Decentralization and Health Outcomes

8/4/2019 Decentralization and Health Outcomes

1/28

http://jhm.sagepub.com/Journal of Health Management

http://jhm.sagepub.com/content/11/1/167The online version of this article can be found at:

DOI: 10.1177/097206340901100112

2009 11: 167Journal of Health ManagementNitya Mohan

Health : An Analysis of Outcomes for Women in Two South Indian StatesDemocratic Decentralisation and the Millennium Development Goals for

Published by:

http://www.sagepublications.com

can be found at:Journal of Health ManagementAdditional services and information for

http://jhm.sagepub.com/cgi/alertsEmail Alerts:

http://jhm.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://jhm.sagepub.com/content/11/1/167.refs.htmlCitations:

at Universitaetsbibliothek on June 18, 2011jhm.sagepub.comDownloaded from
http://jhm.sagepub.com/http://jhm.sagepub.com/http://jhm.sagepub.com/http://jhm.sagepub.com/content/11/1/167http://jhm.sagepub.com/content/11/1/167http://www.sagepublications.com/http://jhm.sagepub.com/cgi/alertshttp://jhm.sagepub.com/cgi/alertshttp://jhm.sagepub.com/subscriptionshttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsPermissions.navhttp://jhm.sagepub.com/content/11/1/167.refs.htmlhttp://jhm.sagepub.com/content/11/1/167.refs.htmlhttp://jhm.sagepub.com/content/11/1/167.refs.htmlhttp://jhm.sagepub.com/http://jhm.sagepub.com/http://jhm.sagepub.com/http://jhm.sagepub.com/http://jhm.sagepub.com/content/11/1/167.refs.htmlhttp://www.sagepub.com/journalsPermissions.navhttp://www.sagepub.com/journalsReprints.navhttp://jhm.sagepub.com/subscriptionshttp://jhm.sagepub.com/cgi/alertshttp://www.sagepublications.com/http://jhm.sagepub.com/content/11/1/167http://jhm.sagepub.com/


2/28

Journal of Health Management, 11, 1 (2009): 167193SAGE Publications Los Angeles London New Delhi Singapore Washington DCDOI:10.1177/097206340901100112

Democratic Decentralisation and theMillennium Development Goals forHealth: An Analysis of Outcomes for

Women in Two South Indian States

Nitya Mohan

In the context of the targets for primary health identified by the MDGs this article evaluatesthe link between decentralisation and positive outcomes for women and children. UsingIndia as a case study, the article traces the changes in health attainments as a result of decen-tralisation reforms.

The evidence presented, drawn from the experiences of two states, speaks to the relevanceof such a link. Despite the heterogeneity of contexts and in implementation, in generaldemocratic decentralisation has enhanced health outcomes for women in the selected village

Panchayats. However, the article unearths significant differences in the impacts of decentral-isation between the two states.

The variations in outcomes between the two states are found to be linked to the architectureof decentralisation design as well as to non-statutory provisions that can create a process of

path-dependency towards achieving MDGs. The article also flags key methodological com-plexities inherent in the current MDG framework with respect to the actualization of thegoals of equity and access to primary health.

In September 2000, 147 heads of nations endorsed the Millennium Develop-ment Goals (MDGs) to address the worlds greatest developmental challenges

by the year 2015. Half these goals, either directly or indirectly, are concernedwith aspects of primary healthcare. These goals include: eradicating extremepoverty and hunger (Goal 1), reducing child mortality (Goal 4), improvingmaternal health (Goal 5), combating Human Immunodeficiency Virus/

Acquired Immune Deficiency Syndrome (HIV/AIDS), malaria and otherpreventable diseases (Goal 6) and environmental sustainability includingensuring sustainable access to safe drinking water (Goal 7).

http://jhm.sagepub.com/http://jhm.sagepub.com/http://jhm.sagepub.com/http://jhm.sagepub.com/


3/28

168 Nitya Mohan

Journal of Health Management, 11, 1 (2009): 167193

In the context of developing nations such as India, the public sector iscentral to the provision of most primary healthcare services relating to theMDGs.1 However, one of the main challenges to attaining the MDGs is thelack of capacity in the public sector to provide these basic services. Publichealth institutions, especially in rural areas, are neither universal in reachnor adequate in quality. For instance, the government healthcare sector inIndia has tended to adopt rigid, vertical approaches, which have not beenamenable to the needs of their intended beneficiaries. This is compoundedby structural obstacles to accessing healthcare particularly for poor women,

whose utilisation of services has historically been limited due to deep-setbiases, lack of acknowledgement of health needs, discriminatory attitudesin diagnosis and low levels of awareness which have negatively influencedtheir health-seeking behaviour (Koenig et al. 2000; Sen et al. 2007: 682).This has led to a wide gap between the providers of health services on theone hand and families and communities on the other.

One of the principal reasons for the lack of access to health services hasbeen the absence of mechanisms that enable those in need of the servicesto demand quality services, monitor their availability and supervise their

management. Democratic decentralisation is seen as one way of improvingequity, management, accountability and responsiveness of government healthservices through the reform of the public sector. Decentralisation institu-tionalises opportunities for citizens (particularly the poor and marginalised)to participate in and influence decision-making and resource allocation, em-powering them to become active agents of development (Oates 1972).2 Inaddition, the opening up of new political spaces at the local level is especiallyrelevant for women who have been traditionally excluded from decision-making processes (Manor 1999).

