Gender Disparity in Intra-Household Health Care Expenditure: - - PDF document

16

Gender Disparity in Intra-Household Health Care Expenditure: Empirical Evidence from India Abstract

“Excess female death” in India has sparked great attention in investigate gender discrimination in health care expenditure (HCE) for hospitalization in India. This study examines the intra- household gender disparity in HCE in treatment of illness and examines effect of demographic and socio-economic factors on gender disparity in HCE, using 25th scheduled for two rounds (60th and 71st) of the NSSO hospitalized cases. Descriptive statistics and bivariate analysis are used to describe the characteristics of sample study and to estimate average HCE. Oaxaca- Blinder decomposition used to understand the contribution of demographic and socio-economic factors results of gender gap in HCE. Results showed that there is a huge gender disparity in average HCE and disparity has been increased in 2014. Decomposition results suggest that about 84% gender difference in HCE is due to male-female difference in socio-economic, demographic and healthcare-related factors. Education level, type of disease, level of care and duration of stay in hospital are contributing towards widening the male-female gap in HCE. And 18% difference in male-female HCE is due to the effect of these factors. To reduce the gap between male-female OOP expenditure, we need to economically empower the women through improving education status and changes in gender attitude.

17

Introduction Gender disparity in health care and morbidly in India has been well documented in recent

• decades. The female advantage in life expectancy at birth is a recent phenomenon in India,

unlike in many parts of the world (Canudas-Romo et al., 2015; Saikia et al. 2011). Also, the female advantage in overall life expectancy at birth masks the disadvantage spread across ages: In India, females are still subject to feticide and excess mortality (Bongaats, 2015; Sudha, 1999). Several research has shown that there is a significant variation in the health status of population and utilization of health care services (Batra et al., 2014; Baeten et. al., 2013; Joe et. al., 2008; Nikiema et. al., 2008; Purohit & Siddiqui, 1994). The literature on the social determinants of health showed how social and cultural factors affect health and longevity (Wilkinson & Marmot, 2003). One such factor is gender-based discrimination in health care utilization and lower health investment results of worsts health status of women and higher mortality compared to men. Recently, many researches has focused on gender discrimination and child health care, shows that parents are preferring to provide treatment for boys compared to girls when a household is facing tight resource constraints. For instance, Borooah (2004) shows that girls children were facing biased in getting proper nutritious and to be fully immunized results of excess female mortality (Rose, 1999) and decline sex ratios.. Anderson and Ray (2009, 2012) has shown that poor treatment and care at home of the female was leading causes of the risk of excess compared to males at each stage of lives. In a patriarchal society where female face discriminatory behavior in term of health care, nutrition intake, education and other opportunity, In India context are especially important to study the effect of gender on health. However, in India, the effect of gender on treatment seeking behavior within the household are relatively got less attention. While Pandey et al. (2002), find gender discrimination in the treatment of disease like diarrhea in rural West Bengal. Gosoniu et. al., (2008) showed that female suffering from tuberculosis did not get treatment at the appropriate time. There was a gender disparity in intra-household health care financing strategy among children (Behrman, 1988; Asfaw et. al., 2007, 2010). Asfaw (2007) showed that females have less probability to be hospitalized before their death. In India, very few study focused on gender disparity in Intra-household health care expenditure (HCE). This study

18

examines the gender disparity in average HCE and Intra-household HCE for inpatient care in India.

