research article explaining sex differentials in child mortality in...

Research ArticleExplaining Sex Differentials in Child Mortality in IndiaTrends and Determinants

Shrikant Kuntla1 Srinivas Goli2 and Kshipra Jain3

1 International Institute for Population Sciences Govandi Station Road Deonar Mumbai Maharashtra 400088 India2Department of Development Studies Giri Institute of Development Studies Sector 0 Aliganj Housing Scheme Lucknow 226024 India3 Department of Economics University of Rajasthan Jaipur Rajasthan 302004 India

Correspondence should be addressed to Srinivas Goli sirispeaks2ugmailcom

Received 13 April 2014 Accepted 9 October 2014 Published 6 November 2014

Academic Editor Sally Guttmacher

Copyright copy 2014 Shrikant Kuntla et alThis is an open access article distributed under the Creative CommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

This study has twofold objectives (1) to investigate the progress in sex differentials in child mortality in India in terms of withinand between group changes and (2) to identify the factors explaining the sex differentials in child mortality and quantify theirrelative contributions We have used three rounds of the National Family Health Survey (NFHS) data 1992 to 2006 Life tableapproach and Pyatt and Oaxaca decomposition models were used as methods of analyses The results revealed that though sexdifferential in child mortality is still high in India it declined during 1992 to 2006 (Gini index from 036 to 024) This declinewas primarily led by a change in within inequality of female child mortality (Gini index from 018 to 014) Among the selectedpredictors breastfeeding (40) birth order (24) antenatal care (9) and motherrsquos age (7) emerged as critical contributorsfor the excess female child mortality in India From the findings of this study we suggest that any efforts to do away with genderdifferences in child survival should focus more on within female child disparity across different population subgroups alongsidemale-female disparity Implications are advanced

1 Introduction

In most populations female mortality rates are lower thanmale mortality rates and females are more than maleswhich remain consistent across all the ages [1] Howeverhistorically South Asia is known for its male skewed childsex ratio Researchers have attributed sex differentials in childmortality as one of the primary factors contributing to itsskewed sex ratios in 0ndash6 year population Indiarsquos SampleRegistration System (SRS) has been indicating the presenceof sex differentials in child mortality since 1970s The recentSRS report has also indicated a large gap (9 per 1000 livebirths) in terms of under-five mortality rate among malesand females which further increases in rural areas [2] Whileall the Indian states show excess female under-five mortalitythe discrimination is huge in Jharkhand Chattisgarh UttarPradesh Rajasthan and Bihar (Figure 1) Figure 1 suggeststhat despite the explicitly oriented efforts to eliminate healthinequalities gender disparities in childhood mortality stillpersists in India The issue of sex differentials in child

mortality has remained a major hurdle for achieving genderequity and Millennium Development Goals-4 (MDG-4) inIndia

There are plenty of studies that have attempted to explainsex differentials in child mortality and factors associatedwith it specifically in the case of India [3ndash17] The thrust ofthese studies had been about the level of gender disparityin child survival nutrition and health care of children andfactors associated with it According to Das Gupta [6] excessfemale child mortality is an inherent part of family buildingstrategy where girls are considered as a burden and boys asresources Various studies [6ndash8 18ndash20] have revealed parentaldiscrimination in the provision of basic food and health careagainst their own daughters which is typically understood asone of the causes for excess female child mortality Furtherabusing the girl child within the household is also a commonpractice [19] Even though a number of policy interventionsare in place the persistent excess of female child mortalitycontinues to speculate about the authentication of reportedpredictors and their volume of contributions [21ndash31]

Hindawi Publishing CorporationInternational Journal of Population ResearchVolume 2014 Article ID 649741 7 pageshttpdxdoiorg1011552014649741

2 International Journal of Population Research

102030405060708090

Indi

aA

ndhr

a Pra

desh

Ass

amBi

har

Chat

tisga

rhD

elhi

Guj

rat

Har

yana

Him

acha

l Pra

desh

Jam

mu

Kash

mir

Jhar

khan

dKa

rnat

aka

Kera

laM

adhy

a Pra

desh

Mah

aras

htra

Odi

sha

Punj

abRa

jstha

nTa

mil

Nad

uU

ttar P

rade

shW

est B

enga

l

MaleFemale

per1

000

live b

irths

Und

er-fi

ve m

orta

lity

Figure 1 Under-five mortality rates by sex of the child India andstates 2010

Recently some studies [27ndash34] have tried to provideinsights on existing inequalities but there is no evidence ofdecomposition of the relative proportional contribution ofsocioeconomic and demographic factors in sex differentialsin child mortality The lack of information on the exactrelative contribution of different socioeconomic and demo-graphic predictors of excess female child mortality obscuresthe planning process that can reduce gender disparity inchild mortality Thus it is imperative to estimate the relativeproportional contribution of predictors as it allows the policymakers to prioritize the factors and consequent interventionsto bring balance in survival chances of male and femalechildren in India The other question which needs ourattention is whether the reported sex differentials in childmortality consist of only the difference between male-femalechildmortality or withinmale and female differences in childmortality are also contributing to it Hence decompositionof between and within male-female differences in childmortality are also important in policy perspective With thisrationale this study has twofold objectives (1) to investigatethe progress in sex differentials in child mortality in Indiain terms of between and within group changes and (2) toidentify the factors explaining the sex differentials in childmortality and quantify their relative contributions

