Decentralization and Decision Space in the Health Sector: A Case Study From Karnatka, India

Rao Seshadri, Kotte et al Page 1

Decentralization and Decision Space in the Health Sector: A case study from Karnataka, India

Shreelata Rao Seshadri, Sandesh Kotte, Latha N and Kalyani Subbiah DRAFT – MARCH 2015 PLEASE DO NOT CITE WITHOUT PERMISSION


1.1 Introduction

Decentralization has a long history in India. Panchayati Raj Institutions (PRI), or local self-government, was established as a three-tier system of decentralized governance at district (Zilla Panchayat), block or taluk (Taluk Panchayat) and village or gram (Gram Panchayat) levels as early as 1957. This was the beginning of several attempts to promote decentralization of funds, functions and functionaries at policy and program levels. Decentralization received its current impetus after the 73rd Constitutional Amendment (1993), which devolved specific powers related to 29 subjects. While the Amendment enabled such devolution at the federal level, the further interpretation and implementation of the Amendment was left to the discretion of state governments. While this provided political support for decentralization, it is more recently that bureaucratic decentralization has been on the agenda of government and donor agencies as a way of enhancing efficiency as well as equity in provision and distribution of services.1,2,3 The call for both political and bureaucratic decentralization has partly been in response to the Health for All agenda as articulated by the Alma Ata Declaration (1978); as well as a growing recognition that citizens need to be given voice to make governments more responsive, cost-effective and accountable.4 Others have argued that decentralisation is a way by which the centralised state can reconnect with social groups from which they have become increasingly distant.5,6 Thus, many developing countries, including India, have adopted decentralization as an institutional response for more effective service delivery.7

Progress towards decentralized governance in India has been slow for several reasons.8,9,10

Although some measure of success in terms of improved infrastructure and equipment, and expanded healthcare delivery have been reported from Kerala where decentralization was undertaken in a comprehensive manner,11 evidence from elsewhere is less encouraging. Political 1 2 Mills et al. Health systems decentralization: Concepts, issues and country experience. World Health Organization, Geneva; 1990. 3 Balarajan, Y., Selvaraj, S., & Subramanian, S. V. (2011). Health care and equity in India. The Lancet, 377(9764), 505-515. 4 Volcker, P. A., and W. F. Winter. Democracy and public service. In Deregulating the public sector: Can government be improved? J. J. DiIulio, Jr. (ed); Washington, DC: Brookings Institution; 1994. 5 Manor, J. (2007). Successful governance reforms in two Indian states: Karnataka and Andhra Pradesh. Commonwealth & Comparative Politics, 45(4), 425-451 6 Manor, J. (2011). Perspectives on Decentralization. 7 Bardhan P. Decentralization of Governance and Development. The Journal of Economic Perspectoves, Vol 16, No 4; 2002. 8 Besley, T., Pande, R., Rahman, L., & Rao, V. (2004). The politics of public good provision: Evidence from Indian local governments. Journal of the European Economic Association, 2(2‐3), 416-426. 9 Singh, N. (2008). Decentralization and public delivery of health care services in India. Health Affairs, 27(4), 991-1001. 10Peters, D. H., Rao, K. S., & Fryatt, R. (2003). Lumping and splitting: the health policy agenda in India. Health Policy and Planning, 18(3), 249-260. 11 Rajesh K and M Benson Thomas. Decentralization and interventions in the health sector. Journal of Health Managemen; tDecember 2012 vol. 14 no. 4 417-433.


and bureaucratic resistance at the state level to sharing powers and resources with lower levels of government has been one reason;12 paucity of funds has also curtailed the functioning of many local bodies. Further, there has been lack of clarity about the responsibilities of these local bodies. As a result, most of these elected bodies function merely as administrative units.

Given the challenges, it is fair to ask whether India is ready for decentralization at all.13

Some have argued that decentralization may lead to inequities in access and inefficiency.14, 15 In addition, there is evidence to indicate that almost a decade of experience with decentralization in India has resulted in little change in the functioning of the health sector.16,17 The reasons given include lack of funds, complex organizational structures and lack of community participation in service delivery.18,19

1.2 Decentralization under the National Rural Health Mission (NRHM) Partly in response to the issues identified above, the Government of India (GOI) attempted

to address issues of health sector governance in the design of the National Rural Health Mission (NRHM).20 It has actively tried to promote decentralization in the health sector, with specific guidelines for the allocation and management of funds and responsibilities related to service delivery at different levels of the health system. This agenda has been largely focused on bureaucratic/programmatic decentralization, which has meant the creation of institutional structures and cadres of workers who are closer to the community.21 However, elected representatives (PRI members) have been systematically involved in the newly constituted institutional structures at all levels in an attempt to provide community voice. The institutional

12 Rao CH. Decentralized planning: An overview of experience and prospects. Economic and Political Weekly; 411-416, 1989. 13 Anirudh Krishna. 2005. “Are Villagers Ready for Decentralization?” In L.C.Jain, ed., Decentralization and Local Governance. New Delhi: Orient Longman. 14 Litvack, J. I., Ahmad, J., & Bird, R. M. (1998). Rethinking decentralization in developing countries. World Bank Publications. 15 Robinson, M. Does decentralization improve equity and efficiency in public service delivery provision? IDS Bulletin, 38(1), pp7-17; 2007. 16 Mahal, A., Srivastava, V., & Sanan, D. (2000). Decentralization and public sector delivery of health and education services: The Indian experience (No. 20). ZEF discussion papers on development policy. 17 Varatharajan, D., Thankappan, R., & Jayapalan, S. (2004). Assessing the performance of primary health centres under decentralized government in Kerala, India. Health Policy and Planning, 19(1), 41-51. 18 See for instance: Hammer, J., Aiyar, Y., & Samji, S. (2007). Understanding government failure in public health services. Economic and Political Weekly, 4049-4057 19 Véron, R., Williams, G., Corbridge, S., & Srivastava, M. (2006). Decentralized corruption or corrupt decentralization? Community monitoring of poverty-alleviation schemes in Eastern India. World Development, 34(11), 1922-1941. 20 Presently renamed as National Health Mission (NHM) but in this article we refer to it as NRHM as our field work was conducted prior to the change in nomenclature. 21Berman, P. A., Gwatkin, D. R., & Burger, S. E. (1987). Community-based health workers: head start or false start towards health for all?. Social Science & Medicine, 25(5), 443-459.


arrangements for decentralization under NRHM are uniform for all states, even though health is listed as a ‘state subject’ under the Indian Constitution, in that implementation of all health programs is the responsibility of the state government, even if funding is partially or wholly provided by the central government.