Data source

This study used 25th scheduled for two rounds (60th and 71st) of the National Sample Survey Organization (NSSO) data. The NSSO is a nationwide, large-scale population-based survey data and collected by MoSPI, Government of India (GOI). Since 1950, the NSSO has been collecting household-level data on socio-economic status of the study population, as well as health and morbidity status and health care consumption in India. The 60th and 71st round data were collected in 2004, and 2014 between January to Jnue adopted a two-stage stratified random

• design. In the 60th round, the information was collected from 73868 households, and the sample

size for male and female was 195,712 and 187,626. While in the 71st round, the information was collected from 65932 households and the sample sizes for male and female was 168,697 and 164,407. The information on health expenditure was collected separately for inpatients and outpatients. Details of ailments and hospitalization were collected in the reference period of last 365 days. HCE information collected at disaggregated level that includes total amount spent on medicines cost, doctor's fee, diagnostic tests charge, other medical expenses (blood, oxygen, attendant charges, personal medical appliances, physiotherapy, etc.), bed charges, transportation fee. For each individual of the households, the detail information about sex, age, morbidity status communicable and non-communicable disease, treatment status and hospitalization were recorded, and for each patient, the episode of ailments in last 356 days, treatment status, type of health care facility used, medical and non-medical HCE, sources of healthcare finance to

• vercome hospitalization cost and duration of hospitalization were collected using a
• questionnaire. This study used HCE of hospitalization (medical and nonmedical expense) for

each hospitalized (inpatient care) person during the last 365 days before the survey. HCE used in this study is converted to constant prices by GDP deflator of 175 (2014) at base price index (2005=100).

Methodology

Measures

19

Outcome variable

In the both round of the survey, data on HCE for hospitalization was collected separately for each episode of hospitalization. Along with medical expenditure, other expenditures recorded

• separately. Medical expenditure constituted by expenditure on medicine, bed charge for

hospitalized treatment, charge for diagnostic test, doctor fee. The other expenditure includes all expenditures related to the treatment of an ailment incurred by the households, but expenditure regarding medical treatment is excluded. Other medical cost included all transport cost paid by the households connection with the treatment of patient, food and lodging expenditure of the escort(s) during last one year. The total expenditure constituted by sum of the medical expenditure and other expenditure. We have estimated gender disparity in average HCE (Inter- household) and Intra-household (with in household) HCE. Gender disparity in intra household HCE expenditure estimated only for 2014. The maternity cases are not included in this study since those are restricted to only one gender. To estimate per capita annual HCE, expenditure of all episodes for an individual has added in case of more than one episode of the same individual for the same disease. The explanation about the predictor variables is given in the detail.

Predictor variables

Age group: Age of the individuals has categorized into three groups: 0-14 years, 15-49 years and 50 and above 50 years. Education level: The educational status of an individual has coded into five categories: no education, up to the primary, up to secondary, up to higher secondary, Graduate and above. These groups had classified in such as way that they have a distinct effect on the health care spending. Religion: The religious categories have classified in Hindu-1, Muslim-2 and Others-3. Others religion constituted by Christianity, Sikhism, Jainism, Buddhism, Zoroastrianism, others. Social group: The social groups are recoded into three groups: Scheduled Caste (SC)/Schedule Tribe (ST), Other Backward Castes (OBCs) and General Castes.

20

Economic indicator: The collection of income of the household is always a challenging

• exercise. This is based on total annual consumption expenditure made by a household. We have

categories it in five distinct categories according to the rank of 20 percent. The categories are Poorest-1, Poorer-2, Middle-3, Richer-4, Richest-5 Level of care: Type of healthcare facility of inpatient care is categories in two categories Public- 1 and Private-2 health care facility Place of Residence: Place of residence is coded into Rural-1 and-2 Type of disease: Type of disease categories in Communicable, Non-communicable and others

• disease. Detail of the disease categories given in Appendix S.1

Methods

The descriptive statistics and bivariate analysis used to describe the characteristics of sample study (inpatients) and to estimate average HCE separately for male and female by background

• characteristics. We considered only annual inpatient HCE in this study.

To estimate, gender disparity in intra-household HCE, we selected the household has at least one male and female hospitalized. The difference in the HCE within the household calculated by the total expenditure of male subtracts by the total expenditure of female. In the case of more than

• ne male or female hospitalized within the household then, we have taken then an average of

males expenditure and average of females expenditure then an average expenditure of males is subtracted by an average expenditure of females. Many individuals hospitalized more than one times due to the same individual has suffered from different ailments or hospitalized in a different hospital, the total expenditure for that person is the sum of all episodes of expenditures. The process to calculate gender disparity in intra-household HEC is described through following equations.