2 Methods

21 Data Source Data from the three rounds of the NationalFamily Health Survey (NFHS) 1992ndash2006 have been usedto assess the trends in child mortality stratified by sexThe NFHS is an equivalent of worldwide DemographicHealth Survey (DHS) and a widely accepted source of datafor estimating demographic and public health indicators inIndia Three rounds of NFHS have been conducted by theInternational Institute for Population Sciences (IIPS) andMacro International under the aegis of theMinistry of Healthand FamilyWelfare IndiaTheNFHS collects information on

fertility mortality morbidity and maternal and child healthA similar sampling scheme was adopted in all the threerounds and provides sufficiently large sample sizes (morethan 90000 households were interviewed in each round) tocarry out analysis at the national as well as the state levelThe data were collected using similar interview schedulesand the household and eligible female informant responserates were consistently above 90 across the three roundsTo make the estimates representative and comparable acrossthe three survey rounds and to account for the multistagesampling design adopted in the three rounds of the NFHSwe used appropriate weights in the analysis The detailsof the sampling weights are given in NFHS reports of thevarious rounds For details on sampling see IIPS and MacroInternational 1992-93 1998-99 and 2005-06 [21ndash23]

22 Analysis Life table approach and Pyatt and Oaxacadecomposition models were used as methods of analyses

Life-Table Approach Child mortality rates were estimatedbased on the simple life table approach using STATA 110(Table 2) This approach builds up a 12ndash59 month matrix ofexposure and deaths for each month that the child was alivebetween the first and fifth birthday of life Child mortalityrates are calculated using the following steps

(1) We calculated the date of birth of the child as a centurymonth code (cmc) denoted by bcmc

(2) We calculated the date of death of the child as acentury month code (dcmc) In this process age atdeath in months is calculated for all deaths imputingthe age in months

(3) After tallying the deaths and exposure for eachmonth the probability of dying at exact age 119909 (119902

119909)

is calculated by dividing the deaths at age 119909 by theexposure at age 119909 for each month between first andfifth birthday of life

(4) We calculated the probability of surviving to age 119909 as119897119909= 119897119909minus 1 lowast (1 minus 119902

119909) with 119897

119900= 1

(5) We calculated the probability of dying by age 119909 as 1 minus119897119909

where 119902 is the probability of dying and 119897 is the probability ofsurviving

Pyattrsquos Gini Decomposition Model Pyatt [24] has given thedecompositionmodel ofGini coefficient Gini indexwas usedto calculate the change in inequality in childmortality amongthe male-female population of Indian states during 1992ndash2006 Further the Gini index in child mortality among themale-female population across Indian stateswas decomposedinto between group inequality within group inequality andinequality due to group overlapping

Let a population of ldquo119899rdquo individuals with mortality vector(1199101 1199102 1199103 119910

119899) and mean mortality 119910 be desegregated in

ldquo119896rdquo subgroups with 119899 = sum119896119895=1119899119895and subgroup mean is 119910

119895

International Journal of Population Research 3

The Gini index between subgroups 119895 (male) and ℎ(female) can be expressed as

119866119895ℎ=

1

119899119895119899ℎ(119910119895+ 119910ℎ)

119899119895

sum119894=1

119899ℎ

sum119903=1

10038161003816100381610038161003816119910119895119894minus 119910ℎ119903

10038161003816100381610038161003816 (1)

If 119865(119910) is the cumulative distribution function of mortal-ity then expected mortality difference between group 119895 and ℎcan be defined as

1198891

119895ℎ= int120572

0

119889119865119895(119910) int119910

0

(119910 minus 119909) 119889119865ℎ(119909)

for 119910119895119894gt 119910ℎ119903 119910119895gt 119910ℎ

1198892

119895ℎ= int120572

0

119889119865ℎ(119910) int119910

0

(119910 minus 119909) 119889119865119895(119909)

for 119910119895119894lt 119910ℎ119903 119910119895gt 119910ℎ

(2)

The relative mortality affluence is defined as

119863119895ℎ=1198891119895ℎminus 1198892119895ℎ

1198891119895ℎ+ 1198892119895ℎ

(3)

If the population share in subgroup 119895 is 119901119895= 119899119895119899 and

mortality share in subgroup 119895 is 119904119895= 119901119895119910119895

119910 then thecontribution to total inequality attributable to the differencebetween the 119896 population subgroup is defined as

119866119887=

119896

sum119895=1

119896

sumℎ=1 119895 =ℎ

119866119895ℎ119863119895ℎ(119901119895119904ℎ+ 119901ℎ119904119895) (4)

The Gini index for subgroup 119895 is given by

119866119895119895=sum119899119895

119894=1sum119899119895

119903=1(119910119894119895minus 119910119903119895)

21198992119895119910119895

(5)

