DOI: 10.1111/j.1752-5209.2009.0028.x
Delivering Public Services in the Developing World:
Frontiers of Research
Daniel Rogger1
University College London and Institute for Fiscal Studies
Abstract: This essay presents a view of the frontiers of research on public
service delivery in the developing world, based on a series of interviews with
researchers and practitioners actively working in this eld. It recognizes the
lasting contribution of the theoretical framework laid down by the World
Development Report 2004 that emphasized accountability, and the randomized
evaluations that have taken place to test and develop this theory. Research
on other questions, such as those relating to the analysis of politics and the
structure and organization of government, is at an earlier stage, and is likely to
need a more structural approach. There are many questions still to be answered
in this eld.
Introduction
Public expenditure in the developing world is
often ineective and inecient. Projects and
programs are implemented over long periods to
a low quality, and subsequent welfare impacts
are meagre. As World Bank (2004) observes,`too
often, services fail poor people - in access, in
quantity, in quality.'
Even when governments in the developing
world are able to muster resources and direct
these toward public goods provision, welfare im-
pacts are not ensured. As World Bank (2008, 2)
states:
`There is a temptation to view the
relationships between welfare out-
comes and [public expenditure] sim-
plistically: if more money is spent on
basic services, welfare outcomes will
improve. However, this view
ies in
the face of the empirical fact that
there is a weak correlation between
spending and outcomes.'
The `public service delivery' research agenda
aims to understand the sources of this discrep-
ancy. It analyzes the mechanisms required to
eectively transform resources into public goods.
For many years, the economics profession fo-
cussed on corruption and the weak state of the
private sector as the core constraints to the ade-
quate provision of public goods.2 More recently,
an increasingly subtle approach has been taken
to this research that investigates the broader in-
stitutional context of delivering public services
in the developing world.
This essay aims to sketch the frontiers of
that research agenda. It integrates messages
from a series of interviews with researchers ac-
tively working in this eld.3 A list of those
interviewed can be found at the end of the essay,
and includes a range of academics, development
practitioners and World Bank researchers. It is
not a literature review, but rather a collection
of thoughts by leading experts in service deliv-
ery. A more thorough treatment of the existing
literature is given in Rogger (2009).
Building a coherent theory
The `public service delivery chain' is made up
of a plethora of institutions and interactions.
1d.rogger@ucl.ac.uk. I am grateful to the editor and an anonymous referee for useful comments.
2Rose-Ackerman (1999) is a classic example of this literature. Svensson (2005) and Pande (2008) provide an
update on more recent work in corruption.
3Most of the interviews took place face-to-face in December 2008 in either London or Washington.
Oxonomics 4 (2009). c
2009 The Author. Journal compilation c
2009 The Oxonomics Society
Published by Wiley-Blackwell
19

Daniel Rogger Delivering Public Services in the Developing World
From bargaining over public good allocations
in Parliament to implementing public projects
at the local level, each step in delivering pub-
lic goods is a link in the public service delivery
chain. The framework for understanding this
chain laid out in the World Development Report
2004: Making Services Work for Poor People
(henceforth WDR) has provided the foundation
for much of the research undertaken since.
At its core is the notion that `successful ser-
vices for poor people emerge from institutional
relationships in which the actors are accountable
to each other' (World Bank, 2004, 46).4 In the
WDR, accountability is dened as a relation-
ship among actors that has ve features: dele-
gation, nance, performance, information about
performance and enforceability. For example,
a government may delegate primary healthcare
to a private provider, provide nance, and then
track performance, threatening a punishment if
the contract is broken (information about per-
formance and enforceability).
The relationships underlying this statement
have typically been sorted into a `long route'
of accountability (working through the political
process) and a `short route' (the client-provider
relationship). If a public service fails, a citizen-
client can lobby her politicians for change, or if
there is choice, merely choose a better provider.
These `key relationships of power' are displayed
in the World Bank's `accountability triangle'
(Figure 1).
The WDR describes how components of each
route may break down and thus hinder service
delivery.5 For example, politicians may not lis-
ten to their constituents, and there may be lit-
tle choice in provider of public services. The
WDR drew from existing evidence and the World
Bank's own experiences to provide suggestive ev-
idence on what might make a dierence. Much
of the research since then has focussed on testing
the hypotheses laid out in the report.
Advancing theory through testing
More progress has been made on testing some
aspects of the triangle than others. For exam-
ple, client power is increasingly well understood.
Simply providing public goods may not be suf-
cient to induce take-up, even when the client
is given a choice of providers. Information cer-
tainly helps, but is not always decisive (see, for
example, the work by Svensson (2005)). Agents
respond to quality (although this relationship
needs to be better understood).
Pay for performance works in many settings,
but not in all. Mookherjee (1997) lays out the
conditions under which pay for performance
schemes are welfare improving (reducing). He
nds that `institutional parameters', such as
the precise range of tools available for provid-
ing incentives, the extent of discretion available
to bureaucrats and the relevant dimensions of
bureaucratic performance, determine the eec-
tiveness of a pay-for-performance scheme. This
nicely summarizes the now large empirical lit-
erature on this subject. A typical example is
Kingdon and Teal (2007), which nds private
schools in India benet from linking pay to per-
formance, whilst public schools do not.
In terms of relationships to the state, un-
derstanding of the compact (the relationship
between providers and the state) is far more
developed than that of voice (the relationship
between citizens and politicians). This is true
throughout much of the higher echelons of gov-
ernment. Little is known about the objectives
of public sector principals in general.6
Similarly, some sectors are better understood
than others (education and health versus water
and sanitation for example). In education, much
is now understood about teacher incentives and
school-based management (see the work by Mu-
ralidharan and Sundararaman (2008), who em-
phasize the role of improved incentive schemes
for teachers over grants for student-level inputs
and the addition of support teachers). The con-
4A number of interviewees questioned the WDRs emphasis on accountability (due to demand side failures, questions
of feasibility and so on). This was not the dominant consensus, but it is important to keep in mind the potentially
restrictive nature of the approach.
5Thus, it provides (testable) predictions as to the mechanisms that drive service delivery. For example, the model
predicts that improvements in citizen information about politician behaviour should, conditional on the compact,
improve the services delivered to the citizenry. Similarly, it predicts that consumer choice will unconditionally improve
services delivered. For a critique of this model, see Rogger (2009).
6This impacts on our ability to dene appropriate principal-agent problems and conceptualize eciency.
20 Oxonomics 4 (2009). c
2009 The Author. Journal compilation c
2009 The Oxonomics Society