With particular reference to the healthcare sector, decentralisation isviewed as an important tool for implementing primary healthcare policiesby strengthening patient leverage, responsitivity and enhanced local servicedelivery (Green 1992). From this perspective, decentralisation is a meansof empowering local communities, facilitating multisectoral coordinationof activities at the local level, fostering community participation as well asencouraging participatory planning more attuned to local needs (Mills 1994;Taal 1993).3 In economics literature, one of the most commonly cited reasonsfor advocating health sector decentralisation has been the ability to address

variations in preferences across regions or population sub-groups for the

http://jhm.sagepub.com/http://jhm.sagepub.com/http://jhm.sagepub.com/


4/28

Democratic Decentralisation and the Millennium Development Goals 169


services that governments provide (Rondinelli et al. 1983). In the absenceof cost savings or inter-jurisdictional external effects, decentralisation canimprove collective societal welfare by providing a set of health goods andservices that are more sensitive to varying local needs than the provision of ahomogenous set of services by a central government (Mishra 2005).

Decentralisation reforms are thus an important means through which thegoals of human development, the main elements of which are articulatedthrough the MDGs, can be achieved (UNDP 2007). However, in the finalanalysis, the question of the success of decentralisation will depend on whetherequity and access to care, particularly for disadvantaged groups including

women has improved.4 This is critical to interpreting progress on overallhealth outcomes and indeed, to attaining the MDGs.

In the specific context of the targets for primary healthcare identifiedby the MDGs, this article evaluates the link between decentralisation andpositive health outcomes for women and children. Using India as a casestudy, the article traces the changes in health attainments relating to MDGtargets for these groups in rural India as a result of decentralisation reforms.

An analysis of health outcomes for women in the two southern states shouldprovide salient lessons with respect to the effectiveness of decentralisation inadvancing the MDGs.

The article is divided into three sections. The first section briefly sets outthe background to decentralisation in India and outlines the methodologyemployed by this article in the outcomes analysis. The next section examinesthe changes in health attainments for women and children with respect tofour selected MDGs, across a 15-year time period. The final section tracesthe conditions that facilitate decentralised systems to achieve the MDGs.This section also serves as a conceptual and methodological critique of thecurrent MDG framework with respect to the objective of gender equity.

Background and Methodology

This article evaluates progress in actualising MDGs for women and childrenas a result of decentralisation reforms. India makes a useful case study forevaluating whether decentralisation does indeed promote improved outcomesas it has a long tradition of emphasising and institutionalising local self-government and grassroots level democracy. In 199293, the Government

of India passed a series of constitutional reforms in order to democratise and



5/28

170 Nitya Mohan


empower a third tier of government at the sub-state level called PanchayatiRaj Institutions (PRIs). The reforms mark a transition from a two-tier sys-tem of governance (union and state) to a three-tier system, comprising theunion, the states and the Panchayats. This three-tier structure of rural localgovernment is based on the principles of democratisation, devolution ofpowers and resources for planning as well as community involvement in theimplementation of development programmes. Healthcare was one of thefirst subjects devolved to PRIs in India.

Geographically, the article focuses on two southern states in India(Karnataka and Kerala) because they are leading examples of innovation indecentralisation in India, albeit in two very different waysKarnataka wasone of the first states to institute decentralisation and Kerala has arguablybeen the most comprehensive in its decentralisation programme. Within thesegeographies, we look at the specific experience of decentralisation in rela-tion to health outcomes for rural women in India. Seven outcome indicatorscorresponding to four MDGs are evaluated. These goals are: reducing childmortality (Goal 4); improving maternal health (Goal 5); combating preven-tative diseases (Goal 6) and; ensuring environmental sustainability (Goal 7)(see Box 1).

In this study, the assessment of equity in health outcomes focuses on therelative as well as absolute gaps in the health status of women in the post-decentralisation context. By gaps, we mean differences in the health status ofa particular sub-section of the population relative to other sections, segregated(wherever possible) on the basis of three specific parameters (geography (rural/urban),5 religion and caste6). For the purpose of the analysis, the situation isregarded as equitable if there is a reduction in gaps between the parametersof segregation over time.

The outcomes analysis is based on secondary data (National Family Health

Survey rounds 1, 2 and 3) collected at three periods of time, one prior todecentralisation (199293), and the second and third after decentralisation(199899 and 200405 respectively).7 The comparison between three timeperiods of health outcomes is expected to help trace changes in the relativehealth status along a range of parameters of disaggregation against the back-drop of increased PRI involvement in the management and supervision ofhealthcare.

As a caveat, while the article recognises the difficulty of establishing a directand rigorous link between decentralisation and improvements in outcomes

and well-being, variations in outcomes observed are likely to be a part of



6/28

Box1

SelectedMDGs,Tar

getsandIndicators

Goals

Targets

Indicators

Goal4:Reducechildmortality

Reducebytwo-thirdsbetw

een1990and

2015theunder-fivemorta

lityrate

Under-fivemortalityrate

Infantmortalityrate

Goal5:Improvematernalhealth

Reducebythree-quartersbetween1990

and2015thematernalmortalityratio

Maternalmortalityratio

Proportionof

birthsattendedtobyskilledhealth

personnel

Goal6:Combatpreventativediseases

Havehaltedby2015andbeguntoreverse

thespreadofHIV/AIDS

Havehaltedby2015andbeguntoreverse

theincidenceoftuberculo

sisandother

majordiseases

Currentcontraceptiverate

Percentageof

populationwithtuberculosis

Goal7:Ensureenvironmentalsustain

ability

Halveby2015thepropor

tionofpeople

withoutbasicsanitation

Proportionof

populationwithaccesstoimproved

sanitation

Sou

rce:http://www.un.org/millennium

goals/;Self.