Average gender disparity in Intra-household healthcare expenditure:

Sum of the healthcare expenditure of male members: ∑ ∑ ∑ ……………(1)

21

Similarly, Sum of the health expenditure for female members: ∑ ∑ ∑ …………….(2) Where, k represents number of episodes of the patient; m and f represent the number of male and female members in household hospitalized during 365 days; M and F represent the sex of patient (male and female). Now, average expenditure for males in the first household= ∑

∑ ∑ ∑

………………….(3) Average expenditure for males in the first household= ∑

∑ ∑ ∑

………………..........(4) Finally, Average of Gender Disparity in Intra-household healthcare expenditure= ∑

[∑

∑ ∑ ∑

∑

∑ ∑ ∑

]

..…………..(5) N is total number of households.

Oaxaca Decomposition

To quantify the role of demographic, socio-economic and other disease-related variables in HCE, we used decomposition technique propounded by Oaxaca (year & Reference). The core idea is to find out the inequality in HCE is the consequence of the distribution of a set of the difference in the socio-economic and demographic factors. For example, disparity in HCE results of difference in quality and type of healthcare facility, education, economic status and social factors such as caste and religion. This study used Oaxaca decomposition to explain the gap in the means of HCE between male and female. The gaps between mean outcomes ymale and yfemale , is equal to …………(2)

22

Where xmale and xfemale are vectors of explanatory variable evaluated at the means for the male and female. Further, we estimated how much of the overall gap or the gap specific to any one of the x’s is attributed to (i) difference in x (sometimes called the explained components), rather than (ii) difference in the β’s is (sometimes called unexplained components). The gap between male and female HCE could be explained in either of two ways. ………………………(3) ………………………(4) Where ∆x= xmale- xfemale and ∆β= βmale - βfemale In the third equation, the difference in the x’s weighted by the coefficient of the female group and the difference in the coefficient is weighted by the x’s of the male group. Whereas in the fourth equation the x’s are weighted by the coefficient of male and difference in coefficient is weighted by the x’s of the female. The general form of decomposition can be write as …………… (5) =E+C+CE The equation 5 shows the gap between average HCE of male and female as a result of difference in the distribution of endowments (E), coefficients (C), and the interaction (CE) of endowments (E) and coefficients (C).

Descriptive statistics

Table 1 shows the descriptive statistics of study variables; gender wise separately for two rounds

• f the survey, 2004 and 2014. The results demonstrate that the gender difference in medical and

total HCE has increased from 2004 to 2014. Average medical and total HCEs for male are significantly higher compared to female in 2004 and 2014. During 2014, the average medical expenditure has increased two-fold for males whereas for females it only shows a slight increase.

23

Figure 1 shows that overall HCE also increased in 2014, for males the increase is higher compared to females. Figure 1. Average healthcare expenditure for male and female during 2004 and 2014 Furthermore, the sample distribution by demographic and socioeconomic characteristics shows that 21 % of males and 14 % of females were under 15 years of age during 2014. Similarly, 38 % males and 33% females in the sample belong to 50 and above age group. The most of the sample lives in rural areas. The analysis shows that nearly 26 % male and 40 % female were illiterate while 26 % male and 23 % female had completed primary schooling, 14 % male and 12 % female completed up to secondary education and 8 % male and 5 % female have studied up to graduate and above. By religion, more than 78 % male and 76 % female sample belong to the Hindu community, 13 % male and 14 % female to Muslim and around 9 % of the sample for both sex to other religious categories includes Christians, Sikhs, Jain and Buddhist, etc. By caste group, the result shows that around 28 % belonged to SC/ST, 40 % belonged to OBC and 32 % belonged to General Caste. By the economic background, 28 % male and female were poorest, and 19 % were the richest background. 2000 4000 6000 8000 10000 12000 14000 16000 2004 2014 8843 15606 7695 11189 HCE (INR) year Male Female

24

Table 1. Descriptive statistics of study sample, India, NSS, 2004, 2014.

