Decentralization and public services: the case of immunization

https://doi.org/10.1016/j.socscimed.2003.10.013Get rights and content

Abstract

This study examines the impact of political decentralization on an essential public service provided in almost all countries: childhood immunization. The relationship is examined empirically using a time-series data set of 140 low- and middle-income countries from 1980 to 1997. The study finds that decentralization has different effects in low- and middle-income countries. In the low-income group, decentralized countries have higher coverage rates than centralized ones, with an average difference of 8.5 percent for the measles and DTP3 vaccines. In the middle-income group, the reverse effect is observed: decentralized countries have lower coverage rates than centralized ones, with an average difference of 5.2 percent for the same vaccines. Both results are significant at the 99 percent level. Modifiers of the decentralization-immunization relationship also differ in the two groups. In the low-income group, development assistance reduces the gains from decentralization. In the middle-income group, democratic government mitigates the negative effects of decentralization, and decentralization reverses the negative effects of ethnic tension and ethno-linguistic fractionalization, but institutional quality and literacy rates have no interactive effect either way. Similar results are obtained whether decentralization is measured with a dichotomous categorical variable or with more specific measures of fiscal decentralization. The study confirms predictions in the theoretical literature about the negative impact of local political control on services that have public goods characteristics and inter-jurisdictional externalities. Reasons for the difference between low- and middle-income countries are discussed.

Introduction

Decentralization has become an increasingly familiar theme in development theory and practice over the past two decades. Yet the literature on decentralization has concentrated on case studies and theoretical discussions rather than empirical analysis, and few studies have explored the practical consequences of decentralization policies in a quantitative way. The purpose of the present study is to empirically test the impact of decentralization on an essential public service provided in all countries—childhood immunization—and to see if the various benefits proposed for decentralization are, in practice, seen. Given the continued interest in decentralization in developing countries, the findings are expected to be of interest to those in the development community as well as to health specialists concerned specifically with immunization.

Examining the impact of decentralization on immunization services has several objectives. First, examining the decentralization-immunization relationship can help planners and policymakers to predict the likely consequences of decentralizing processes—most of which originate as political decisions outside the health sector—on immunization and other preventive and public health services within the health sector. Decentralization has featured prominently in health sector reform efforts in many countries, and there has been considerable discussion of its potential to negatively affect health services such as immunization and family planning (Kolehmainen-Aitken & Newbrander, 1997; Melgaard, 1998; WHO, 1999; Feilden and Nielson, 2001; POLICY Project, 2001). Published studies of national immunization and family planning programs have also given prominence to the effect of decentralization on these services (Msambichaka, 1998; Feilden et al., 1999; POLICY Project, 2001). As a rule, these studies have used case studies, qualitative methods and managerial indicators to examine the impact of decentralization. The aim of the present study is to complement these works by using a quantitative approach to measure the approximate extent to which decentralization affects the delivery of immunization services, as well as to explore some of the channels through which it works.

Second, if immunization is accepted as a proxy for other (similar) public services, the analysis can illuminate the broader effects of decentralization on public service provision as a whole. Immunization has a number of unique characteristics, so its use as a proxy for all public services is not without shortcomings. But given the ready availability of data on immunization coverage, and the reasonable quality and comparability of these data across time and space, they nevertheless provide a useful lens through which to examine the impact of decentralization policies in a quantitative way (Gauri & Khaleghian, 2002). Caution is required when generalizing such findings, except perhaps to similar services such as malaria control or aspects of maternal and child health: but the exercise can nonetheless provide helpful insights into the effects of decentralization and a platform for testing some of the theoretical benefits proposed for it. This study therefore contributes to the small but emerging literature that uses quantitative analysis to examine the impact of decentralization on public services, examples of which include Bird, Ebel, and Wallich (1995), West and Wong (1995), Isham and Kähkönen (1999), Akin, Hutchinson, & Strumpf (2001) and Faguet (2001).

Section snippets

Definitions

For the purposes of this paper decentralization is defined as the presence of taxing, spending or regulatory authority on the part of subnational authorities such as state, provincial, district or municipal governments. Deconcentration of the health sector (also known as administrative or ministerial decentralization) is not included, since our primary question is about the implications of political decentralization rather than of organizational arrangements within a given sector;1

Decentralization and public services

Many of the proposed benefits of decentralization are based on the premise that it brings local decision makers closer to the constituencies they serve. Implicit in this are assumptions about the nature of information available to local decision makers, the presence of effective channels for the public to express wants and preferences, and the incentive environment motivating decision makers to respond. Each of these assumptions leads to specific conclusions regarding the benefits of