The goal of the NRHM was as follows: …..to provide effective health care to the rural

population, especially disadvantaged groups including women and children, by improving access, enabling community ownership and demand for services, strengthening public health systems for efficient service delivery, enhancing equity and accountability and promoting decentralization (ref. NRHM framework document). The Mission specifically sought to strengthen the linkages between the health system at primary health care (PHC) level and the community, and provide more effective avenues for communities to express their voice, in several ways:

(i) District Health Committees have been established in each district, responsible for ensuring

decentralized planning, better coordination in the implementation of various health schemes and making available funds through a flexible pooling mechanism;

(ii) Village Health Sanitation and Nutrition Committees (VHSNC) have been constituted at the village level. They are anchored by a PRI member of the village or Gram Panchayat and including other community, civil society and Health Department outreach workers, and are given an annual grant to fund community health initiatives;

(iii) An additional community health worker - the Accredited Social Health Activist (ASHA) – has been appointed to support the VHSNC and to provide community-based primary care; and

(iv) Facility-based committees called Arogya Raksha Samithis (ARS) have been established at all District Hospitals, Community Health Centers (known as Taluk Hospitals in Karnataka) and PHCs comprised of the selected staff of the respective facility, civil society members and at least one PRI member. The experience of decentralization under NRHM has been mixed: multi-state evaluation

of NRHM have concluded that there has been an easing of infrastructural constraints; however, problems relating to human resource recruitment and deployment persist.22,23 This is confirmed by evaluations of the NRHM that reveal concerns with regard to availability of human resources of all cadres and at all levels, both in sufficient numbers and of adequate quality. The importance of creating appropriate learning opportunities for different cadres of health managers and workers has been recognized, with a greater focus on roles and responsibilities for those more closely

22 Sinha, A. In defence of the National Rural Health Mission. Economic and Political Weekly, pp72-75; 2009. 23 Gill, K (2009). A primary evaluation of service delivery under the National Rural Health Mission: Findings from a study in Andhra Pradesh, Uttar Pradesh, Bihar and Rajasthan. Working Paper 1.2009; Planning Commission of India; 2009.


engaged in service delivery, and a more system-wide strategic focus for higher level officials.24 Capacitating district level cadres has been shown to significantly impact their capacity for evidence-based planning.25 This is critical, when considering evidence from Karnataka that shows an absolute mismatch between program implementation plans, funding allocations and actual expenditures.26 Other studies are also critical of the translation of NRHM objectives regarding increasing community engagement and ownership due to lack of clarity in defining the planning process.27 Importantly, there are several issues relating to the recruitment and capacity of the backbone of NRHM: the ASHA worker, and her awareness of her role in the healthcare delivery system.28,29 There are issues as well with the establishment and capacity building of the bodies aimed at ensuring decentralized governance under the NRHM such as the VHSNC and the ARS.30,

31 A study in Uttarakhand demonstrated that members of VHSNCs had very low awareness of the role of the Committee, indicating a clear need for further capacity building of these village-based bodies.32 This is supported by a study from Maharashtra which found very low levels of awareness of the functions of VHSNCs among PRI members and ASHAs, the two most crucial actors in the decentralization framework of the NRHM.33 A study on expenditures under NRHM showed that the bulk of discretionary funds were utilized in March, that is, the end of the fiscal year, and had little to do with being responsive to the needs of the community. The same study found that differences of opinion on the proper use of discretionary funds between the PRI and the Health Department personnel were a significant barrier to efficient use of funds; as also delays in the receipt of funds and lack of clarity on expenditure guidelines.34 Strong state and local institutional capacity, particularly to mobilize strong partnerships and support from civil society organizations

24 Bossert T, S Mazumdar and P Belli. Decentralization of health in the Indian state of West Bengal: Analysis of decision space, institutional capacities and accountability. Working Paper 68499; CSSS, Calcutta; 2010. 25 Shukla A, R Khanna and N Jadhav (2014). Using community-based evidence for decentralized health planning: insights from Maharashtra, India. Health Policy Plan. (2014)doi: 10.1093/heapol/czu099. First published online: October 1, 2014. 26 Gayathri K. District level NRHM funds flow and expenditure: sub-national evidence from the state of Karnataka. Working Paper 278; Institute for Social and Economic Change, Bangalore, India; 2012. 27 Agarwal D. Universal access of health care for all: Exploring roadmap. Indian J Community Med. 2012 Apr-Jun; 37(2): 69–70. 28 Husain Z. Health of the National Rural Health Mission. Economic and Political Weekly, 46(4), 53; 2011. 29 Program Evaluation Organization (PEO). Evaluation study of National Rural Health Mission in 7 states. Delhi: Planning Commission, Government of India; 2011. 30Bajpai, N., Sachs, J. D., & Dholakia, R. H. (2010). Improving access and efficiency in public health services: mid-term evaluation of India's national rural health mission. SAGE Publications India. 31 Kaur, M., Prinja, S., Singh, P. K., & Kumar, R. (2012). Decentralization of health services in India: barriers and facilitating factors.WHO South East Asia Regional Office, New Delhi. 32 Semwal V, SK Jha, CMS Rawat, S Kumar, A Kaur (2013). Assessment of Village Health Sanitation and Nutrition Committees in Nainital District of Uttarakhand. Indian Journal of Community Health; Vol. 25, No. 4 (2013). 33 Sah PK, AV Raut, CH Maliye, SS Gupta, AM Mehendale and BS Garg. Performance of village health, nutrition and sanitation committee: A qualitative study from rural Wardha, Maharashtra. The Health Agenda, Vol I, Issue 4; 2013. 34 Choudhury M, HK Amarnath and BB Dash. Selected aspects of NRHM at the state-level: A focus on Rajasthan and Karnataka. National Institute of Public Finance and Policy; 2013.


such as the PRI, has been identified as a strong predictor of the successful implementation of NRHM initiatives.35

1.3 Research Question

After almost a decade of implementation of NRHM, there is a need to take stock of the

experience so far and examine (i) whether decentralization has stayed at the level of a mandate or whether it has genuinely taken place at the district level and below; and (ii) whether there is any correlation between an increase in decentralized autonomy (decision space) and health outcomes.

Using Bossert’s framework (1998), we study the perception of power and autonomy with

respect to certain critical administrative functions at the district level and below. Bossert’s framework introduces the concept of “Decision-Space”, defined as the choices allowed by central authorities to be utilized by local authorities. The framework maps the range of choice allowed to local officials along a series of functional dimensions. These choices define the working of the decentralized bodies.36 Within this decision space or range of choices, local authorities make innovative choices that are different from the choices they made before decentralization and different from directed change that central authorities impose on localities which have not been decentralized. Bossert (2011) makes a distinction between de jure decision space or the degree of autonomy in local decision-making and use of funds provided by the system; and de facto decision space or the degree of autonomy exercised in practice.

In this paper, we use the example of Karnataka to critically examine whether and to what

extent autonomy has genuinely devolved to lower levels of the health system under the NRHM. Karnataka is considered a pioneer in decentralization as the state started decentralizing processes even before the Constitutional Amendment of 1993. Karnataka has shifted powers with respect to all 29 items set aside for local government. There is evidence that indicates that decentralization efforts in the state have not achieved their objectives.37 Given the evidence linking decentralized autonomy/decision space with improved planning and management capacity, there is a critical need for a rigorous study, using a tested framework, to better understand the dimensions of de facto decentralization in the health sector.