Variable Categories 2004 2014 Male Female Male Female Mean/ % ±CI Mean/% ±CI Mean/ % ±CI Mean% ±CI Medical Cost 7276 ±285 ±6408 ±313 ±11902 ±488 ±8629 ±334 Total Cost 8843 ±330 ±7695 ±373 ±15606 ±571 ±11189 ±415 Age (in years) 0-14 22.62 ±1.36 ±14.44 ±1.21 ±21.33 ±1.16 ±13.72 ±0.99 15-49 44.22 ±1.61 ±57.12 ±1.7 ±40.92 ±1.4 ±52.5 ±1.44 50 and Above 33.16 ±1.53 ±28.43 ±1.55 ±37.75 ±1.38 ±33.77 ±1.37 Sector Rural 65.22 ±1.54 ±63.77 ±1.65 ±54.84 ±1.41 ±53.69 ±1.44 Urban 34.78 ±1.54 ±36.23 ±1.65 ±45.16 ±1.41 ±46.31 ±1.44 Education No education 34.2 ±1.56 ±50.18 ±1.73 ±26.42 ±1.25 ±40.02 ±1.42 Up to primary 29.74 ±1.5 ±23.75 ±1.47 ±25.91 ±1.24 ±23.59 ±1.23 Up to secondary 25.3 ±1.43 ±19.09 ±1.36 ±13.7 ±0.98 ±11.97 ±0.94 Up to higher secondary 4.83 ±0.7 ±3.3 ±0.62 ±19.64 ±1.13 ±15.29 ±1.04 Graduate and above 5.94 ±0.78 ±3.68 ±0.65 ±8.29 ±0.78 ±5.53 ±0.66 Religion Hindu 80.08 ±1.3 ±79.18 ±1.39 ±78.15 ±1.17 ±76.33 ±1.23 Muslim 11.8 ±1.05 ±12.66 ±1.14 ±13.1 ±0.96 ±14.15 ±1.01 Others 8.12 ±0.89 ±8.17 ±0.94 ±8.74 ±0.8 ±9.52 ±0.85 Social group SC/ST 26.56 ±1.43 ±26.17 ±1.51 ±27.96 ±1.27 ±28.06 ±1.3 OBC 38.91 ±1.58 ±38.72 ±1.67 ±39.56 ±1.39 ±39.76 ±1.41 General 34.54 ±1.54 ±35.11 ±1.64 ±32.48 ±1.33 ±32.18 ±1.35 Wealth quintile Poorest 21.34 ±1.33 ±20.04 ±1.37 ±27.36 ±1.27 ±27.69 ±1.29 Poorer 23.1 ±1.37 ±22.76 ±1.44 ±15.2 ±1.02 ±14.96 ±1.03 Middle 20.61 ±1.31 ±20.9 ±1.4 ±20.27 ±1.14 ±20.45 ±1.17 Richer 15.95 ±1.19 ±16.78 ±1.28 ±17.8 ±1.09 ±18.66 ±1.13 Richest 19 ±1.27 ±19.53 ±1.36 ±19.36 ±1.12 ±18.24 ±1.12 Level of Care Public 46 ±1.62 ±44.05 ±1.7 ±43.15 ±1.41 ±44.64 ±1.44 Private 54 ±1.62 ±55.95 ±1.7 ±56.85 ±1.41 ±55.36 ±1.44 Type of disease Communicable 26.6 ±1.52 ±26.6 ±1.52 ±25 ±1.23 ±32.3 ±1.35 Non Communicable 53.43 ±1.71 ±53.43 ±1.71 ±51.54 ±1.42 ±50.08 ±1.44 Others 19.98 ±1.37 ±19.98 ±1.37 ±23.46 ±1.2 ±17.62 ±1.1 Region North 13.83 ±1.12 ±13.87 ±1.19 ±14.48 ±1 ±14.84 ±1.03 Central 17.61 ±1.24 ±18.11 ±1.32 ±16.94 ±1.07 ±17.44 ±1.1 East 18.18 ±1.25 ±17.24 ±1.3 ±17.53 ±1.08 ±18.01 ±1.11 Northeast 9.72 ±0.96 ±9.5 ±1.01 ±10.8 ±0.88 ±11.33 ±0.92 West 12.73 ±1.08 ±12.39 ±1.13 ±14.35 ±1 ±13.01 ±0.97 South 24.84 ±1.4 ±25.72 ±1.5 ±20.09 ±1.14 ±20.19 ±1.16 Others 3.1 ±0.56 ±3.17 ±0.6 ±5.82 ±0.66 ±5.18 ±0.64