Decentralization and the health sector

Experience in the health sector has confirmed many of the problems summarized above. Community participation has seldom materialized as expected, capacity constraints have prevented decentralized managers from carrying out their new functions or adopting innovative approaches, accountability has failed to emerge, and failures of institutional design have left both central and local authorities confused about their responsibilities and relationships. Case studies have illustrated these issues in

Decentralization and immunization

The impact of decentralization on immunization is a matter of special concern for three reasons. First, immunization is among the cheapest and most effective health interventions available and is a core element of public health programs in all countries. Properly carried out, immunization programs can have a dramatic impact on childhood morbidity and mortality from communicable diseases, especially in developing countries and especially among the poor (Jamison, Mosley, Measham, & Bobadilla, 1993

Data and variables

To examine this question, we use a cross-sectional time-series of data on low- and middle-income countries from 1980 to 1997. Data on immunization coverage were obtained from WHO and UNICEF. Data were obtained on coverage rates for two vaccines: measles vaccine, a single vaccine usually administered at around 9 months of age, and DPT3 vaccine, the third of a three-vaccine series against diphtheria, pertussis (whooping cough) and tetanus that is usually administered at around 10–16 weeks of age.

Methods

Our empirical analysis has two objectives: first, to characterize differences in immunization coverage between decentralized and non-decentralized countries, controlling for other determinants such as national income and contact with donors; and second, to examine how these effects are modified in the presence of factors such as democracy, illiteracy, institutional quality and ethnic heterogeneity. To do this, we model immunization coverage rates using the general formYit=DECit+Xit+Zit+eit,

Decentralization is associated with higher immunization coverage rates in low-income countries, but lower coverage rates in middle-income countries

The parameter estimate for the decentralization variable is large and statistically significant in both low- and middle-income countries. Surprisingly, however, the sign of the coefficient is different for each: it is positive in low-income countries, indicating higher coverage rates in decentralized countries than centralized ones, and negative in middle-income countries, indicating the reverse (Table 3, Table 4). Other things being equal, decentralization is associated with an 8.8 percent

Conclusions

Decentralization is not without benefits. While empirical support may be limited, there are numerous theoretical and common sense reasons to expect a positive relationship between decentralization and various aspects of government performance. One thing is clear, however: decentralization, for all its benefits, is not a panacea. Given proper design and the right environment, decentralization can be an effective reform that meets many of its theoretical expectations. Absent these conditions,

Acknowledgements

The author gratefully acknowledges the comments and assistance of Timothy Baker, William Reinke, Laura Morlock, Carl Taylor, Larry Moulton, Manning Feinleib, Varun Gauri, Ruth Levine, Paul Fife, Imran Hafiz and Hedy Sladovich. The findings, conclusions and interpretations expressed in this paper are those of the author and do not necessarily reflect those of the World Bank, its Executive Directors or the countries they represent.

The word “processed” describes informally reproduced works that

References (91)

  • X. Zhang et al.

    Knowledge, attitude and practice survey on immunization service delivery in Guangxi and Gansu, China

    Social Science & Medicine

    (1999)
  • X. Zheng et al.

    The reforms of the Chinese health care system: county level changesThe Jiangxi study

    Social Science & Medicine

    (1995)
  • Akin, J., Hutchinson, P., & Strumpf, K. (2001). Decentralized and government provision of public goods: the public...
  • J.-L. Araújo

    Attempts to decentralize in recent Brazilian health policyIssues and problems, 1988–1994

    International Journal of Health Services

    (1997)
  • O. Azfar et al.

    Decentralization, governance and public servicesThe impact of institutional arrangements

    (1999)
  • O. Azfar et al.

    Conditions for effective decentralized governanceA synthesis of research findings

    (2001)
  • Bardhan, P., & Mookherjee, D. (1998). Expenditure decentralization and the delivery of public services in developing...
  • Beck, T., Clarke, G., Groff, A., Keefer, P., & Walsh, P. (2000). New tools and new tests in comparative political...
  • N. Beck et al.

    What to do (and not to do) with time-series cross-section Data

    American Political Science Review

    (1995)
  • Besley, T., & Coate, S. (1999). Centralized vs. decentralized provision of local public goods: a political economy...
  • R. Bird et al.

    Decentralization of the socialist stateIntergovernmental finance in transition economies

    (1995)
  • A-E. Birn et al.

    To decentralize or not to decentralize, is that the question? Nicraguan health policy under structural adjustment in the 1990s

    International Journal of Health Services

    (2000)
  • Bos, E., & Batson, A. (2000). Using immunization coverage rates for monitoring health sector performance, measurement...
  • Bossert, T., Beauvais, J., & Bowser, D. (2000). Decentralization of health systems: preliminary review of four country...
  • Bossert, T. J. (2000). Guidelines for promoting decentralization of health systems in Latin America. LAC Health Sector...
  • T.J. Bossert et al.