In this context, the main question we want to address in this paper is:

35 Schweitzer J. Improving health services in India: A different perspective. doi: 10.1377/hlthaff.27.4.1002Health Aff July 2008 vol. 27 no. 41002-1004. 36 Bossert, T. (1998). Analyzing the decentralization of health systems in developing countries: decision space, innovation and performance. Social science & medicine, 47(10), 1513-1527. Also, Mitchell, A., & Bossert, T. J. (2010). Decentralisation, Governance and Health‐System Performance:‘ Where You Stand Depends on Where You Sit’. Development Policy Review, 28(6), 669-691 and Bossert, T. J. (2014). Decentralization of Health Systems: Challenges and Global Issues of the Twenty-First Century. In Decentralizing Health Services (pp. 199-207). Springer New York. 37 Center for Good Governance. Inter-state Study on Decentralization. Hyderabad; 2004.


Does the degree of decentralization as implemented under the NRHM correlate strongly

with perceived decision space of Health Department officials and PRI members at the district level and below?

As a corollary to this, we ask another question that is little addressed in the literature.

Ultimately, the avowed goal of decentralization is improved service delivery. The assumption is that the implementation of decentralized governance will result in better management of health programs, that are more responsive to local needs. This will finally result in improved health outcomes.38,39 We test this assumption by asking the following question:

Does greater perceived decision space by Health Department officials and PRI members

at the district level and below result in better managed health systems? We argue that the mechanistic application of NRHM provisions, in the absence of genuine

capacitation and empowerment of lower level functionaries, defeats the ultimate objective of decentralization. We submit that the success of decentralization as a mechanism for greater autonomy and accountability at lower levels of the health system depends to a large extent on the vision driving the initiative from the top and the perception of the effectiveness of decentralization by administrative and political personnel at lower levels.40 In the final analysis, it is the perception of empowerment of the elected representatives and the administrative functionaries at different levels of the health system which will be instrumental in operationalizing decentralization.41

This paper presents the findings of our analysis. We first provide an overview of the

conceptual framework for the analysis as well as a review of literature relating to decentralization in India. Next, we present the study methodology, field sites and stakeholders for the research. The third section presents the findings from the data. The final section concludes, with suggestions for further research. 2.1 Research Design and Methodology 2.1.1. Decision Space Assessment Survey:

38 Saide MAO and DE Stewart. Decentralization and human resource management in the health sector: a case study (1996098) from Nampula, Mazambique. Intl J of Health Planning and Management 2001; 16: 155-168. 39 Grundy J, Healy V, Gorgolon L, Sandig E. Overview of devolution of health services in the Philippines. Rural and Remote Health 3: 220. (Online) 2003. 40 Narayana K. and K. Hari Kurup. Decentralization of the Health Care Sector in Kerala. Centre for Development Studies: Working Paper No. 298; 2000. 41 Kalita A, S Zaidi, V Prasad and VR Raman. Empowering health personnel for decentralized health planning in India: The Public Health Resource Network. Human Resources for Health 2009, 7:57.


The study is based on a survey instrument administered in 2012-13 in five districts of Karnataka. The survey instrument was based on instruments used in similar studies (in Pakistan, West Bengal for example). The instrument was reviewed in consultation with officials of the Department of Health and Family Welfare, Government of Karnataka as well as local health researchers, and modified to be specific to the context in Karnataka. The survey instrument was designed to elicit information from selected respondents on perceived decision space in their current institutional role. Responses were elicited on the following functional dimensions:

Table 1: Health Functions measured in the Study

Function Description Survey Questions Finance and Budgeting

This function related to the respondent’s ability to influence the allocation and expenditure of program budgets and locally raised funds.

There were nine questions that focused on whether the respondent is involved in the budgeting process; change budgeting decisions; have access to program funds; and influence the allocation of non-program funds such as the Local Area Development funds or user fees collected at the facility level.

Contract Management

The NRHM has made provision for contracting in the services of various cadres of health personnel in an effort to fill human resource gaps. In addition, provision is also made for the contracting in of certain services such as cleaning and laundry services. However, this facility is under-utilized due to lack of clarity in contractual arrangements.

There were six questions that focused on whether the respondent was empowered to enter into such contractual arrangements; and whether they were authorized to make payments to contractors.

Service Delivery

This function relates to the actual delivery of health services at the district level and below, particularly the programs under the ambit of the NRHM (reproductive and child health, vertical disease control programs).

There were seventeen questions that focused on whether or not the respondent was empowered to undertaking procurement of drugs and supplies based on local need; to expand availability of certain services based on local demand; to re-deploy infrastructure and equipment based on local need; and their role in monitoring service quality.

Human Resources

Human resource issues are widely recognized as a critical challenge in service delivery. This section focused on the ability to use available human resources more effectively and

There were nine questions that focused on whether the respondent was authorized to monitor vacancies and fill them as required; to recruit new staff; to create new posts; introduce additional training programs; and


employ additional resources as needed.

whether they face any pressure in recruitment and transfers.

Performance Monitoring

This section relates to the monitoring function of various respondents.

There were four questions that focused on whether the respondent was directly responsible for monitoring any program; take steps to re-orient the program based on findings; and reward good performance in any way.

Access Rules This section related to the ability of PRI and civil society members to support the health sector; as well as the ability of the officials of the Health Department to garner support from civil society.

There were twelve questions that focused on whether the respondent could mobilize funds to hold additional health programs (for example, health camps) for the poor; create additional facilities through independent funding; and enhance community participation in health programs.

The response to each question was recorded on a scale of 1-5, where 1 represents the lowest perceived decision space and 5 represents the highest. 2.1.2 Sample Selection: Overall, in five districts, we administered the questionnaire to 93 respondents selected through a (Two-stage) Cluster Stratified Random Sampling methodology. In any one district, the study period was three days. The selection of districts was based on the district classification of the High-Powered Committee for Redressal of Regional Imbalances (HPCRRI; 2002) headed by the late D.M. Nanjundappa (the Nanjundappa Committee) categorized regions and districts in Karnataka based on their performance on the Social Infrastructure Index (SII). The SII was a composite of a Health Index and Education Index, comprised of a total of seven indicators, including number of doctors per 10,000 population, number of government beds per 100,000 population, percentage of households with drinking water facility, literacy rate, and pupil-teacher ratio. Blocks were then scored as being more or less ‘backward’ based on their performance. The four divisions/regions of the state were stratified by rank: Gulbarga was ranked lowest, with 94% of the blocks rated ‘backward’; and Mysore ranked highest with 45% of the blocks rated ‘backward’. Bangalore and Belgaum division, with 71% and 61% ‘backward’ blocks respectively, scored in the middle, and were clubbed into one strata. Each strata was made up of a district cluster: seven each in the lowest and middle ranking clusters, and eight in the highest. Of these, about 30% were selected in the lowest and middle ranking clusters and 13% in the highest ranking cluster (see Table 2a).

