
Working Paper 5248. Washington, DC: World Bank.
King, E and B Ozler. 2000. "What’s decentralization got to do with learning?
Endogenous school quality and student performance in Nicaragua." Development
Research Group, World Bank, Washington, DC.
Litvack, J., Junaid Ahmad and Richard Miller Bird. 1998. Rethinking decentralization in
developing countries. Washington, D.C.: World Bank.
Litvack, J., J. Seddon, J. Ahmad and Institute World Bank. 1999. Decentralization
briefing notes. Washington D.C.: The World Bank.
Loayza, N., J. Rigolini, V. Jamele and O. Calvo-Gonzalez. 2011. “More than you can
handle: decentralization and spending ability of Peruvian municipalities.” Policy
Research Working Paper No 5763. Washington, D.C.: World Bank.
Lockheed, M. E. and Q. Zhao. 1993. "The empty opportunity: Local control and
secondary school achievement in the Philippines." International Journal of
Educational Development 13(1):45-62.
Manor, J. 1999. The Political Economy of Democratic Decentralization. Washington,
DC: The World Bank
Oates, W. 1972. Fiscal Federalism. New York: Harcourt Brace.
Paes de Barros, R. and R. Mendonca. 1998. "The impact of three institutional
innovations in Brazilian education." Organization matters: agency problems in
health and education in Latin America:75.
Parker, C. 2005. “Teacher Incentives and Student Achievement in Nicaraguan”. In
Incentives to Improve Teaching, edited by E. Vegas. Washington D.C.: World Bank.
Piriou-Sall, S. 1998. “Decentralization and Rural Development: A Review of Evidence.”
Washington, DC: Manuscript.
Robalino, David A., Oscar F. Picazo and A. Voetberg. 2001. Does fiscal decentralization
improve health outcomes? : evidence from a cross-country analysis. Washington,
D.C.: World Bank.
Rondinelli, D. A., G. S. Cheema and J. Nellis. 1983. “Decentralization in Developing
Countries: A Review of Recent Experience.” World Bank Staff Working Paper
No.581. Washington, DC: World Bank.
Rubinfeld, D. 1987. “The Economics of the Local Public Sector.” In Auerbach, A. and
M. Feldstein (eds.), Handbook of Public Economics. Oxford: North-Holland.
Santibañez, L. 2006. "School-Based Management Effects on Educational Outcomes: A
Literature Review and Assessment of the Evidence Base." The World Bank,
Washington, DC (processed).
Sawada, Y. and A. Ragatz. 2005. "Decentralization of education, teacher behavior, and
outcomes." In Incentives to Improve Teaching, ed. E. Vegas. Washington D.C.:
World Bank.
46 of 46
Schwartz, B. J., D. K. Guilkey, R. Racelis and Measure Evaluation University of North
Carolina at Chapel Hill. 2002. Decentralization, allocative efficiency and health
service outcomes in the Philippines: MEASURE Evaluation, Carolina Population
Center, University of North Carolina at Chapel Hill.
Shah, A., T. Thompson, and H.F. Zou. 2004. “The impact of decentralization on service
delivery, corruption, fiscal management and growth in developing and emerging
market economies: A synthesis of empirical evidence.” CESifo DICE Report,
1/2004: 10-14.
Skoufias, E. and J. Shapiro. 2006. "Evaluating the impact of Mexico's quality schools
program: the pitfalls of using nonexperimental data."
Skoufias, E., A. Narayan, K. Kaiser, and B. Dasgupta. 2011. “Electoral Accountability,
Fiscal Decentralization and Service Delivery in Indonesia”. Policy and Research
Series Paper 5614. Washington DC: World Bank.
Smoke, P. 2001. “Fiscal Decentralization in Developing Countries: A Review of Current
Concepts and Practice.” Democracy, Governance and Human Rights Programme
Paper No. 2. Geneva: UNRISD.
Summers, A.A., and A.W. Johnson. 1994. “A review of the evidence on the effects of
school-based management plans”.
Treisman, D. 2002. “Decentralization and the Quality of Government.” Unpublished
paper, Department of Political Science, UCLA.
Treisman, D. 2007. The Architecture of Government: Rethinking Political
Decentralization. New York: Cambridge University Press.
Uchimura, H and JP Jütting. 2009. "Fiscal Decentralization, Chinese Style: Good for
Health Outcomes?" World Development 37(12):1926-1934.
Woessmann, L. 2003. "Schooling resources, educational institutions and student
performance: the international evidence." Oxford Bulletin of Economics and
Statistics 65(2):117-170.