The distribution of the sample regarding the type of health facility shows that patients were predominantly using the private health care, 56 % males and females were hospitalized in private hospitals while only 44 % male and female were hospitalized in public hospitals. The sample distribution of hospitalization by type of disease shows that 25 % male and 32 % females were hospitalized due to communicable diseases, around 50 % male and female hospitalized due to

25

non-communicable diseases and the percentage of male (24 %) hospitalized due to other diseases/causes is higher than females (18 %) during 2014. Region wise sample distribution shows that 15 percent sample comes from the North region, 17 % belonged from Central, 18 % belonged to East, 20 % belonged to South, and 5 % belonged to Other regions of India. Figure 2. Histogram and Epanechnikov kernel density function of health care expenditure, India, NSS, 2004, 2014.

.1 .2 .3 5 10 15 logexp_2004 .1 .2 .3 5 10 15 logexp_2014

26

Average healthcare expenditure for Male and Female

Table 2 shows the average HCE at current and constant price for male and female by different socio-economic background characteristics. Average HCE is shown in Indian (INR) rupees. The results show that average HCE increased in 2014 compared to 2004. The HCE is always higher for male as compared to their female counterparts. The average HCE which was INR 8843 for male and INR 7695 for a female in 2004 had increased to INR 15606 for male and INR 11189 for female in 2014. The HCE is associated with the age group of the patients, among 0-14 year age group HCE is INR 8166 for male and INR 7858 for Female, among 15-49 age group INR 14975 for male and INR 10590 for females. Among higher age groups HCE is 20299 for male and 13271 for female. The result shows that gap between male-female HCE is higher in higher age group (Figure 3). Figure 3 & 4. Average healthcare expenditure for male-female by age group and place of residence during 2014. Figure 4 shows that in urban area people incurred higher HCE (Male=INR 20470, Female=INR 14979) than their rural counterparts (Male= INR 13057, Female= INR 9163). Similarly, people who were educated, in particular those with education level up to graduation and above (Male= INR 30761, Female= INR 20524) were spending more while the less spending was for illiterate people (Male= INR 9163, Female= INR 8830). Average HCE was higher among the people belonging from Other (Male= INR 19239, Female= INR 13100) religions followed by Hindu (Male= INR 15693, Female= INR 11385) and Muslim (Male= INR 19239, Female= INR 13100)

8166 14975 20299 7858 10590 13271 5000 10000 15000 20000 25000 0-14 15-49 50 and Above HCE (INR) Age group Male Female 13057 20470 9163 14979 5000 10000 15000 20000 25000 Rural Urban HCE (INR) Sector Male Female