Table 2a: Selection of SII-Cluster and Study District


SII-Strata (Ranked by Social Infrastructure Indices)

Division in Strata (% ‘Backward’ Blocks)

District-Cluster in Selected SII-Strata

Random District Selection

Fraction of District-Clusters Sampled

Lowest Gulbarga (94%) 7 1 14%

Middle Bangalore (71%), Belgaum (61%) 7 2 29%

Highest Mysore (45%) 8 2 25% Notes: Definitions of SII and ‘Backward’ Blocks are from the Nanjundappa Committee.

Study respondents were selected from Health Department officials and PRI members at the district level and below. The details of group specific Ns and additional sample information is in Table 2b.

Table 2b: Selection of Study Respondents

Strata (Designation)*

Strata Population per District during Study Period

(approx.)

Sampled in each District during Study

Period (approx.)

Stratified Random Sample

Fraction (approx.)

N (Total=93)

Senior Health Managers (SHM)

7 7 100% (census of strata)

34

Program Managers (PM)

5 to 6 5 to 6 100% (census of strata)

27

Junior Health Managers (JHM)

9 to 10 1 to 2 12.5% to 25% 8

Qualified Panchayati Raj Members (PRI-Q)

5 5 100 % 24

*Definitions: Senior Health Managers (SHM): Includes health sector decision-makers within the district – District Collector, Chief Executive Officer of the Zilla Parishad, District Surgeon, District Health Officer, and Taluk Health Officer. Program Managers (PM): Managers of specific health programs, including Reproductive and Child Health Officer, TB Officer, District Program Manager – NRHM, District Program Management Officer – Karnataka Health Systems Development and Reform Project, District Malaria Officer. Junior Health Managers (JHM): Medical Officer, Primary Health Center. Qualified Panchayati Raj Members (PRI-Q): Includes VHSNC anchors, member of ARS, member of District Health Society, district Member of Legislative Assembly.


2.1.3 Sample Stratification: The sample was stratified in two ways for data collection and analysis: first, by division/region according to the Social Infrastructure Index (SII) ranking (see Table 2a); and second, by designation: Senior Health Managers (SHM), Program Managers (PM), Junior Health Managers (JHM) and Panchayat Raj members belonging to the District Health Society, the VHSNC or the ARS (PRI-Q) members (see Table 2b). We did this to better understand perceptions of different actors who represented districts with significantly different social development outcomes. We also examined each question included in the survey: using the available description of roles and responsibilities of each designation strata from the NRHM website, we segregated the questions according to those that applied to specific designation stratas, removing those that did not apply (for example, contract management was irrelevant for PRI members). 2.1.4 Response Rates

a. Overall Respondent Response Rate was 100% since respondents were contacted in advance to ensure their availability during the study period.

b. Response Bias: The survey form was self-administered by respondent to minimize

confirmation biases due to administrator interactions. In addition, only designation was recorded, without any personal information, in order to minimize response biases.

2.1.5 Indicators of Health Systems Performance:

We next selected a set of health systems performance indicators. Data for these indicators was collected from various sources.

Table 3: Intermediate Indicators of Health Systems Performance by District

Highest SII Middle SII Lowest SII S.No. Chmrng Mandya Tumkur Kolar Yadgir 1 Total funds available to

expenditure (in %) 47 45 42 51 46

2 % of Untied funds utilized-2011-12

88 84.4 67.3 - -

3 Number of Staff contracted 3 11 8 10 6 4 % of Institutional delivery 96.3 96.9 96.4 93.5 76.7 5 % Women with complete

ANC 96.7 96.5 97.6 92.3 72.2

6 % Full immunization against Estimated live births

82.5 83 93.7 83.7 55.9

7 % of PHC MOs posts filled 84.8 81.2 80 74.7 70.4


8 % of ANM posts filled 83 77 79 77 62 9 % of specialist posts filled

up 60.6 76.5 64 85.2 41.1

10 Target achieved in VHSC members training (%)

100 91 100 100 93.3

11 Target achieved in ARS members training (%)

NA 89.3 98.5 100 89.8

Data Sources according to S.No:

1, 3, 7, 8, 9, 10 and 11: Data from Ministry of Health and Family Welfare, Govt of Karnataka. Accessed by authors. 2: Highlights from the reports of the Regional Evaluation Teams During 2012-13, Statistics Division, Ministry of Health & Family Welfare, Government of India. 4, 5, 6: District Level Household Survey 4 (2012-13).

2.2 Data Analysis

For the analysis, a composite score was calculated as the unweighted mean of all the individual responses provided by the respondent for the questions relating to any specific functional area (Bossert et al, 2011). So, for example, we took the sum of the perception scores of SHM for all nine questions in the functional category “Finance and Budgeting”; the composite score of SHM was then calculated as the unweighted mean to yield a nine item Likert scale with five points.

The analysis was conducted in two stages. In the first stage, the composite scores of the four designation strata were compared along all six functional categories and tested for significance. Similarly, the composite scores of the three SII divisions were compared along all six functional categories and tested for significance. Finally, the composite scores of the two strata – designation and SII divisions – were compared and tested for significance.

In the second stage, we looked at how perceived decision space in the six functional categories impacted performance on a set of selected intermediate indicators of health systems performance by district (Table 3). The composite scores of the two strata – designation and SII divisions – were correlated with eleven output indicators representing the six functional areas and tested for significance. The proposition for this analysis is that perceived decision space by the different designations (SHM, PM, PRI and JHM) in all functional areas (finance and budgeting, contract management, service delivery, human resource management, performance management and access rules) is systematically correlated with health systems indicators.

3.1 Results


We first present the findings of the first stage of analysis: comparisons of the composite decision space scores by designation and by SII division, with tests of significance.


Table 4: Analysis of Mean Decision Space by Designation and Function SHM PM t-test JHM t-test PRI t-test Mean SD Mean SD Mean SD Mean SD

Fin & Budg 2.3 1 2.2 1.1 t=0.46 2.8 1 t=1.35 1.5*** 0.8 t=5.01

Cont Mgmt 1.7 0.9 1.7 0.8 t=0.30 1.2*** 0.3 t=4.81 Service Del 2.8 0.9 2.2*** 1 t=3.30 3.6*** 0.6 t=3.67 2.6 1.9 t=0.56 HR Mgmt 1.7 0.6 1.9 1 t=1.08 1.9 0.5 t=1.26 Perf Mon 3.7 1.3 3.5 1.4 t=-0.78 4.1 1 t=1.18 2.8*** 1.3 t=0.21 Access 2.1 0.8 2.6* 1.5 t=1.9 2 0.4 t=0.79 2.7** 1.1 t=2.41

PM JHM t-test PRI t-test Mean SD Mean SD Mean SD Fin & Budg 2.2 1.1 2.8 1 t=1.64 1.5*** 0.8 t=3.38 Cont Mgmt 1.7 0.8 1.2*** 0.3 t=4.81 Service Del 2.2 1 3.6*** 0.6 t=6.49 2.6** 1.9 t=2.55 HR Mgmt 1.9 1 1.9 0.5 t=0.09 Perf Mon 3.5 1.4 4.1 1 t=1.74 2.8** 1.3 t=2.56 Access 2.6 1.5 2*** 0.4 t=4.71 2.7 1.1 t=0.25