7/28

172 Nitya Mohan


a changing trend, attributable, at least in some measure, to the changingpolitical and economic context since decentralisation. The point is thus,to examine the trends in the health status of rural women in the two statesunder study, in the context of decentralisation of healthcare.

Findings

As highlighted in the previous section, this article is specifically concernedwith health outcomes and utilisation rates across a range of reproductive and

child health (RCH) indicators that are under the scope of decentralised gov-ernments and where local governments have a direct degree of responsibility.These indicators are now examined.

Child Mortality

With respect to measuring progress on targets for child mortality, two indi-cators have been used: under-five mortality rate and infant mortality rate.

Under Five Mortality and Infant Mortality8

A comparison of under-fivemortality rates (U5MR) across the three surveys indicates that there has beena substantial decline in mortality in both the states (see Table 1). For instance,in Karnataka the number of under-five deaths per 1,000 live births decreasedfrom 102 deaths in 199293 (NFHS 1) to 83 deaths in 199899 (NFHS 2)to 55 deaths in 200405 (NFHS 3). In Kerala too, there has been a dramaticdecline in U5MR across the surveys. NFHS 1 and NFHS 2 recorded IMRsof 40 and 26 respectively for Kerala. Comparing these estimates with theNFHS 3 estimate of 16 indicates that U5MR declined by 24 deaths per

1,000 live births over the survey period.

Table 1

Under Five Mortality Rate per 1,000 Live Births by State

(Figures in brackets indicate percentages)

U5MR NFHS 1 NFHS 2 NFHS 3

Karnataka 102 (10.2%) 83 (8.3%) 55 (5.5%)Kerala 40 (4%) 26 (2.6%) 16 (1.6%)

Sources: IIPS 1995 (various reports), IIPS 2000 (various reports), IIPS 2008 (various reports).



8/28



Table 2 presents infant mortality rates (IMR) for both the states acrosstime. In both Karnataka and Kerala, the IMR per 1,000 live births declineddramatically from NFHS 1 to NFHS 3. In Karnataka, IMR declined from75 deaths per 1,000 live births to 43 deaths across the 15-year period, whilein Kerala, IMR declined by nearly 50 per cent over the same period. Chart 1compares IMR and U5MR for India across the three surveys. As the trendin the chart indicates, early childhood mortality rates for India as a wholehave also declined over time.

Table 2

Infant Mortality Rate per 1,000 Live Births by State(Figures in brackets indicate percentages)

Total

IMR NFHS 1 NFHS 2 NFHS 3

Karnataka 75 (7.5%) 62 (6.2%) 43 (4.3%)Kerala 31 (3.1%) 21 (2.1%) 15 (1.5%)

Sources:IIPS 1995 (various reports), IIPS 2000 (various reports), IIPS 2008 (various reports).

Chart 1Early Childhood Mortality Rates across NFHS 1, NFHS 2 and NFHS 3: India

Sources:IIPS 1995 (various reports), IIPS 2000 (various reports), IIPS 2008 (various reports).Note: IMR and U5MR are represented per 1,000 live births.



9/28

174 Nitya Mohan


Maternal Health

With respect to measuring progress on maternal health, two indicators havebeen used: maternal mortality ratio and proportion of births attended to byskilled health personnel.9

Maternal Mortality RateThe maternal mortality rates (MMR) for thetwo states as well as for India as a whole are presented in Table 3.10 Whereasnationally and for Kerala, MMR has been declining over time, in Karnatakaprogress has been slow. In Karnataka, the MMR increased quite dramaticallyfrom 195 deaths per 10,000 live births in 1998 to 266 deaths in 19992001.

Although the MMR came down to 228 in 200103, this figure is still higherthan the 1997 estimates indicating that progress is tardy. Kerala on the otherhand had an MMR of 110 deaths per 10,000 live births (200103), whichis half the MMR in Karnataka.

Table 3

Maternal Mortality Rate per 10,000 Live Births: State-wise and India

State-wise Maternal Mortality Rate

(1997, 1998, 19992001 and 200103)

States 1997 1998 199901 200103

India 408 407 327 301Karnataka 195 195 266 228Kerala 195 198 149 110

Sources: Ministry of Statistics and Programme Implementation, Government of India; Dir-ectorate General of Health Services, Ministry of Health & Family Welfare, Governmentof India (2001).

Assistance in Delivery With respect to the indicator of assistance in deliveryby trained personnel, we see improvements across all parameters of segregationin both the states over time. As Tables 4 and 5 indicate, in Karnataka acrossall religious and caste groups, the proportion of rural women who wereassisted by doctors at the time of delivery increased quite dramatically fromNFHS 1 to NFHS 3. For instance, among Hindu women, the proportionof doctor assisted deliveries increased by 125 per cent, while for Muslim

women, the corresponding figure was 138 per cent over the 15-year period.Among SC/ST women, the proportion of doctor assisted deliveries went up



10/28

Table4

AssistanceinDeliverybyReligion:RuralKarnataka(inpercent)

Karnataka

Hindu

Muslim

Other

NFHS1

NFHS2

NFHS3

NFHS1

NFHS2

NFHS3

NFHS1

NFHS2

NFHS3

Doctor

20

28

45

21

23

50

44

46

68

AN

M/Nurse/

M

idwife/LHV

19

19

14

22

22

14

13

19

21

Other

60

53

41

58

54

36

44

35

11

Tot

al

100

100

100

100

100

100

100

100

100

Sou

rces:

IIPS1995(Karnataka),IIP

S2000(Karnataka),IIPS2008(Ka

rnataka).