27

• religion. By caste affiliation of patients, the results show that absolute HCE on health care

among the patients from general caste (Male= INR 20245, Female= INR 15233) is higher than their other socially disadvantaged groups SC/ST (Male=9952, Female=8084) and OBC (Male=15607, Female=10324). Based on the economic status of the patient, mean HCE increases with an increase in the household income status: Patients from richest income group are spending significantly higher amounts (Male= INR 34889, Female= INR 21508) than the Poorest income group (Male= INR 9194, Female=6545). Figure 5 shows that average HCE was higher among the patients who were hospitalized in private hospitals (Male= INR 21032, Female= INR 15800) when compared to those hospitalized in Public (Male= INR 6676, Female= INR 4232) hospitals. Figure 5 & 6. Average healthcare expenditure for male-female by type of healthcare facility used and type of diseases during 2014 Figure 6 shows that average HCE is higher among the patient who suffered from other diseases (Male= INR 19717, Female= INR 15447) followed by the patients who suffered Non- communicable (Male= INR 18409, Female= INR 13280) and communicable (Male= INR 6012, Female= INR 5649) diseases. In Central region (Male=17769, Female=11206) higher HCE was incurred during 2014 whereas HCE was higher in North region (Male= INR 12433, Female= INR 11236) during 2004.

6676 21032 4232 15800 5000 10000 15000 20000 25000 Public Private HCE (INR) Type of Healthcare facility Male Female 6012 18409 19717 5649 13280 15447 5000 10000 15000 20000 25000 CD NCD Others HCE (INR) Type of disease Male Female

28

Table 2. Average annual healthcare expenditure for inpatient by sex and background characteristics, India, NSS, 2004, 2014

Variables Current Price Constant Price Categories 2004 2014 2004 2014 Male Female Male Female Male Female Male Female Age (in years) 0-14 4618 4026 14291 13752 4712 4108 8166 7858 15-49 9112 7499 26206 18533 9298 7652 14975 10590 50 and Above 10927 9478 35524 23224 11150 9671 20299 13271 Sector Rural 7497 6623 22850 16035 7650 6758 13057 9163 Urban 11407 9608 35823 26213 11640 9804 20470 14979 Education Level

• f women

No education 5839 5623 16036 15452 5958 5738 9163 8830 Up to primary 6765 6999 22115 17459 6903 7142 12637 9977 Up to secondary 9819 8966 24052 18487 10019 9149 13744 10564 Up to higher secondary 14459 9814 33114 23204 14754 10014 18922 13259 Graduate and above 23344 21933 53831 35917 23820 22381 30761 20524 Religion Hindu 8783 7193 27462 19924 8962 7340 15693 11385 Muslim 7784 6864 23445 16063 7943 7004 13397 9179 Others 8834 12393 33669 22925 9014 12646 19239 13100 Social group SC/ST 5797 4881 17416 14147 5915 4981 9952 8084 OBC 7849 7433 27312 18067 8009 7585 15607 10324 General 11622 9451 35428 26658 11859 9644 20245 15233 Wealth quintile Poorest 4729 4268 16090 11453 4826 4355 9194 6545 Poorer 6511 5384 17040 15015 6644 5494 9737 8580 Middle 8205 6337 20585 16081 8372 6466 11763 9189 Richer 9463 8426 28248 22579 9656 8598 16142 12902 Richest 15583 14247 61056 37639 15901 14538 34889 21508 Level of Care Public 4792 3965 11683 7406 4890 4046 6676 4232 Private 11265 9865 36806 27650 11495 10066 21032 15800 Type of disease Communicable 4080 3605 10521 9886 4163 3679 6012 5649 Non Communicable 11428 9630 32215 23240 11661 9827 18409 13280 Others 7334 7065 34505 27032 7484 7209 19717 15447 Region North 12184 11011 31095 19611 12433 11236 17769 11206 Central 9432 8203 31449 21693 9624 8370 17971 12396 East 7057 6350 20901 15452 7201 6480 11943 8830 Northeast 4795 6883 13004 15536 4893 7023 7431 8878 West 9356 7674 28370 25035 9547 7831 16211 14306 South 7742 6660 29410 18085 7900 6796 16806 10334 Others 6681 6144 19605 22401 6817 6269 11203 12801 Total 8666 7541 27311 19580 8843 7695 15606 11189