JHM PRI t-test Mean SD Mean SD Fin & Budg 2.8 1 1.5*** 0.8 t=3.65 Cont Mgmt 1.2 0.3 Service Del 3.6 0.6 2.6*** 1.9 t=4.6 HR Mgmt 1.9 0.5 Perf Mon 4.1 1 2.8*** 1.3 t=3.68 Access 2 0.4 2.7*** 1.1 t=5.49

***p=<.005;** p=<.05; *p=<0.10


The data in Table 4 indicate that: (i) Perception of decision space in Finance and Budgeting is relatively low across all roles,

with no significant difference between mean scores of the different designations. Only the PRI have scored significantly lower on this function than all other designations;

(ii) Similarly, with means ranging between 1.2-1,7, perception of decision space in Contract Management is also uniformly low across all designations, with JHM scoring significantly lower than the others;

(iii) Human Resource Management is another area where all designations have scored uniformly low with means ranging from 1.7-1.9;

(iv) Performance Monitoring scored the highest, with scores ranging from 3.7-4.1 for respondents from the Health Department. Only PRI scored significantly lower than all other designations on this function;

(v) With regard to Access Rules, PRI members have scored significantly higher than SHMs and JHMs, but comparable to PMs. There is a wide variance within both these groups, however, with a SD of 1.1 and 1.5 respectively.

Overall, perception of decision space across several functional areas is uniformly low, and

the variation in some areas is high. The group that appears to score significantly lower as compared to the other designations on all functions (except Access Rules) are the PRI members. As a critical partner in the government’s decentralization strategy, this evidence of low perceived empowerment among PRI members is a cause for concern.

We next look at variation in perceived decision space among the four designations in the three SII strata. Here again, while the overall perception of decision space is low, there are distinct patterns that emerge.


Table 5: Analysis of Mean Decision Space by District and Roles SHM

SII-Division High SII-Division Medium t-test SII-Division Low t-test SII-Division Medium SII-Division Low t-test

Mean SD Mean SD df=10 Mean SD df=18 Mean SD Mean SD df-10 Fin & Bud 2.4 1 2 1 t=1.13 2.2 1 t=0.74 2 1 2.2 1 t=0.54 Cont Mgmt 1.8 1 1.5 0.8 t=1.1 1.4* 0.3 t=1.9 1.5 0.8 1.4 0.3 t=0.69 Service Del 2.8 0.9 2.9 0.9 t=0.43 2.3* 0.4 t=2.08 2.9 0.9 2.3* 0.4 t=2.07 HR Mgmt 1.7 0.8 1.9 0.5 t=0.93 1.5 0.4 t=1.13 1.9 0.5 1.5** 0.4 t=2.30 Perf Mon 3.4 1.3 4.2* 1.3 t=1.92 3.9* 0.6 t=1.8 4.2 1.3 3.9 0.6 t=0.57 Access Rules 2 1 2 0.6 t=0.23 2.4 0.2 t=1.67 2 0.6 2.4** 0.2 t=2.3

PM


Mean SD Mean SD df=11 Mean SD d=10 Mean SD Mean SD df=11 Fin & Bud 2.5 1 2.3 1.2 t=0.57 1.2*** 0.3 t=4.47 2.3 1.2 1.2*** 0.3 t=3.32 Cont Mgmt 1.8 0.8 1.7 1 t=0.10 1.1** 0.3 t=2.41 1.7 1 1.1* 0.3 t=2.1 Service Del 2.4 0.9 2.2 1.1 t=0.69 1.3*** 0.6 t=4.42 2.2 1.1 1.3** 0.6 t=3.0 HR Mgmt 1.9 1 2 1.1 t=0.35 1.9 0.8 t=0.02 2 1.1 1.9 0.8 t=0.36 Perf Mon 3.9 1.2 3.4 1.6 t=1.19 2.8*** 1.2 t=3.32 3.4 1.6 2.8 1.2 t=1.29 Access Rules 3.4 1.3 2.2*** 1.3 t=3.32 1.8*** 1.5 t=4.4 2.2 1.3 1.8 1.5 t=1.14

PRI


Mean SD Mean SD df=8 Mean SD df=11 Mean SD Mean SD df=8 Fin & Bud 1.6 0.7 1.6 0.9 t=0.15 1 0 t=3.02 1.6 0.9 1* 0 t=2.0 Service Del 4 1.7 1.2*** 0.7 t=12.5 1*** 0 t=6.29 1.2 0.7 1 0 t=1.0 Perf Mon 3.8 1 1.8*** 0.8 t=7.63 1.6*** 1 t=8 1.8 0.8 1.6 1 t=0.96 Access Rules 3.2 1.2 2.3*** 0.7 t=3.48 1.5*** 0.6 t=4.85 2.3 0.7 1.5** 0.6 t=3.12

JHM


Mean SD Mean SD df=3 Mean SD df=1 Mean SD Mean SD df=3 Fin & Bud 3.3 0.4 2.5 1.3 t=1.16 2.8 0.4 t=1.67 2.5 1.3 2.8 0.4 t=0.52 Cont Mgmt 1.4 0.6 1.2* 0.2 t=2.6 1 0.6 t=1.0 1.2* 0.2 1 0.6 t=1.73 Service Del 4 0.1 3.2*** 0.2 t=7.4 4 0.1 t=2.0 3.2*** 0.2 4*** 0.1 t=6.66 HR Mgmt 2.2 0.9 1.7** 0.3 t=3.68 2.2 0.9 1.7** 0.3 2.2** 0.9 t=3.68

Perf Mon 4.8 0.4 3.8 1.2 t=1.54 4.8 0.4 t=2.5 3.8 1.2 4.8 0.4 t=0.52

Access Rules 2.3 0.1 1.8 0.4 t=2.16 2.3* 0.1 t=7 1.8 0.4 2.3* 0.1 t=0.72


***p=<.005;** p=<.05; *p=<0.10

What clearly emerges is that the low SII-Division scores significantly lower across most functions among all designations (Table 5):

(i) SHM in SII-Division Low score significantly lower on Contract Management, Service Delivery and Performance Monitoring than SHM in SII-Division High; and significantly lower on Service Delivery, Human Resource Management and Access Rules as compared to SII-Division Medium;

(ii) PM in SII-Division Low score significantly lower on all functions other than Human Resource Management as compared to SII-Division High; and significantly lower on Finance and Budgeting, Contract Management and Service Delivery as compared to SII-Division Medium;

(iii) PRI in SII-Division Low score significantly lower on Service Delivery, Performance Monitoring and Access Rules as compared to SII-Division High; and significantly lower on Finance and Budgeting and Access Rules as compared to SII-Division Medium; and

(iv) JHM in SII-Division Low score significantly lower on Access Rules as compared to SII-Division High; and significantly lower on Service Delivery, Human Resource Management and Access Rules as compared to SII-Division Medium.