Notes:ANM=Auxiliarynursemidwife;LHV=Localhealthvolunteer.



11/28

176 Nitya Mohan


from 11 per cent in NFHS 1 to 35 per cent in NFHS 3. Further, gaps be-tween SC/STs and the better-off other castes declined dramatically betweenNFHS 1 and NFHS 3.

Finally, the proportion of assisted deliveries by midwives and nurses de-clined slightly across the three surveys but these results must be interpretedin light of the sharp increases in assistance by doctors. What is also notableis that the proportion of deliveries that were assisted by untrained personnel(indicated by the category other) declined over the 15-year period acrossall caste and religious sub-groups.

As Tables 6 and 7 indicate, in rural Kerala too, remarkable progress hasbeen made in terms of the increase in the proportion of assisted deliveries bydoctors. This is true across all religious and caste sub-groups. For instance,the proportion of deliveries by doctors for Muslim women increased by62 per cent between NFHS 1 and NFHS 3 and among SC/ST women,the corresponding ratio went up from 63 per cent to 88 per cent. Gapsbetween Hindus and Muslims as well as SC/STs and other castes also de-clined substantially between NFHS 1 and 3. Another prominent trendis the reduction in the proportion of deliveries conducted by unqualifiedpersonnel. As Table 6 indicates, 17 per cent of Muslim women were assistedby unqualified personnel in NFHS 1. By NFHS 3, this proportion had comedown to 0 per cent, indicating the huge strides taken toward safe motherhoodstrategies in the state.

Preventative diseases

This article utilises two indicators to measure progress on preventative dis-eases: current contraceptive use11 and proportion of population sufferingfrom tuberculosis.

Table 5

Assistance in Delivery by Caste: Rural Karnataka (in per cent)

SC/ST Other

Karnataka NFHS 1 NFHS 2 NFHS 3 NFHS 1 NFHS 2 NFHS 3

Assistance in deliveryDoctor 11 17 35 23 32 51

ANM/Nurse/Midwife/LHV 13 17 14 21 20 14Other 76 67 51 56 48 35Total 100 100 100 100 100 100

Sources:IIPS 1995 (Karnataka), IIPS 2000 (Karnataka), IIPS 2008 (Karnataka).



12/28

Table6

AssistanceinDeliverybyReligion:RuralKerala(inpercent)

Hindu

Muslim

Other

Kerala

NF

HS1

NFHS2

NFHS3

NFHS1

NFHS2

NFHS3

NFHS1

NFHS2

NFHS

3

Doctor

84

94

94

60

87

97

92

96

96

AN

M/Nurse/Midwife/LHV

10

3

5

24

3

2

5

3

4

Other

6

3

1

17

11

0

3

2

0

Tot

al

100

100

100

100

100

100

100

100

100

Sou

rces:IIPS1995(Kerala),IIPS2000(Kerala),IIPS2008(Kerala).



13/28

178 Nitya Mohan


Current Contraceptive UseAn important indicator of the reproductivehealth of women is the use of contraception. What is remarkable is thatthe three rounds of NFHS data indicate that dramatic improvements incontraception rates (modern method, any method) have been achieved inboth the states (Tables 8, 9, 10, 11). The proportion of women not using anycontraception also declined in the two states across the different parametersof segregation.

It is noteworthy that across SC/ST households in Karnataka, the proportion

of households using modern forms of contraception increased from 43 percent (NFHS 1) to 65 per cent (NFHS 3). Gaps between disadvantaged SC/SThouseholds and other households using modern forms of contraception alsodramatically decreased between the survey periods indicating greater equityin outreach services (Table 9). With respect to religion as the parameter ofsegregation, we find that the gaps between Hindu and Muslim householdsusing modern contraception have also declined (Table 8).

In Kerala, the figures for any contraceptive use among SC/ST householdsincreased marginally from 73 per cent (NFHS 1) to 75 per cent (NFHS 3)

(Table 11). What is interesting in the Kerala case is that the proportion ofwomen from the disadvantaged SC/ST castes that used contraception washigher than more advantaged other castes, flagging the success of targetedfamily planning programmes. Consequently, gaps in equity among castegroups in the post-decentralisation period are positively biased towardsSC/STs. Given that 67 per cent of rural users in Kerala obtain their contra-ceptives from the public health sector (NFHS 2 Kerala), the high rates ofcontraception indicates that needs for family planning services are beingmet effectively by the network of PHCs, sub-centres and medical staff in

conjunction with PRIs.

Table 7

Assistance in Delivery by Caste: Rural Kerala (in per cent)

SC/ST Other

Kerala NFHS 1 NFHS 2 NFHS 3 NFHS 1 NFHS 2 NFHS 3

Doctor 63 82 88 77 92 97ANM/Nurse/Midwife/

LHV

21 8 8 10 2 3

Other 16 10 4 13 6 0Total 100 100 100 100 100 100

Sources: IIPS 1995 (Kerala), IIPS 2000 (Kerala), IIPS 2008 (Kerala).