Note: Expenditure in Indian currency (INR)

29

Gender disparity in Healthcare expenditure (Table 3)

Gender disparity in average HCE by socio-economic and demographic characteristic is shown in Table 3. The fact that gender disparity does exist in HCE is abundantly clear; the average HCE was reported higher for males than females in both round of survey (2004= INR 1148 and 2014= INR 4418). Figure 7 & 8. Gender disparity in healthcare expenditure by survey year and age groups in 2004 and 2014 Figure 9 & 10 Gender disparity in healthcare expenditure by place of residence and region, India, 2004, 2014

1148 4418 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 D1(2004) D2(2014) Male-Female Exp (INR) Year 1000 2000 3000 4000 5000 6000 7000 8000 0-14 15-49 50 and Above Male-Female exp (INR) Age Ggoup

D1(2004 ) D2(2014 )

891 3894 1835 5492 1000 2000 3000 4000 5000 6000 D1(2004) D2(2014) Male-Female HCE (INR) Year Rural Urban

1197 1254 721

• 2131

1716 1104 6562 5575 3114

• 1447

1906 6472

• 3000
• 2000
• 1000

1000 2000 3000 4000 5000 6000 7000

Male-Female HCE (INR) Regions

D1(2004) D2(2014)

30

Figure 7 shows that gender disparity in average HCE increased four times in 2014 compared to

• 2004. Figure 8 shows gender disparity in HCE allocation is more apparent in higher age groups

than lower age group patients. Gender disparity in HCE increased in higher age group and reduced among children during the period 2004 to 2014. In urban areas, female incurred much lower HCE than males compared to rural areas. By education level, gender disparity in HCE among the illiterate was less (2004= INR 221, 2014= INR 333) than educated particularly among graduate and above (2004= INR 1439, 2014= INR 10236). Gender disparity was higher among Hindus in 2004, but during 2014 it has increased among Other religious group too. Patients who used private health facility show higher disparity compared to public health service users. When seen across causes, a significant disparity in HCE is observed. Non-communicable disease (2004= INR 1834, 2014= INR 5128) shows higher disparity because of more expensive/higher cost of treatment compared to communicable disease (2004= INR 485, 2014= INR 363). Figure 10 shows the region-wise analysis of HCE shows that in Northeast, females show better position compared to Western, North and Central region of India.

31

Table 3 Gender Disparities in average annual healthcare expenditure for inpatient during 2004 to 2104 by background characteristic’s, India, NSS

Variable Categories D1 D2 D2/D1 Age (in years) 0-14 603 308 0.51 15-49 1645 4385 2.67 50 and Above 1479 7029 4.75 Sector Rural 891 3894 4.37 Urban 1835 5492 2.99 Education Level of women No education 221 333 1.51 Up to primary

• 238

2661

• 11.18

Up to secondary 870 3180 3.66 Up to higher secondary 4740 5663 1.19 Graduate and above 1439 10236 7.11 Religion Hindu 1623 4308 2.65 Muslim 939 4218 4.49 Others

• 3632

6140

• 1.69

Social group SC/ST 934 1868 2.00 OBC 425 5283 12.43 General 2215 5011 2.26 Wealth quintile Poorest 471 2650 5.63 Poorer 1150 1157 1.01 Middle 1907 2574 1.35 Richer 1058 3239 3.06 Richest 1362 13381 9.82 Level of Care Public 843 2444 2.90 Private 1429 5232 3.66 Type of disease Communicable 485 363 0.75 Non Communicable 1834 5128 2.80 Others 275 4271 15.53 Insurance No 521 4897 9.40 Yes 11988 20172 1.68 Region North 1197 6562 5.48 Central 1254 5575 4.45 East 721 3114 4.32 Northeast