There are significant differences in scores between SII-Division High and SII-Division Medium as well, but these are not so extensive. Among SHM and PM, significant differences are observed in Performance Monitoring and Access Rules respectively. Among PRI in the two divisions, the differences in scores are more widespread, with PRI in SII-Division Medium scoring significantly lower on Service Delivery, Performance Monitoring and Access Rules.

Overall, the emerging pattern is mixed: all cadres of health personnel in low SII districts appear to score themselves significantly lower on decision space than corresponding cadres in higher SII districts. However, the difference in scores between medium and high SII districts is less consistent; among SHM and PM, there is virtually no difference in perceived decision space. The critical difference here appears to be the perception of the PRI, who perceive themselves to be significantly more empowered in the higher SII districts.

Finally, we compare the composite SII-Division scores for each functional category (Table 6). Table 4 supports combining of decision space scores across designation strata as, with a few exceptions, differences in perceived decision space scores are not significant. This analysis demonstrates more clearly the finding noted in the previous section: low SII districts score significantly lower than both medium and high SII districts in perceived decision space on all functions. The differences between medium and high SII districts are less wide ranging, with SII-Division Medium districts scoring significantly lower on Service Delivery and Access Rules. Importantly, the score of all districts on Human Resource Management is uniformly low and there was no significant difference between the districts. This points to a narrow sense of


autonomy in being able to make decisions relating to human resources at the district level and below, regardless of designation.

Table 6: Analysis of Mean Decision Space by SII-Divisions

SII-Division

High SII-Division

Medium t-test

SII-Division Low

t-test

SII-Divison Medium

SII-Division Low

t-test

Mean SD Mean SD df=35 Mean

SD

df=43 Mean SD Mean

SD

df=35

Fin & Bud 2.2 1 2.1 1.1

t=0.7 1.7***

0.9

t=3.65 2.1 1.1 1.7**

0.9

t=2.08

Con Mgmt 1.8 0.9 1.6 0.8

t=1.46 1.2***

0.3

t=3.41 1.6 0.8 1.2**

0.3

t=2.53

Service Del 3.1 1.3 2.3*** 1.1

t=4.5 1.9***

1.1

t=6.06 2.3 1.1 1.9**

1.1

t=2.06

HR Mgmt 1.8 0.8 1.9 0.8

t=0.51 1.8

0.6

t=0.02 1.9 0.8 1.8

0.6

t=0.51

Perf Mon 3.7 1.2 3.3 1.5

t=1.67 3.1***

1.3

t=3.43 3.3 1.5 3.1

1.3

t=0.66

Access Rules 2.7 1.2 2.1*** 0.9

t=3.9 1.9***

0.9

t=4.4 2.1 0.9 1.9

0.9

t=1.45

***p=<.005;** p=<.05; *p=<0.10

3.2 How does perceived Decision Space relate to health systems performance?

We then explored the relationship between performance of the SII-Division strata on the selected Health Systems indicators and perceived decision space on the six functions. The data was reduced to a common unit of analysis – the district level. Correlations on the composite scores (average for the district on the Likert scales) for each functional area with the data on eleven indicators of health systems performance were run for the five districts. The results are shown in Table 7. A strong correlation has been defined as a coefficient of magnitude greater than 0.6, whatever the significance level.42

Table 7: District-level Bivariate Correlations (Pearsons) – Functions and Selected Health Systems Indicators

Decisi

on Space Score

Total funds

available

Untied

funds

Staff contrac

ted

% Instituti

onal delivery

% Complete ANC

Full immunization

PHC MO posts filled

ANM posts filled

Specialist

posts filled

VHSC training target

ARS traini

ng target

42 See inter alia Plug I et al. Socio-economic inequalities in mortality from conditions amenable to medical interventions: do they reflect inequality of access or quality of health care? BMC Public Health 1:346; 2012.


utilized

Fin and Bud score

-0.12 0.94 -0.06 0.92* 0.92* 0.77 0.98** 0.94* 0.51 0.29 0.21

Contract

Mgmt score

-0.44 0.35 0.13 0.88* 0.89* 0.78 0.92* 0.81 0.44 0.06 0.02

Service Del score

0.04 0.86 0.21 0.69 0.66 0.43 0.75 0.62 0.53 -0.26 -0.31

HR Mgmt score

-0.89* -0.85 0.38 0.18 0.21 0.33 0.12 -0.01 -0.08 -0.26 -0.11

Perf Mon -.07 .99 -.23 .71 .70 .463 .93* .76 .29 .04 -.24

Access Rules Score

-0.18 0.88 -0.49 0.57 0.57 0.35 0.91* 0.68 0.03 0.11 -0.38

*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).

Greater perception of decision space in financing and budgeting are significantly

correlated with utilization of untied funds; indicators of service delivery such as institutional deliveries, percentage of women receiving full ante-natal care, and full immunization of children in the age group 0-59 months; and human resource issues such as filling PHC medical officer and Auxiliary Nurse Midwife (ANM) posts.

Perceived decision space in contract management is positively correlated with all three service delivery indicators as well as two of the three indicators of human resource management. Not surprisingly, higher decision space scores in service delivery are correlated significantly with indicators of service delivery as well as the availability of human resources. It is also correlated with utilization of untied funds, which was the intention of the NRHM: to make available flexible financing to respond to local service needs.

Performance Monitoring is another area where high decision space scores are correlated with multiple indicators of health system performance. These include better utilization of flexible financing, measures of service delivery as well as human resource availability. Higher perceived decision space on Access Rules is correlated with better availability of human resources at the PHC level and below: this could reflect the fact that this level of health services is directly under the control of the PRI in Karnataka, and the higher perceived decision space of PRI members on Access Rules is perhaps exercised on ensuring availability of health personnel.


Overall, several of the functional categories are significantly correlated with indicators of health systems performance. These findings are indicative and provide insights that need to be studied further for more definitive results.

4 Discussion

We started out with the question: Does the degree of decentralization under the NRHM correlate strongly with perceived decision space of Health Department officials and PRI members at the District level and below?

Evidence from previous studies of decision space in Africa and the Philippines (Bossert et al, 2002), West Bengal, India (Bossert et al, 2010) and Pakistan (Bossert et al, 2011) showed that (i) despite a uniform mandate for decentralization, different district-level respondents experience varying degrees of decision space; (ii) when examined in the context of four broad health functions i.e. planning, budgeting, human resource management and service delivery, decision space in the different functions were significantly correlated with each other; and (iii) the success of decentralization is highly dependent on institutional capacity to effectively exercise decision space.

Our analysis confirms the first of these findings: on the six functions measured in this study

– finance and budgeting, contract management, service delivery, human resource management, performance monitoring and access rule - there is a wide variation in perceptions of decision space across districts and across designations. Even though the guidelines for the decentralization of funds, functions and functionaries are common across the board, variation in their application and interpretation has resulted in uneven uptake of the intended autonomy.