14/28

Table8

CurrentContraceptiveUsebyRelig

ion:RuralKarnataka(inpercent)

Hindu

Muslim

Other

Karnataka

NFHS

1

NFHS2

NFHS3

N

FHS1

NFHS2

NFHS3

NFHS1

NFHS2

NFHS3

Anymethod

49

59

66

34

42

56

45

54

59

Anymodernmethod

48

58

66

33

42

56

42

51

58

Anytraditionalmethod

1

1

1

1

1

1

3

3

2

Notusinganymethod

51

41

34

66

58

44

55

46

41

Tot

al

100

100

100

100

100

100

100

100

100

Sou

rces:IIPS1995(Karnataka),IIPS2000(Karnataka),IIPS2008(Karn

ataka).



15/28

Table9

CurrentContraceptiveUsebyCaste:RuralKarnataka(inpercent)

SC/ST

Other

Karnataka

NFHS1

NFHS2

NFHS3

NFHS1

NFHS2

NFHS3

Anymethod

43

52

66

49

59

65

Anymodernmethod

43

51

65

47

58

65


1

1

1

1

1

1

Notusinganymethod

57

48

34

51

41

35

Tot

al

100

100

100

100

100

100

Sou




16/28

Table10

CurrentContraceptiveUsebyReligion:RuralKerala(inpercent)

Hindu

Muslim

Other

Kerala

NFHS

1

NFHS2

NFHS3

NFHS1

NFHS2

NFHS3

NFHS1

NFHS2

NFHS3

Anymethod

71

72

75

36

47

53

71

72

78

Anymodernmethod

63

65

65

31

41

44

59

61

58


8

7

10

6

6

9

12

12

20

Notusinganymethod

29

28

25

64

53

47

29

28

22

Tot

al

100

100

100

100

100

100

100

100

100

Sou




17/28

182 Nitya Mohan


While examining the data by religion we note that there have beensharp increases in the use of any method of contraception among differentreligious sub-groups (Table10). This trend is particularly visible amongMuslim women. More crucially, gaps in equity between Hindu and Muslimhouseholds using any contraception declined sharply between NFHS 1 andNFHS 3, which is indicative of a positive trend.

Tuberculosis Using NFHS data, the experience of morbidity in termsof preventable diseases such as tuberculosis (TB) has been presented inTable 12. Due to the small size of the NFHS sample and the difficulties indisaggregation, only state-wise data is presented here. We note that a higherproportion of women in Kerala have TB relative to women in Karnataka.However, in both the states there has been a decline in TB prevalence ratesfrom NFHS 1 to NFHS 3, which is indicative of a positive trend.

Table 12

Prevalence of Tuberculosis by State across Surveys (in per cent)

Tuberculosis NFHS 1 NFHS 2 NFHS 3

Karnataka 1 3 1Kerala 6 5 3

Sources: IIPS 1995 (various reports), IIPS 2000 (various reports), IIPS 2008 (various reports).

Environmental Sustainability

To evaluate progress on environmental sustainability, this section probes the

level of basic sanitary facilities in the two states. The provision of sanitation is

Table 11

Current Contraceptive Use by Caste: Rural Kerala (in per cent)

SC/ST Other


Any method 73 74 75 61 62 67Any modern method 69 69 68 52 54 56Any traditional

method

5 6 7 8 8 12

Not using any method 27 26 25 40 38 33Total 100 100 100 100 100 100




18/28



directly within the realm of Panchayat responsibility in India. Improvementsin water sources and sanitation facilities are critical as they have a significantinfluence on the health of household members, especially children (NFHS 2,Karnataka).12 Lack of access to facilities is a particular problem for girls and

women because of issues of privacy and safety.

Sanitation As indicated by the NFHS data for Karnataka, there has beena marked increase in the percentage of households with improved toiletfacilities between the three rounds for all parameters of segregation (Tables 13and 14).13 Conversely, there were also reductions in the proportion of house-holds with non-improved facilities across all groups between NFHS 1 and 3.

Among Muslims, the change in the proportion of households having im-proved toilets was remarkablefrom 12 per cent households in NFHS 1to 35 per cent in NFHS 3 (Table 13). Considering the data by caste, theproportion of SC/ST households with improved sanitation increased from2 per cent to 10 per cent (Table 14). However, differences within caste andreligious sub-clusters were very high, indicating variations in equity betweendifferent groups in Karnataka.

The NFHS data also highlights the great strides that have been madein Kerala in terms of construction of sanitary facilities (Tables 15 and 16).

Across all religion and caste groups, there has been a substantial increased inimproved facilities. For instance, among caste groups, the proportion of SC/ST households with improved facilities doubled between NFHS 1 and 3.

Further, gaps in equity in terms of proportion of households possessingownpucca toilets narrowed among caste and religious sub-categories. Wealso find strictly decreasing proportions of households with non-improvedfacilities across all parameters between the three rounds.

Conclusion

Within the context of womens empowerment, engagement through localgovernments is the first step towards greater democratisation and conse-quently towards transformative outcomes for vulnerable communities. Acomparison of secondary data over time indicates that decentralisation ofpublic healthcare responsibilities to local governments in India has resultedin augmenting health (and health-related) outcomes for rural women. An

observation clearly discernible from the data is a reduction in inequity over



19/28

Table13

H

ouseholdSanitationFacilitiesby

Religion:Karnataka(inpercent)

Hindu

Muslim

Other

Karnataka

NFHS1

N

FHS2

NFHS3

NFHS1

NFHS2

NFHS3

NFHS1

NFHS2

NFHS3

Improved

6

12

19

12

8

35

33

52

4

4

Non-improved

94

88

81

88

92

65

67

48

5

6

Other

0

0

0

0

0

0

0

0

0

100

100

100

100

100

100

100

100

10

0

Sou


ataka).