Within group inequality index is the sum of Gini indices forall subgroups weighted by the product of population sharesand mortality shares of the subgroups

119866119908=

119896

sum119895=1

119866119895119895119901119895119904119895 (6)

If subgroups are not overlapping total inequality can beexpressed as the sum of within group and between groupindices But if subgroups are overlapping we can add anothercomponent which is a part of between group disparitiesissued from the overlap between the two distributions whichmeasures the contribution of the intensity of transvariationThe contribution of the transvariation between the subpopu-lations to 119866 is

119866119905=

119896

sum119895=1

119896

sumℎ=1 119895 =119896

119866119895ℎ(1 minus 119863

119895ℎ) (119901119895119904ℎ+ 119901ℎ119904119895) (7)

Thus Gini index can be decomposed into three compo-nents within group inequality between group inequality andinequality due to group overlapping

119866 = 119866119908+ 119866119887+ 119866119905 (8)

Oaxaca Decomposition Taking into account the utility ofregression based decomposition analyses we have also usedthe Oaxaca decomposition model which is appropriate forbinary models to decompose the gender differences in childsurvival into relative contribution attributable to explana-tory factors [25] Mathematically Oaxaca decomposition isexpressed in three equations (1)The gap between the meanoutcomes 119910Male and 119910Female is equal to

119910Maleminus 119910

Female= 120573

Male119909Maleminus 120573

Female119909Female (9)

where 119909Male and 119909Female are vectors of explanatory variablesevaluated at the means for the Male and Female respectively

Further we estimated how much of the overall gap orthe gap specific to any one of the 119909rsquos is attributable to (1)difference in 119909rsquos called the explained component rather than(2)differences in120573s (also called the unexplained component)This is mathematically expressed as


Female= Δ119909120573

Female+ Δ120573119909

Male (10)


Female= Δ119909120573

Male+ Δ120573119909

Female (11)

whereΔ119909 = 119909Maleminus119909Female andΔ120573 = 120573Maleminus120573Female show thatin (10) (unexplained part) the differences in 119909rsquos are weightedby the coefficients of the female group and differences inthe coefficients are weighted by the 119909rsquos of the male groupwhereas in (11) (explained part) the differences in the 119909rsquos areweighted by the coefficient of male group and the differencein coefficients are weighted by the 119909rsquos of female group

23 Variables Used in the Study The socioeconomic anddemographic variables were dichotomized into disadvan-tageous and advantageous groups (based on inputs fromthe simple bivariate analyses) to perform the differentialdecomposition analyses place of residence as ruralurbancaste SCs or STsothers religion as Hindu religionnon-Hindu religion motherrsquos age as risky (less than 19 yearsand above 30 years) age groupnonrisky (20ndash29 years) agegroup education of mother as no educationeducation massmedia exposure of mother as noyes mothersrsquo body massindex less than 185 kgm2more than 185 kgm2 economicstatus as poornonpoor place of delivery as homeother birthweight as lowhigh birth order less than 3above 3 antenatalcare receivednot received and breast feeding less than 6monthsabove 6 months For more details on variables seeIIPS and Macro International 2005-06 [23]

3 Results

31 Trends in Difference of Male-Female Child Mortality Thechild mortality estimates (

41199021) by sex are given in Table 1

The results showed that in 1992 female children were at a


Table 1 Trends in child mortality (41199021) by sex of the child in India and major states 1992ndash2006

StatesNFHS-1 (1992-93) NFHS-2 (1998-99) NFHS-3 (2005-06)

1198631minus 1198633

M F 1198631

(F minusM) M F 1198632


(F minusM)Delhi 136 212 76 106 134 28 76 89 13 63Haryana 184 432 248 138 302 164 104 21 106 142Himachal Pradesh 176 253 77 90 93 03 50 42 minus08 85Punjab 127 23 103 59 238 179 60 155 95 08Bihar 345 535 19 314 436 122 245 404 159 31Madhya Pradesh 467 568 101 494 663 169 247 326 79 22Rajasthan 265 422 157 294 523 229 185 262 77 8Uttar Pradesh 385 656 271 288 534 246 217 432 215 56Assam 529 596 67 214 169 minus45 226 299 73 minus06Orissa 161 234 73 296 279 minus17 311 265 minus46 119West Bengal 217 354 137 185 239 54 131 149 18 119Gujarat 271 386 115 251 314 63 97 214 117 minus02Maharashtra 191 236 45 155 20 45 79 91 12 33Andhra Pradesh 215 276 61 166 278 112 92 132 4 21Karnataka 256 334 78 211 238 27 147 131 minus16 94Kerala 10 94 minus06 6 45 minus15 14 24 1 minus16Tamil Nadu 29 232 minus58 127 158 31 49 105 56 minus114India 294 42 126 249 367 118 142 229 87 39

Table 2 Pyattrsquos Gini decomposition of change in male-female differential in child mortality in India during 1992ndash2006

Decomposed categories1992-93(NFHS-1)

1998-99(NFHS-2)

2005-06(NFHS-3)