Second, we found that there were some areas where perceived decision space was

uniformly low, regardless of designation or location. For example, Human Resource Management was an area where all designations, including SHMs, perceived themselves to have low autonomy in decision-making. This was true also of contract management. These were also the two areas that scored uniformly low irrespective of SII-Division level (high, medium or low). This is a matter of concern, since under NRHM there are separate funds and guidelines provided for contracting out selected services and human resources. This raises questions related to (i) the awareness of SHMs of the provisions under NRHM for addressing human resource challenges; and (ii) whether there is an enabling environment within the state to apply these provisions, in terms of systems, procedures and clearances.

Third, apart from on Access Rules, PRI members scored significantly lower than all other

designations on all other functional categories. Since they are the primary instrument of political decentralization, this is a cause for concern. One of the major objectives of the NRHM was to involve the community in decentralized health care delivery and allied activities. The idea was that greater community participation would result in greater systemic responsiveness and


accountability, and allow for need-based planning of health interventions. Unless they see themselves as having the autonomy to meaningfully participate in the planning, resourcing and implementation of programs, it is unlikely that to build their sense of ownership of these programs. Engaging with the community to enhance demand for services as well as working with communities to hold the system accountable to their needs is a central role envisaged for people’s organizations under the NRHM. If this essential function is not being served, then mechanisms need to be developed to support this objective. Although training of PRI members in their roles and responsibilities has been undertaken, there is a need for more handholding and regular refresher training. Other active community-based organizations such as Self-Help Groups and Co-operative Groups could be engaged to support such efforts. This finding could have implications beyond the health sector, since many other departments have adopted a similar decentralization strategy vis-à-vis the PRI.

Fourth, the lowest SII-Division scores significantly lower on decision space as compared

to the high and medium SII-Divisions on several functional categories. This points to functionaries in poorly performing districts having a poorer sense of autonomy and empowerment than their counterparts in better performing districts.

Next, we asked: Does greater perceived decision space by Health Department officials and

PRI members at the district level and below result in better managed health systems? With regard to health systems performance, the study provides important insights into

functional areas and their relationship with key health outputs; and could give guidance on areas that require strengthening in order to bring about improvements in key health indicators. The results of the study present evidence of a positive correlation between decentralisation and health systems performance, unlike cross-country comparative studies that have found insufficient evidence for it.43 In the case of Pakistan, a correlation was found between decision space and health service delivery; but in other areas, the outcomes were either mixed or hard to explain.44 We find that decision space scores in several functional categories correlate significantly with selected indicators of health system performance. Perceived decision space in Finance and Budgeting and Performance Monitoring, for example, are significantly correlated with measures of financial autonomy, service delivery and human resource management. Human Resource Management shows the least correlation with most indicators. This analysis requires to be taken further more rigorously as a future research effort.

The findings of the study have important implications for the future implementation of decentralization measures under the NRHM. As the flagship program whose objective it is to

43Bossert TJ, Beauvais JC. Decentralization of health systems in Ghana, Zambia, Uganda and the Philippines: a comparative analysis of decision space. Health Policy and Planning 17(2002):14–31. 44 Bossert, T. J., & Mitchell, A. D. (2011). Health sector decentralization and local decision-making: Decision space, institutional capacities and accountability in Pakistan. Social Science & Medicine, 72(1), 39-48.


provide people-centered health services, the NRHM needs to be strengthened on many fronts. The analysis of the data shows us that, even after almost a decade of implementation, (i) there is a wide gap in the stated guidelines for decentralization under the NRHM, and the perceptions of what this decentralization means in action by various Health Department personnel and PRI members; (ii) this gap in perception is significantly greater in low SII-Division districts as compared to better performing ones; and (iii) this gap in perception could have important impacts on health outputs at the district level.

Given this, there is a need to focus on moving decentralization from being de jure to de

facto, by genuinely providing autonomy in key functional areas down to the district level and below. This will require political will, with the state government committing itself to devolving more powers to the lower levels of government, particularly in areas such as finance and budgeting and human resource management, which appear currently to be highly centralized. On the other side, functionaries of both the Health Department and PRI need to be capacitated to take ownership of the programs with a clear understanding of their roles and responsibilities. The perception of decision space is critical: when functionaries at the district level feel empowered and able to function with autonomy, they can have a substantial impact on key indicators of health systems performance. The challenge for policy makers, and state and national managers, is to strategically capacitate the district team to be able to fulfill the expectations of the NRHM. This effort needs to go beyond meeting the annual training ‘targets’, and should systematically assess and address the capacity gaps of different cadres of personnel. A strategic plan for on-going capacity building and handholding as required of all personnel at the district level and below, including particularly PRI members who are meant to catalyze community participation in health service delivery, is urgently required. Decentralized decision-making can have far-reaching benefits on program implementation and health system performance; but this can happen only when the relevant functionaries truly experience the impact of decentralization in terms of control over resource allocation, service delivery and health outcomes. 5 Limitations of the Study

As mentioned earlier, the study covers a of health managers in five districts of Karnataka. As the district is the unit of analysis for performance indicators, this resulted in a small sample of five district-level records, with limited variation in the data. Therefore, the correlational analysis presented in Table 7 is exploratory in nature, and could for the basis for the development of hypotheses to be tested in larger sample descriptive research. In addition, this study represents a cross-sectional data analysis, whereas the dynamics of the empowerment-performance relationship could be better studied through longitudinal data as this would allow the opportunity to determine causal directions and tease out time-lag effects. The study raises many important questions on the factors that influence individual perceptions of decision space, and how they can be fine-tuned to ensure better results. Widening the scope of the study to (i) include motivational variables at the respondent unit of analysis and (ii) more variation in the district unit of analysis variable through


a national level study. The findings of the study point to the need to build a much more comprehensive evidence-base on how to make a district-level health system function effectively.


References

1. Agarwal D. Universal access of health care for all: Exploring roadmap. Indian J

Community Med. 2012 Apr-Jun; 37(2): 69–70. 2. Anirudh Krishna. 2005. “Are Villagers Ready for Decentralization?” In L.C.Jain, ed.,

Decentralization and Local Governance. New Delhi: Orient Longman. 3. Bajpai, N., Sachs, J. D., & Dholakia, R. H. (2010). Improving access and efficiency in

public health services: mid-term evaluation of India's national rural health mission. SAGE Publications India.

4. Bardhan P. Decentralization of Governance and Development. The Journal of Economic Perspectoves, Vol 16, No 4; 2002.

5. Balarajan, Y., Selvaraj, S., & Subramanian, S. V. (2011). Health care and equity in India. The Lancet, 377(9764), 505-515.

6. Berman, P. A., Gwatkin, D. R., & Burger, S. E. (1987). Community-based health workers: head start or false start towards health for all?. Social Science & Medicine, 25(5), 443-459.

7. Besley, T., Pande, R., Rahman, L., & Rao, V. (2004). The politics of public good provision: Evidence from Indian local governments. Journal of the European Economic Association, 2(2‐3), 416-426.

8. Bossert, T. (1998). Analyzing the decentralization of health systems in developing countries: decision space, innovation and performance. Social science & medicine, 47(10), 1513-1527.