20/28

Table14

HouseholdSanitationFacilitiesb

yCaste:Karnataka(inpercent)

SC/ST

Other

Karnataka

NFHS

1

NFHS2

NFHS3

NFHS1

NFHS2

NFHS3

Improved

2

5

10

8

16

24

Non-improved

98

95

90

92

84

76

Other

0

0

0

0

0

0

100

100

100

100

100

100

Sou


ataka).



21/28

Table15

HouseholdSanitationFacilitiesbyReligion:Kerala(inPercent)

H

indu

Muslim

Other

Kerala

NFHS1

N

FHS2

NFHS3

NFHS1

NFHS2

NFHS3

NF

HS1

NFHS2

NFHS3

Improved

60

76

93

68

83

98

71

86

94

Non-improved

40

24

7

33

17

2

29

14

5

Other

0

0

0

0

0

0

0

0

1

100

100

100

100

100

100

100

100

100

Sou




22/28



a majority of selected indicators for both Kerala and Karnataka. Further, thedata indicates that the proportion of households suffering from preventa-tive illnesses such as tuberculosis as well as household attainments of RCHcare including maternal and child mortality rates, contraceptive use and safedeliveries have increased between the survey periods across all the parametersof segregation. There have also been substantial improvements in trends inhealth infrastructure particularly in terms of sanitation at the village level asa result of decentralisation across both states.

However, despite the success of decentralisation reforms, there is muchground to be covered in terms of achieving MDG targets by 2015. This isvery obvious in the case of Karnataka. For instance, U5MR in Karnatakadecreased by 46 per cent between 199293 and 200405 but by 2015,it must reduce by 67 per cent. On the other hand in Kerala, U5MR hasdeclined substantially by 60 per cent since 199293. MMR in Karnatakaincreased dramatically by 17 per cent between 1997 and 2003. However,by 2015, it must reduce by nearly 75 per cent. In Kerala, MMR declined by44 per cent in the same period. Finally, the 7th MDG requires that by 2015

the proportion of people without basic sanitation should be halved. Whileprogress has been made in this direction in both the states, the data indicatesthat in Karnataka there is still a long way to go to realise this target.

Thus, a striking paradox that emerges from the data is that Keralaperforms much better in terms of achieving MDG targets than Karnataka.This is surprising in light of Karnatakas institutional experience vis--visdecentralisation.14 Yet, decentralisation has had a more muted effect on

womens health outcomes in Karnataka than in Kerala. How can we explainthe differential impacts in the two states?

Table 16

Household Sanitation Facilities by Caste: Kerala (in per cent)

SC/ST Other


Improved 44 68 83 66 81 96Non-improved 55 32 17 35 19 4Other 0 0 0 0 0 0

100 100 100 100 100 100




23/28

188 Nitya Mohan


The Kerala model of decentralisation indicates that the success of suchreforms is linked to both the architecture of the decentralisation design(institutional mechanisms and other features of design) as well as to non-statutory provisions such as a vibrant civil society, levels of literacy andawareness that create a sustainable process of path-dependency towardsemancipatory opportunities for women. We now turn our attention brieflyto these factors.

The Kerala Model

As the experience of decentralisation in Kerala suggests, the effective design ofdecentralised systems is a necessary prerequisite for the success of such systems.In order to render decentralisation more effective, consideration of key factorsrelating to the empowerment of Panchayats are critical. These factors includea greater degree of financial and administrative decentralisation to the locallevel along with a concomitant strengthening of channels of accountabilityand transparency in local government functioning. Such factors are essentialto sustain positive outcomes from decentralisation, provide safeguards for themarginalised and lead to improved outcomes. In Karnataka, many of these

factors are as yet inchoate.Keralas success with decentralisation has also been abetted by the inter-

action of a multiplicity of enabling social factors. One of the salient featuresof Keralas development experience is the role of education. The success of theliteracy movement and the states active interest in promoting universal access

was instrumental in spreading education throughout Kerala and ensuringgender parity in school enrolments. The spread of education catalysed anincreased level of social consciousness in society. This resulted through publicdemand and supply, in wider access to and awareness of healthcare amongthe population.15

Other landmarks in the development history of Kerala include powerfulcaste and social reform movements in the early 20th century, radical landreforms, the vast public distribution of food networks as well as a host of socialsecurity and welfare measures. This commitment has paid off in dramaticand lasting improvements in the quality of life for its rural people (Dreze andSen 1995). This has bestowed a measure of economic freedom, especiallyamong socially marginalised communities and has led to an enhancementin their overall well-being.