Gini indices Gini indices Gini indices Contribution from between inequality 0004 099 0004 102 0001 053Contribution from overlap 0177 4895 0173 4891 0135 4941Contribution from within inequality 0181 5005 0177 5072 0137 5006Total inequality 0361 100 0353 100 0273 100Note figures might be affected by rounding

higher risk of mortality in all the states except in Kerala andTamil Nadu where female children reflected a better positionA positive shift towards gender equalities in child survivalwas clearly evident from the estimates of child mortalitytrends during 1992 to 2006 Yet female children consistentlyremained at a higher risk though there was a decline inthe volume of risk The state specific analysis revealed thatover the time Kerala and Tamil Nadu turned favourabletowards male child whereas Himachal Pradesh Orissaand Karnataka emerged as the states favourable towardsfemale child Uttar Pradesh consistently showed greater sexdifferentials in child mortality with a difference of 21 deathsper 1000 live births between male and female children Thestate was able to reduce a difference of only 6 deaths per 1000live births during 1992 to 2006 A similar trend was reflectedin Bihar and Haryana where male-female difference in childmortality was quite high with an impressive progress during1992 to 2006 A more unexpected outcome was observed intwo of the southern states that is Kerala and Tamil Nadu

which initially showed a female advantage in child survivalbut this replaced with male advantage in the latest periodNevertheless the results in Table 1 clearly evident that femalechild mortality was always higher in the north and centralregions of India with Uttar Pradesh and Madhya Pradeshbeing the states with highest female child mortality The sexdifferentials in child mortality decreased in almost all thestates with the exception of Assam (67 to 73) Gujarat (115 to117) Kerala (minus06 to 1) and Tamil Nadu (minus58 to 56) whereit was slightly increased in 2006 as compared to 1992

32 Pyattrsquos Decomposition Results Consistent with the resultsof Table 1 the results in Table 2 showed that considerablegender inequalities persist in child mortality during 1992ndash2006 However this difference was less in 2005-06 (119866 = 023)as compared to 1992-93 (119866 = 036) The decompositionof gender based inequalities in child mortality revealed thatwithin male and female inequalities contribute more than


Table 3 Oaxaca decomposition contribution of selected predictors to child mortality difference between female and male children India2005-06

Summary of Oaxaca decompositionCoef Std error

Female 014 00154Male 009 00122Difference 005 00196Explained 004 00139Unexplained 001 00141 explained 8195 mdash unexplained (residual) 1805 mdash

Details of explained partPredictors contribution to total difference Std errorRural place of residence 598 00010Motherrsquos no education 473 00012Motherrsquos age (15 to 19 and 30+) 561 00008Mothers do not have mass media exposure 231 00014Non-Hindu religion 022 00003SCsSTs caste 061 00006Noninstitutional place of deliveryhome delivery minus139 00012Poor household economic status 393 00004No antenatal care received 914 00013ge3 birth order 2427 00017lt6 months breastfeeding 4046 00036lt185 kgm2 motherrsquos body mass index 001 00001Low birth weight 412 00003Total explained part 100

betweenmale-female inequalitiesThe contribution of withinmale and female inequalities to total gender inequalitiesin child mortality was consistently over 50 for all thethree rounds of the NFHS This gives a strong message thatrather than male-female inequalities inequalities within thefemale and male children by socioeconomic demographicand health care characteristics of the households contributedmore to the total inequality in the child mortality Theother important finding from Table 2 was the contribution ofoverlapping factors which was consistently high (nearly 49)during 1992ndash2006 Here overlapping factors were neitherthe individual factors of female nor the individual factors ofmale These factors can be referred to parents communityand other contextual factors which might be having an equalprobability of favoring either of them Therefore the entirefemale child population cannot be treated as a disadvantagedgroup rather a certain section of this population is in adisadvantageous position depending on their exposure toparents community and other contextual discriminationsFurther to identify the explanatory factors in the form ofsocioeconomic demographic and health care characteristicsof the households we have decomposed the sex differentialsin child mortality by their socioeconomic and demographiccharacteristics in the following section

33 Oaxaca Decomposition Results Table 3 presentsthe results of decomposition analysis for proportional

contributions of selected predictors The socioeconomicand demographic predictors can explain up to 82 of thetotal mortality difference between male and female childrenin India The remaining 18 constitutes the unexplainedresidual component The results in Table 3 also illustratethe relative contribution of selected predictors by takingthe total explained components (ie 82) equivalent to100 Among all the explanatory factors breastfeeding upto less than six months contributed 40 of the differencein male and female child mortality Women of birth orderof 3 and above and those who have not received antenatalcare each contributed 24 and 9 respectively to the sexdifferential in child mortality Rural place of residence andmotherrsquos age contributed 5 eachThe variables like motherswith no education no mass media exposure non-Hindureligion scheduled caste and tribe low body mass index(lt185 kgm2) and low birth weight of baby contributed lessthan 5 of the gender differential in child mortality Thenegative contribution (minus139) of home delivery indicatesthat the place of delivery is not a significant factor for sexdifferentials in child mortality in India