9. Bossert TJ, Beauvais JC.(2002). Decentralization of health systems in Ghana, Zambia, Uganda and the Philippines: a comparative analysis of decision space. Health Policy and Planning 17(2002):14–31.

10. Bossert T, S Mazumdar and P Belli. Decentralization of health in the Indian state of West Bengal: Analysis of decision space, institutional capacities and accountability. Working Paper 68499; CSSS, Calcutta; 2010.

11. Bossert, T. J. and Mitchell, A. D. (2011). Health sector decentralization and local decision-making: Decision space, institutional capacities and accountability in Pakistan. Social Science & Medicine, 72(1), 39-48.

12. Bossert, T. J. (2014). Decentralization of Health Systems: Challenges and Global Issues of the Twenty-First Century. In Decentralizing Health Services (pp. 199-207). Springer New York.

13. Center for Good Governance. Inter-state Study on Decentralization. Hyderabad; 2004. 14. Choudhury M, HK Amarnath and BB Dash. Selected aspects of NRHM at the state-level:

A focus on Rajasthan and Karnataka. National Institute of Public Finance and Policy; 2013.

15. Gayathri K. District level NRHM funds flow and expenditure: sub-national evidence from the state of Karnataka. Working Paper 278; Institute for Social and Economic Change, Bangalore, India; 2012.

16. Gill, K (2009). A primary evaluation of service delivery under the National Rural Health Mission: Findings from a study in Andhra Pradesh, Uttar Pradesh, Bihar and Rajasthan. Working Paper 1.2009; Planning Commission of India; 2009.

17. Grundy J, Healy V, Gorgolon L, Sandig E. Overview of devolution of health services in the Philippines. Rural and Remote Health 3: 220. (Online) 2003.


18. Hammer, J., Aiyar, Y., & Samji, S. (2007). Understanding government failure in public health services. Economic and Political Weekly, 4049-4057.

19. Husain Z. Health of the National Rural Health Mission. Economic and Political Weekly, 46(4), 53; 2011.

20. Kalita A, S Zaidi, V Prasad and VR Raman (2009). Empowering health personnel for decentralized health planning in India: The Public Health Resource Network. Human Resources for Health 2009, 7:57.

21. Kaur, M., Prinja, S., Singh, P. K., & Kumar, R. (2012). Decentralization of health services in India: barriers and facilitating factors. WHO Regional Office, New Delhi.

22. Litvack, J. I., Ahmad, J., & Bird, R. M. (1998). Rethinking decentralization in developing countries. World Bank Publications.

23. Mahal, A., Srivastava, V., & Sanan, D. (2000). Decentralization and public sector delivery of health and education services: The Indian experience (No. 20). ZEF discussion papers on development policy.

24. Manor, J. (2007). Successful governance reforms in two Indian states: Karnataka and Andhra Pradesh. Commonwealth & Comparative Politics, 45(4), 425-451.

25. Manor, J. (2011). Perspectives on Decentralization. 26. Mills J.P. Vaughan, D.L. Smith and I.Tabibzadeh (eds.) (1990). Health systems

decentralization: Concepts, issues and country experience. World Health Organization, Geneva; 1990.

27. Mitchell, A. and Bossert, T. J. (2010). Decentralisation, Governance and Health‐System Performance:‘Where You Stand Depends on Where You Sit’. Development Policy Review, 28(6), 669-691.

28. Muraleedharan, K., Chaudhury, N., Hammer, J., Kremer, M., & Rogers, F. H. (2011). Is There a Doctor in the House? Medical Worker Absence in India. Working paper

29. Narayana K. and K. Hari Kurup. “Decentralization of the Health Care Sector in Kerala”. Centre for Development Studies: Working Paper No. 298; 2000.

30. Peters, D. H., Rao, K. S., & Fryatt, R. (2003). Lumping and splitting: the health policy agenda in India. Health Policy and Planning, 18(3), 249-260.

31. Plug I et al. Socio-economic inequalities in mortality from conditions amenable to medical interventions: do they reflect inequality of access or quality of health care? BMC Public Health 1:346; 2012.

32. Poornima and Vyasulu, V., "Women in Panchayati Raj: Grassroots Democracy in India Experience from Malgudi", UNDP, New Delhi, India (1999).

33. Program Evaluation Organization (PEO). Evaluation study of National Rural Health Mission in 7 states. Delhi: Planning Commission, Government of India; 2011.

34. Rajesh K and M Benson Thomas. Decentralization and interventions in the health sector. Journal of Health Managemen; tDecember 2012 vol. 14 no. 4 417-433.

35. Rao CH. Decentralized planning: An overview of experience and prospects. Economic and Political Weekly; 411-416, 1989.

36. Robalino A, David Oscar F. Picazo and Albertus Voetberg. “Does Fiscal Decentralization Improve Health Outcomes? Evidence from a Cross-Country analysis.” World Bank policy research (2001): working paper 2565.

37. Robinson, M. Does decentralization improve equity and efficiency in public service delivery provision? IDS Bulletin, 38(1), pp7-17; 2007.


38. Sah PK, AV Raut, CH Maliye, SS Gupta, AM Mehendale and BS Garg. Performance of village health, nutrition and sanitation committee: A qualitative study from rural Wardha, Maharashtra. The Health Agenda, Vol I, Issue 4; 2013.

39. Saide MAO and DE Stewart. Decentralization and human resource management in the health sector: a case study (1996098) from Nampula, Mazambique. Intl J of Health Planning and Management 2001; 16: 155-168.

40. Schweitzer J. Improving health services in India: A different perspective. doi: 10.1377/hlthaff.27.4.1002Health Aff July 2008 vol. 27 no. 41002-1004.

41. Semwal V, SK Jha, CMS Rawat, S Kumar, A Kaur (2013). Assessment of Village Health Sanitation and Nutrition Committees in Nainital District of Uttarakhand. Indian Journal of Community Health; Vol. 25, No. 4 (2013).

42. Shukla A, R Khanna and N Jadhav (2014). Using community-based evidence for decentralized health planning: insights from Maharashtra, India. Health Policy Plan. (2014)doi: 10.1093/heapol/czu099. First published online: October 1, 2014.

43. Singh, N. (2008). Decentralization and public delivery of health care services in India. Health Affairs, 27(4), 991-1001.

44. Sinha, A. (2009). In defence of the National Rural Health Mission. Economic and Political Weekly, pp72-75.

45. Varatharajan, D., Thankappan, R., & Jayapalan, S. (2004). Assessing the performance of primary health centres under decentralized government in Kerala, India. Health Policy and Planning, 19(1), 41-51.

46. Véron, R., Williams, G., Corbridge, S., & Srivastava, M. (2006). Decentralized corruption or corrupt decentralization? Community monitoring of poverty-alleviation schemes in Eastern India. World Development, 34(11), 1922-1941.

47. Volcker, P. A., and W. F. Winter. Democracy and public service. In Deregulating the public sector: Can government be improved? In J. J. DiIulio, Jr.(ed) Washington, DC: Brookings Institution; 1994.

Decentralization and Decision Space in the Health Sector: A Case Study From Karnatka, India

Documents