24/28



Thus, Keralas steady progress towards MDG targets can be interpretedto be a result of the state having given higher priority to the development ofsocial services in response to organised public demand. All of these conditionsprovided a fertile setting for the institutionalisation of decentralisation thatthe state undertook in the 1990s. In this sense, public action and socialmobilisation have ensured path dependency towards increased political par-ticipation and improved developmental outcomes in Kerala. This has led toan environment that has fostered citizenship opportunities for disadvantagedgroups, particularly women.16 With specific reference to healthcare, an in-crease in participation has resulted in improved needs assessment, increasedaccountability of personnel and enhanced access and quality of healthcarein Kerala. Indeed, Keralas relative success in decentralisation bears out theimportance of social mobilisation and womens education as catalysts forpath dependency and key strategies for achieving strategic improvementsin well-being for women. It is worth acknowledging that these factors (par-ticularly womens literacy) are recognised as contributing to improved healthand well-being in the MDGs in the goals for womens equality and universalprimary education.

However, lessons from the Kerala model of decentralisation underscoreseveral important points and serve as a critique of the MDG framework.First, gender equity is derived from both political and social spheres. Womenseffective citizenship is a combination of political rights as well as enablingsocial conditions that facilitate participation in the public sphere (Yuval-Davis 1997). Because womens responsibilities are traditionally perceived aslying in the private sphere (family and caring roles) and mens gender rolesas being related to decision-making in the public domain, women are oftenexcluded from the realm of public activity. This public/private divide, basedon the classical theories of the social contract, contributes to relationships ofinequality and difference between women and men. None of the MDGs reallyaddress the issue of womens citizenship and effective participation. WhileGoal 3 flags the proportion of sets held by women in national government,the indicators do not acknowledge the importance of womens participationand decision-making rights in the multiple contexts of the household, thecommunity and the state.

What the Kerala model also highlights is the importance of integratedgender-responsive policies that explicitly target excluded communities andserve to reduce gender and class based disparities in society. In this sense,



25/28

190 Nitya Mohan


gender equality is not just a goal in its own right but is central to achievingall other goals (UNDP 2003). Yet, of the eight MDGs, only the 3rd and 5thgoals refer directly to women and (indirectly) to gender relations. Given thatthe MDGs are mutually reinforcing, it is critical to make the gender dimen-sion more explicit in the targets and indicators of the MDGs. Not addressingthe gender issue in an integrated manner is one of the main weaknesses ofthe MDG framework.

Third, under the MDG framework there is an overwhelming emphasison the achievement of targets. However in the process, it is quite easy to lose

sight of the main objectives. For instance, one of the indicators under Goal 5(maternal health) is the proportion of deliveries undertaken by trained birthattendants. Yet, the quality of that care matters as much as the proportion ofassisted deliveries. Thus, achieving the MDGs is not simply about trackingprogress but is inextricably linked to developing sensitive and sustainableapproaches to accomplishing the goals.

Finally, it is important to acknowledge the role of gender relations atall levels (household, community, state) in influencing the achievement oftargets. Gender relations may be perceived to be largely socially imposed as

opposed to being biologically determined. In India (and indeed in manyparts of the world), these relations sustain and reproduce a hierarchy wherewomen are subordinate to men. For gender equity to take root, developmentpolicies must be cognisant of the ideologies and exclusions underpinninggender relations in society. Such policies must be both gender responsive andgender sensitive as the Kerala experience has taught us.

Notes

1. For instance, it has been estimated that 90 per cent of immunisations and 60 per cent of

prenatal care is publicly provided in India (IIPS 2000). Further, deprived sections of the

population like poor women are highly reliant on public health services in rural areas where

private services are either unavailable or are unaffordable.

2. There are two main discourses underpinning decentralisationthe empowerment andefficiency frameworks. The empowerment discourse posits that decentralisation offers an

enabling context to empower citizens by institutionalising processes that catalyse collective

action, agency and social change (Pateman 1970; UNDP 2004). This in turn leads to

broader emancipatory impacts for citizens in terms of the realisation of outcomes for their

well-being. Another rationale is the neo-liberal efficiency argument (Bird 1999). According

to this argument, decentralisation is a means to minimise wasteful spending and encourage

fiscal efficiency, facilitating cost-recovery through mechanisms such as privatisation. Since



26/28



the MDGs are directly relevant to issues of equity and empowerment, this article anchors

its analysis in the empowerment framework.3. Some non-country specific studies draw attention to the potential advantages of decen-

tralised health systems, including improved accountability and logistic systems as well as

a greater degree of resource efficiency (Bossert 1998; Jimnez and Smith 2005).

4. This is particularly relevant for certain services (for instance reproductive and child

healthcare) where gender bias at the local level can create allocational inefficiencies that

result in worsening gender outcomes. As a caveat, analysing health inequities through a

gender lens is problematic. This is because unlike inequalities in education, health has a

physiological basis or at least biological referents. However, more often than not, social

determinants are more instrumental in aggravating gender based health asymmetries than

physiological factors. This article takes the view that inequities in health are for the most

part, socially produced, rather than biologically given. Thus, they can be ameliorated by

changes in the power structures underpinning societal relations.

5. Although the data is segregated according to geography, only rural data is shown. The

rural bias is justified in that over 70 per cent of Indian citizens (and voters) reside in rural

India.6. This article does not consider the parameter of class in its analysis. Although this is a critical

dimension, class-wise data could not be obtained from NFHS 1 and hence comparison

over time was not possible. However, in the Indian context, caste hierarchies are pervasive

and the caste groups at the lower end of the hierarchy (denoted as Scheduled Castes and

Tribes, i.e., SC/STs) suffer from severe social and economic discrimination. Thus, the

category of caste can be taken to approximately reflect class differences.

7. At the outset, it must be said that these surveys were designed to present state and national-

level estimates of demographic trends and were not specifically meant to provide information

on utilisation and health outcomes. However, because they represent a continuum of

information across a 15-year period, these sources were used.