4 Conclusion

This study assessed the trends and determinants of sex differ-entials in child mortality in India during 1992 to 2006 Thefindings foster that there was reduction in sex differentials


in child mortality during 1992 to 2006 Yet the female childcontinues to be at greater risk of death in the majority ofthe states of India with the only exception in Kerala andTamil Nadu where the earlier female advantage in childsurvival chances was replaced with male advantage Thispattern also reflected in the results of recent Census (2011)in the form of a remarkable decline in child sex ratios inmany districts of Tamil Nadu [35] Similarly in case of Keralathe studies documented emerging female discrimination intraditionally matriarchal societies like Nayars [36ndash38] Thuswe can attribute the reversal in female advantage in childsurvival to emergent manifestation female discrimination inthese states The findings of Pyattrsquos decomposition analysessuggested that the discrimination within the female andmale children contributed in greater proportion to totalinequality in child mortality compared to between maleand female differences Moreover the overlapping factorsnamely parental community and other contextual factoralso contributed largely to sex differences in child mortal-ity Thus the contribution from overlapping factors clearlyindicates that male children are treated at par with femalechildren irrespective of the subgroup of the population towhich they belong to Conversely female children are treateddifferently across different subgroups of the population Theresults of the Oaxaca decomposition model suggested thatthe demographic variables such as breastfeeding birth orderantenatal care and the motherrsquos age emerged as critical con-tributors for excess female child mortality compared to othersocioeconomic variables Further variables namely higherorder births and younger maternal age also contributed ina significant proportion to the excess female child mortalityin India The results advance that proximate determinants asdiscussed by Mosley and Chen [26] are playing a greater rolein creating sex differentials in child mortality compared todistal socioeconomic factors

Thoughwe agreewith some of the previous studieswhichhave attributed gender differences in child mortality to girlchild discrimination in accordance with the cultural contextof social and religious values in Indian society [6 11 17ndash20] however our findings are in tune with the studies whichadvanced the argument that socioeconomic factors play acritical role but these factors do not directly influence sexdifferences in child mortality rather they operate throughfactors like breastfeeding birth order and antenatal and childhealth care [27ndash33]This could be one reason why the relativecontribution of the socioeconomic factors in this study wasvery less compared to demographic and health care variablesBreastfeeding variable alone contributed phenomenally tooverall sex differences in child mortality because in absenceof variables like children vaccinated nutrition and otherhealth care variables breastfeeding might be showing acompound effect of all these variables which in turn is closelyassociated with breastfeeding We have not included childimmunization nutrition and other health care variables inOaxaca decompositionmodel because these questions are notasked for dead children in the NFHS surveys This is one ofthe limitations of the analyses based on NFHS data

In conclusion the study advances the argument thatthe gender discrimination is mainly operating through the

unequal provision of breastfeeding negligence of higherorder female births and consequences in terms of allocationof intrahousehold resources in health care provisions Atpolicy perspective this study evidently suggests that theissue of sex differentials in child mortality still remains adaunting task for the policy makers social scientists andpublic health professionals in India Any efforts to eliminategender differences in child survival should focus on withinfemale children discrimination across different populationsubgroups alongside male-female discrimination Also weneed to overcome the discrimination in the terms of proxi-mate factors such as breastfeeding and other nutritional andhealth care requirements and neglect of higher order femalebirths in India The study has some data limitations becauseNFHS collects data on mortality retrospectively for the lastfive years and this has a potential for recall bias especially incase of reporting girlsrsquo deaths in India

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

References

[1] KHill andDMUpchurch ldquoGender differences in child healthevidence from the demographic and health surveysrdquo Populationand Development Review vol 21 no 1 pp 127ndash151 1995

[2] Office of the Registrar General India Sample RegistrationSystem Statistical Report-2010 Office of the Registrar Generalof India Ministry of Home Affairs New Delhi India 2011

[3] P Visaria Pravin ldquoThe sex ratio of population of India andPakistan and regional variations during 1901ndash1961rdquo in Patternsof Population Change in India A Bose Ed Allied PublishersBombay India 1967

[4] P K Bardhan ldquoOn life and death questionrdquo Economic andPolitical Weekly vol 9 no 32ndash34 pp 1293ndash1304 1974

[5] P K Bardhan ldquoSex disparity in child survival in rural Indiardquo inRural Poverty in South Asia T N Srinivasan and P K BardhanEds Columbia University Press New York NY USA 1988

[6] MDasGupta ldquoSelective discrimination against female childrenin rural Punjab IndiardquoPopulation andDevelopment Review vol13 no 1 pp 77ndash100 1987

[7] A M Basu ldquoIs discrimination in food really necessary forexplaining sex differentials in childhoodmortalityrdquo PopulationStudies vol 43 no 2 pp 193ndash210 1989

[8] B D MillerThe Endangered Sex Neglect of Female Children inRural North India Cornell University Press Ithaca NY USA1981

[9] A Sen ldquoFamily and food sex bias in povertyrdquo in Rural Povertyin South Asia T N Srinivasan and P K Bardhan EdsColumbia University Press New York NY USA 1988