    Decentralization of health systems in Latin America

    Pan American Journal of Public Health

    (2000)
  • T.J. Bossert et al.

    ‘Bottom-up’ planning in IndonesiaDecentralization in the ministry of health

    Health Policy and Planning

    (1991)
  • J.C. Caldwell

    Routes to low mortality in poor countries

    Population and Development Review

    (1986)
  • J.C. Caldwell et al.

    The social component of mortality declineAn investigation in South India employing alternative methodologies

    Population Studies

    (1983)
  • D. Campos-Outcalt et al.

    Decentralization of health service in Papua New Guinea

    Health Policy and Planning

    (1989)
  • C. Collins

    Decentralization and the need for political and critical analysis

    Health Policy and Planning

    (1989)
  • C. Collins et al.

    Decentralization and primary health careSome negative implications in developing countries

    International Journal of Health Services

    (1994)
  • K. Deininger et al.

    Measuring Income InequalityA new data base

    World Bank Economic Review

    (1996)
  • W. Dillinger

    Decentralization and its implications for urban service delivery

    (1994)
  • Dollar, D., & Pritchett, L. (1998). Assessing aid—what works, what doesn’t, and why. Policy research report, World...
  • Ebel, R. D., & Yilmaz, S. (2002). On the measurement and impact of fiscal decentralization. Policy research working...
  • England, S., Kaddar, M., Nigam, A., & Pinto, M. (2001). Practice and policies on user fees for immunization in...
  • Faguet, J. -P. (2001). Does decentralization increase responsiveness to local needs? Evidence from Bolivia. Policy...
  • Feilden, R., Firsova, S., Gedlik, G., Huseynov, S.Lisitsin, Y. Seitkazleva, N., & Steinglass, R. (1999). Immunization...
  • Feilden, R., & Nielson, O. F. (2001). Immunization and health reform. Making reforms work for immunization. A reference...
  • Fisman, R., & Gatti, R. (2000). Decentralization and corruption. Evidence across countries. Policy research working...
  • Gauri, V., & Khaleghian, P. (2002). Immunization in developing countries: its political and organizational...
  • L. Gilson et al.

    Local government decentralization and the health sector in Tanzania

    Public Administration and Development

    (1994)
  • Girishankar, N. (1998). Reforming institutions for service delivery: a framework for development assistance with an...
  • M. González-Block et al.

    Health services decentralization in MexicoFormulation, implementation and results of policy

    Health Policy and Planning

    (1989)
  • Cited by (75)

    • Examining decentralization and managerial decision making for child immunization program performance in India

      2023, Social Science and Medicine
      Citation Excerpt :

      Studies of decentralization's impact on childhood immunization highlight key features of the decision space framework and their critical roles in health services delivery. In his study examining the impact of political decentralization on childhood immunization across 138 countries, Khaleghian found that decentralized middle-income countries have lower coverage rates than centralized middle-income countries (Khaleghian, 2004). One explanation offered for these findings was that, without sufficient central direction to make immunization a health priority, local authorities may face a greater diversity of demands from the local community, driving negative impacts on immunization coverage.

    • Building and sustaining public and political commitment to the value of vaccination: Recommendations for the Immunization Agenda 2030 (Strategic Priority Area 2)

      2022, Vaccine
      Citation Excerpt :

      While national governments typically oversee immunization program decision-making and strategic planning, all levels of planning and implementation must work in concert. In decentralized systems, strong commitment, leadership and management capacity, and coordination are often even more critical to ensure efficient, effective, equitable program implementation [26,27]. Immunization programs have traditionally focused on delivery—ensuring the right vaccine, in the right condition and right quantities, reaches the right place at the right time.

    • Decentralization of health care systems and health outcomes: Evidence from a natural experiment

      2017, Social Science and Medicine
      Citation Excerpt :

      In fact, there is some recent research suggesting that decentralization funded primarily through grants is likely to lead to a high local government size, while decentralization funded through own taxes (especially when there is a tax separation scheme) favours the containment of local public expenditure (Liberati and Sacchi, 2013). Potential gains to be realised from decentralization are also conditional on the existence of decentralization of political decision-making authority, and in particular, effective channels for the individuals to express their preferences, and incentives for the policymakers to respond to those preferences (Khaleghian, 2004; Bossert and Mitchell, 2011). The potential benefits of decentralization have to be balanced against the presence of spillover effects and public good characteristics, diseconomies of scale, and heterogeneity of preferences (Besley and Coate, 2003).

    View all citing articles on Scopus
    View full text