8. For both IMR and U5MR, data could not be disaggregated by caste, religion or geography

due to the small size of the sample in both the states. Hence, only state-wise and national

estimates for these two indicators have been presented.

9. These two specific indicators have been selected because they are interlinked. Indeed, the

proportion of births attended to by trained personnel has a direct implication on maternal

mortality rates.

10. Again for MMR, data could not be disaggregated by caste, religion or geography due to the

small size of the sample in both the states. Hence, only state-wise and national estimates

for this indicator have been presented.

11. The indicator current contraceptive use has been employed as an approximate indicator

of HIV knowledge and safe-sex practices in this article. NFHS 3 is the first national survey

in India to provide HIV estimates through testing. Since this was not done in NFHS 1

and 2, the results from round 3 cannot be compared with the earlier rounds.

12. It is estimated that the poor quality of water and sanitation resources accounts for about

10 per cent of the disease burden in developing countries (Government of Karnataka

2001).

13. Improved toilet facilities include facilities with a flush/ pour flush connected to a sewer

system, septic tank or pit latrine, a ventilated latrine, a biogas latrine and a twin pit, com-posting toilet. If a household has any of these types of facilities but shares them with other



27/28

192 Nitya Mohan


households, it has a non-improved facility. This category also includes households without

any facility (NFHS 3).14. Local self-governance has had a long history in Karnataka and the states experience with

respect to decentralisation and Panchayati Raj reforms has been significant (Nataraj and

Anantpur 2004).

15. Scholars concur that one of the most important factors behind Keralas remarkable perform-

ance in reducing fertility is the high level of female education (Bhat and Rajan 1990).

16. Besides mandating a one-third reservation of seats for women, the planning process in

Kerala also provides for a Womens Component Plan comprising 10 per cent of the local

governments budget.

References

Bhat, P.N.M. and S.I. Rajan (1990). Demographic transition in Kerala revisited. Economicand Political Weekly, 25(35/36), 195780.

Bird, R. (1999). Rethinking sub national taxes: A new look at tax assignment. IMF Working

Paper WP/99/165. Washington D.C. International Monetary Fund.Bossert, T. (1998). Analyzing the decentralization of health systems in developing countries:

Decision Space, Innovation and Performance. Social Science and Medicine, 47(10),151327.

Dreze, J. andA. Sen (1995). India: Economic development and social opportunity. New Delhi:

Oxford University Press.Government of India. (2001). Final report of the task force on health and family welfare.

Bangalore, Government of Karnataka.Green, A. (1992).An introduction to health planning in developing countries. Oxford: Oxford

University Press.Indian Institute for Population Sciences (IIPS) (2000). National family health survey

summary, Karnataka, Kerala 198899. Mumbai: Indian Institute for Population

Sciences.

(2008)a. National family health survey summary, India 199899. Mumbai, Indian

Institute for Population Sciences.

(2008)b. National family health survey summary, Karnataka 199899. Mumbai, Indian

Institute for Population Sciences. (2008)c. National family health survey summary, Kerala 199899. Mumbai, Indian

Institute for Population Sciences.

Jimnez, D. and P. Smith (2005). Decentralisation of healthcare and its impact on health

outcomes. York: Health Economics Data Group, University of York.Koenig, M., G. Foo and K. Joshi (2000). Quality of care within the Indian family welfare

programme: A review of recent evidence. Studies in Family Planning, 31(1), 118.Manor, J. (1999). The political economy of democratic decentralization. Washington D.C.:

World Bank.Mills, A. (1994). Decentralization and accountability in the health sector from an international

perspective: What are the choices? Public administration and development, 14,

28192.



28/28



Mishra, U.S. (2005). Understanding health inequity in decentralized health systems of Kerala

State, India (mimeo). Boston.Nataraj, V.K. and K. Anantpur (2004). Delegation to devolution: Karnataka. Chennai: Madras

Institute of Development Studies.Oates. W.E. (1972). Fiscal federalism. New York: Harcourt Brace Jovanivich.Pateman, C. (1970). Participation and democratic theory. Cambridge, M.A.: Cambridge

University Press.Rondinelli, D., J. R. Nellis and G.S.Cheema(1983). Decentralization in developing countries:

A review of recent experience. Washington, D.C.: World Bank.Sen, G., A. Iyer andA. George (2007). Systematic hierarchies and systematic failures: Gender

and health inequities in Koppal district. Economic and Political Weekly, XLII(8),68290.

Taal, H. (1993). Decentralization and community participation for improving access to basic ser-vices: An empirical approach. Florence: UNICEF.

UNDP (2003). Human development report. New York: United Nations. (2004). Decentralized governance for development. New York: United Nations. (2007). Improving local service delivery for the MDGs in Asia: Pilot of methodology

for support to national policy and MDG strategies. Available at http://regional

centrebangkok.undp.or.th/practices/governance/decentralization/documents/

WorkshopReport.pdf (Accessed on 20 November 2007).

UNDP, Millenium Development Goals. Available at http://www.un.org/millenniumgoals/

(Accessed on 20 November 2007).Yuval-Davis, N. (1997). Gender and nation. London, Thousand Oaks and New Delhi, Sage

Publications.

Nitya Mohan is Director, Nand and Jeet Khemka Foundation. E-mail:n.khemka@khemkaf

oundation.org

Decentralization and Health Outcomes

Documents