[10] P K Muhuri and S H Preston ldquoEffects of family compositionon mortality differentials by sex among children in MatlabBangladeshrdquo Population and Development Review vol 17 no 3pp 415ndash434 1991

[11] S Kishor ldquolsquoMayGod give sons to allrsquo gender and childmortalityin Indiardquo American Sociological Review vol 58 no 2 pp 247ndash265 1993


[12] M Murthi A-C Guio and J Dreze ldquoMortality fertility andgender bias in India a district-level analysisrdquo Population ampDevelopment Review vol 21 no 4 pp 745ndash782 1995

[13] M D Gupta and P N M Bhat ldquoFertility decline and increasedmanifestation of sex bias in Indiardquo Population Studies vol 51no 3 pp 307ndash315 1997

[14] F Arnold M K Choe and T K Roy ldquoSon preference thefamily-building process and child mortality in Indiardquo Popula-tion Studies vol 52 no 3 pp 301ndash315 1998

[15] S Parasuraman ldquoDeclining sex ratio of the child population inIndiardquo IIPS Newsletter pp 29ndash39 2001

[16] P N M Bhat and A J F Zavier ldquoFertility decline and genderbias inNorthern IndiardquoDemography vol 40 no 4 pp 637ndash6572003

[17] P Arokiasamy ldquoRegional patterns of sex bias and excess femalechild mortality in Indiardquo Population vol 59 no 6 pp 833ndash8632004

[18] J R Behrman ldquoIntrahousehold allocation of nutrients inrural India are boys favored Do parents exhibit inequalityaversionrdquo Oxford Economic Papers vol 40 no 1 pp 32ndash541988

[19] L C Chen E Huq and S DrsquoSouza ldquoSex bias in the family allo-cation of food and health care in rural Bangladeshrdquo Populationand Development Review vol 7 no 1 pp 55ndash70 1981

[20] M M Rahaman K M Aziz M H Munshi Y Patwari andM Rahman ldquoA diarrhea clinic in rural Bangladesh influenceof distance age and sex on attendance and diarrheal mortalityrdquoThe American Journal of Public Health vol 72 no 10 pp 1124ndash1128 1982

[21] IIPS andMacro International National Family Health Survey 1India International Institute for Population Sciences MumbaiIndia 1992-93

[22] IIPS andMacro International National FamilyHealth Survey 2International Institute for Population Sciences Mumbai India1998-1999

[23] IIPS andMacro InternationalNational Family Health Survey 3International Institute for Population Sciences Mumbai India2005-2006

[24] G Pyatt ldquoOn the interpretation and disaggregation of ginicoefficientsrdquo The Economic Journal vol 86 no 342 pp 243ndash255 1976

[25] R Oaxaca ldquoMale-female wage differentials in urban labormarketsrdquo International Economic Review vol 14 no 3 pp 693ndash709 1973

[26] W H Mosley and L C Chen ldquoAn analytical framework for thestudy of child survival in developing countriesrdquo Population andDevelopment Review vol 10 pp 24ndash45 1984

[27] S Bhalotra ldquoFatal fluctuations Cyclicality in infant mortalityin Indiardquo Journal of Development Economics vol 93 no 1 pp7ndash19 2010

[28] V K Borooah ldquoGender bias among children in India in theirdiet and immunisation against diseaserdquo Social Science andMedicine vol 58 no 9 pp 1719ndash1731 2004

[29] S Jayachandran and I Kuziemko ldquoWhy do mothers breastfeedgirls less than boys Evidence and implications forchild healthin indiardquo Quarterly Journal of Economics vol 126 no 3 pp1485ndash1538 2011

[30] Million Death Study Collaborators ldquoCauses of neonatal andchild mortality in India a nationally representative mortalitysurveyrdquoThe Lancet vol 376 pp 1853ndash1860 2010

[31] V Mishra T K Roy and R D Retherford ldquoSex differentials inchildhood feeding health care and nutritional status in IndiardquoPopulation and Development Review vol 30 no 2 pp 269ndash2952004

[32] E Oster ldquoProximate sources of population sex imbalance inIndiardquo Demography vol 46 no 2 pp 325ndash339 2009

[33] R P Pande ldquoSelective gender differences in childhood nutritionand immunization in rural India the role of siblingsrdquo Demog-raphy vol 40 no 3 pp 395ndash418 2003

[34] U Ram P Jha F Ram et al ldquoNeonatal 1-59 month and under-5 mortality in 597 Indian districts 2001 to 2012 estimates fromnational demographic andmortality surveysrdquoTheLancetGlobalHealth vol 1 no 4 pp e219ndashe226 2013

[35] A Perianayagam and S Goli ldquoProvisional results of the 2011Census of India slowdown in growth ascent in literacy butmore missing girlsrdquo International Journal of Social Economicsvol 39 no 10 pp 785ndash801 2012

[36] A Rondinone ldquoReconsidering the Status of Women in Kerala(India) A geographical analysis based on recent data on emi-gration sex ratio and social statusrdquo Rivista Geografica Italianavol 114 no 2 pp 179ndash205 2007

[37] TV Sekher andNHattiUnwantedDaughters GenderDiscrim-ination in Modern India Rawat Publications New Delhi India2010

[38] P Arokiasamy and S Goli ldquoSkewed child sex ratio in Indiarevisiting ldquolandholding-patriarchy hypothesisrdquordquo Economic andPolitical Weekly vol 47 pp 85ndash94 2012

Submit your manuscripts athttpwwwhindawicom

Child Development Research

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Education Research International


Biomedical EducationJournal of



Psychiatry Journal

ArchaeologyJournal of



AnthropologyJournal of


Research and TreatmentSchizophrenia


Urban Studies Research

Population ResearchInternational Journal of


CriminologyJournal of


Aging ResearchJournal of



NursingResearch and Practice

Current Gerontologyamp Geriatrics Research

Hindawi Publishing Corporationhttpwwwhindawicom

Volume 2014

Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

AddictionJournal of


Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Geography Journal


Research and TreatmentAutism


Economics Research International


102030405060708090

Indi

aA

ndhr

a Pra

desh

Ass

amBi

har

Chat

tisga

rhD

elhi

Guj

rat

Har

yana

Him

acha

l Pra

desh

Jam

mu

Kash

mir

Jhar

khan

dKa

rnat

aka

Kera

laM

adhy

a Pra

desh

Mah

aras

htra

Odi

sha

Punj

abRa

jstha

nTa

mil

Nad

uU

ttar P

rade

shW

est B

enga

l

MaleFemale

per1

000

live b

irths

Und

er-fi

ve m

orta

lity

Figure 1 Under-five mortality rates by sex of the child India andstates 2010

Recently some studies [27ndash34] have tried to provideinsights on existing inequalities but there is no evidence ofdecomposition of the relative proportional contribution ofsocioeconomic and demographic factors in sex differentialsin child mortality The lack of information on the exactrelative contribution of different socioeconomic and demo-graphic predictors of excess female child mortality obscuresthe planning process that can reduce gender disparity inchild mortality Thus it is imperative to estimate the relativeproportional contribution of predictors as it allows the policymakers to prioritize the factors and consequent interventionsto bring balance in survival chances of male and femalechildren in India The other question which needs ourattention is whether the reported sex differentials in childmortality consist of only the difference between male-femalechildmortality or withinmale and female differences in childmortality are also contributing to it Hence decompositionof between and within male-female differences in childmortality are also important in policy perspective With thisrationale this study has twofold objectives (1) to investigatethe progress in sex differentials in child mortality in Indiain terms of between and within group changes and (2) toidentify the factors explaining the sex differentials in childmortality and quantify their relative contributions

2 Methods

21 Data Source Data from the three rounds of the NationalFamily Health Survey (NFHS) 1992ndash2006 have been usedto assess the trends in child mortality stratified by sexThe NFHS is an equivalent of worldwide DemographicHealth Survey (DHS) and a widely accepted source of datafor estimating demographic and public health indicators inIndia Three rounds of NFHS have been conducted by theInternational Institute for Population Sciences (IIPS) andMacro International under the aegis of theMinistry of Healthand FamilyWelfare IndiaTheNFHS collects information on

fertility mortality morbidity and maternal and child healthA similar sampling scheme was adopted in all the threerounds and provides sufficiently large sample sizes (morethan 90000 households were interviewed in each round) tocarry out analysis at the national as well as the state levelThe data were collected using similar interview schedulesand the household and eligible female informant responserates were consistently above 90 across the three roundsTo make the estimates representative and comparable acrossthe three survey rounds and to account for the multistagesampling design adopted in the three rounds of the NFHSwe used appropriate weights in the analysis The detailsof the sampling weights are given in NFHS reports of thevarious rounds For details on sampling see IIPS and MacroInternational 1992-93 1998-99 and 2005-06 [21ndash23]

22 Analysis Life table approach and Pyatt and Oaxacadecomposition models were used as methods of analyses

Life-Table Approach Child mortality rates were estimatedbased on the simple life table approach using STATA 110(Table 2) This approach builds up a 12ndash59 month matrix ofexposure and deaths for each month that the child was alivebetween the first and fifth birthday of life Child mortalityrates are calculated using the following steps

(1) We calculated the date of birth of the child as a centurymonth code (cmc) denoted by bcmc

(2) We calculated the date of death of the child as acentury month code (dcmc) In this process age atdeath in months is calculated for all deaths imputingthe age in months

(3) After tallying the deaths and exposure for eachmonth the probability of dying at exact age 119909 (119902

119909)

is calculated by dividing the deaths at age 119909 by theexposure at age 119909 for each month between first andfifth birthday of life

(4) We calculated the probability of surviving to age 119909 as119897119909= 119897119909minus 1 lowast (1 minus 119902

119909) with 119897

119900= 1

(5) We calculated the probability of dying by age 119909 as 1 minus119897119909

where 119902 is the probability of dying and 119897 is the probability ofsurviving

Pyattrsquos Gini Decomposition Model Pyatt [24] has given thedecompositionmodel ofGini coefficient Gini indexwas usedto calculate the change in inequality in childmortality amongthe male-female population of Indian states during 1992ndash2006 Further the Gini index in child mortality among themale-female population across Indian stateswas decomposedinto between group inequality within group inequality andinequality due to group overlapping

Let a population of ldquo119899rdquo individuals with mortality vector(1199101 1199102 1199103 119910

119899) and mean mortality 119910 be desegregated in

ldquo119896rdquo subgroups with 119899 = sum119896119895=1119899119895and subgroup mean is 119910

119895



119866119895ℎ=

1

119899119895119899ℎ(119910119895+ 119910ℎ)

119899119895

sum119894=1

119899ℎ

sum119903=1

10038161003816100381610038161003816119910119895119894minus 119910ℎ119903

10038161003816100381610038161003816 (1)


1198891

119895ℎ= int120572

0

119889119865119895(119910) int119910

0

(119910 minus 119909) 119889119865ℎ(119909)

for 119910119895119894gt 119910ℎ119903 119910119895gt 119910ℎ

1198892

119895ℎ= int120572

0

119889119865ℎ(119910) int119910

0

(119910 minus 119909) 119889119865119895(119909)

for 119910119895119894lt 119910ℎ119903 119910119895gt 119910ℎ

(2)


119863119895ℎ=1198891119895ℎminus 1198892119895ℎ

1198891119895ℎ+ 1198892119895ℎ

(3)




119866119887=

119896

sum119895=1

119896

sumℎ=1 119895 =ℎ

119866119895ℎ119863119895ℎ(119901119895119904ℎ+ 119901ℎ119904119895) (4)


119866119895119895=sum119899119895

119894=1sum119899119895

119903=1(119910119894119895minus 119910119903119895)

21198992119895119910119895

(5)


119866119908=

119896

sum119895=1

119866119895119895119901119895119904119895 (6)


119866119905=

119896

sum119895=1

119896

sumℎ=1 119895 =119896

119866119895ℎ(1 minus 119863

119895ℎ) (119901119895119904ℎ+ 119901ℎ119904119895) (7)


119866 = 119866119908+ 119866119887+ 119866119905 (8)



Female= 120573






Female= Δ119909120573

Female+ Δ120573119909

Male (10)


Female= Δ119909120573

Male+ Δ120573119909

Female (11)



3 Results







1198631minus 1198633

M F 1198631






1998-99(NFHS-2)

2005-06(NFHS-3)













4 Conclusion









References
















































Psychiatry Journal



















Volume 2014


AddictionJournal of




Geography Journal







119866119895ℎ=

1

119899119895119899ℎ(119910119895+ 119910ℎ)

119899119895

sum119894=1

119899ℎ

sum119903=1

10038161003816100381610038161003816119910119895119894minus 119910ℎ119903

10038161003816100381610038161003816 (1)


1198891

119895ℎ= int120572

0

119889119865119895(119910) int119910

0

(119910 minus 119909) 119889119865ℎ(119909)

for 119910119895119894gt 119910ℎ119903 119910119895gt 119910ℎ

1198892

119895ℎ= int120572

0

119889119865ℎ(119910) int119910

0

(119910 minus 119909) 119889119865119895(119909)

for 119910119895119894lt 119910ℎ119903 119910119895gt 119910ℎ

(2)


119863119895ℎ=1198891119895ℎminus 1198892119895ℎ

1198891119895ℎ+ 1198892119895ℎ

(3)




119866119887=

119896

sum119895=1

119896

sumℎ=1 119895 =ℎ

119866119895ℎ119863119895ℎ(119901119895119904ℎ+ 119901ℎ119904119895) (4)


119866119895119895=sum119899119895

119894=1sum119899119895

119903=1(119910119894119895minus 119910119903119895)

21198992119895119910119895

(5)


119866119908=

119896

sum119895=1

119866119895119895119901119895119904119895 (6)


119866119905=

119896

sum119895=1

119896

sumℎ=1 119895 =119896

119866119895ℎ(1 minus 119863

119895ℎ) (119901119895119904ℎ+ 119901ℎ119904119895) (7)


119866 = 119866119908+ 119866119887+ 119866119905 (8)



Female= 120573






Female= Δ119909120573

Female+ Δ120573119909

Male (10)


Female= Δ119909120573

Male+ Δ120573119909

Female (11)



3 Results







1198631minus 1198633

M F 1198631






1998-99(NFHS-2)

2005-06(NFHS-3)













4 Conclusion









References
















































Psychiatry Journal



















Volume 2014


AddictionJournal of




Geography Journal








1198631minus 1198633

M F 1198631






1998-99(NFHS-2)

2005-06(NFHS-3)













4 Conclusion









References
















































Psychiatry Journal



















Volume 2014


AddictionJournal of




Geography Journal













4 Conclusion









References
















































Psychiatry Journal



















Volume 2014


AddictionJournal of




Geography Journal












References
















































Psychiatry Journal



















Volume 2014


AddictionJournal of




Geography Journal









































Psychiatry Journal



















Volume 2014


AddictionJournal of




Geography Journal





research article explaining sex differentials in child mortality in...

Documents