• 2131
• 1447

0.68 West 1716 1906 1.11 South 1104 6472 5.86 Others 547

• 1598
• 2.92

Total 1148 4418 3.85 Note: D1=Male-Female(2004), D2=Male-Female (2014

32

Average healthcare expenditure by sex and States in India: (Table 4) Figure 11 & 12. shows average HCE in Indian states during 2004 and 2014. The analysis shows that HCE increased in all states in 2014 compared to 2004. Average HCE is varying significantly by states and union territories. Some states shows higher HCE while few states show lesser, It could be attributed to the type and quality of health care facility used by patients. Figure 11. Average healthcare expenditure by Sate and Sex for inpatients during, India, NSS, 2004 Figure 12. Average healthcare expenditure by States for male-female during hospitalization, India, NSS, 2014

5000 10000 15000 20000 25000

Arunachal Pradesh Tripura Meghalaya Daman & Diu Goa Nagar Haveli Mizoram Sikkim Pondicherry J & K Nagaland Manipur Assam Kerala Orissa West Bengal Madhya Pradesh Chhattisgarh Jharkhand Tamil Nadu Karnataka Gujrat Maharashtra Bihar Himachal Pradesh Rajasthan Andhra Pradesh Uttar Pradesh Lakshadweep Haryana Uttaranchal Delhi Punjab Chandigarh

HCE (INR) States Male Female 5000 10000 15000 20000 25000 30000 35000 40000

Meghalaya Nagar Haveli Arunachal Pradesh Manipur Daman & Diu Tripura Nagaland Mizoram Assam Jammu & Kashmir Jharkhand A & N Islands Uttaranchal Chhattisgarh Orissa Sikkim Bihar Pondicherry Lakshadweep Rajasthan West Bengal Gujrat Andhra Pradesh Karnataka Madhya Pradesh Tamil Nadu Kerala Maharashtra Himachal Pradesh Uttar Pradesh Goa Haryana Delhi Punjab Chandigarh

HCE (INR) States Male Female

33

Among the union territories Chandigarh (Male= INR 37501, Female= INR 13406) and Delhi (Male= INR 25374, Female= INR 19804) reported highest HCE during 2014. Among the states average HCE was higher in Punjab, Haryana, Uttar Pradesh; Maharashtra. Another side some states like Jammu & Kashmir, Uttaranchal, Arunachal Pradesh, Manipur, Meghalaya, Assam, Jharkhand, and Chhattisgarh reported low HCE.

Gender disparity in healthcare expenditure by Indian states: (Table 5)

State-wise analysis of gender disparity in HCE demonstrates significant variation among the

• states. It was lower among all states in 2004 and increased during 2014. Some states shows

higher increase such as Chandigarh (INR 28833), Punjab (INR 11341), Haryana (INR 7635), Madhya Pradesh (INR 6985), and Tripura (INR 9812). Figure 13. Gender disparities in health care expenditure for inpatients by states, India NSS, 2004, 2014 Some of the states, particularly Northeast show the better position of female during 2004 but again disparity has increased during 2014. Only a few states such as Uttaranchal, Jharkhand, show a reduction in the gap between male and female average HCE.

• 8000
• 6000
• 4000
• 2000

2000 4000 6000 8000 10000 12000 14000

Assam Meghalaya Mizoram Manipur Uttaranchal Chhattisgarh Jharkhand Daman & Diu Arunachal Pradesh Goa Maharashtra Orissa Nagaland Jammu & Kashmir Nagar Haveli Lakshadweep West Bengal Tripura Gujarat Bihar Sikkim Karnataka Rajasthan Uttar Pradesh Delhi Himachal Pradesh Tamil Nadu Andhra Pradesh Pondicherry A & N Islands Kerala Madhya Pradesh Haryana Punjab

Gender difference in HCE (INR) States

D1(2004) D2(2014)

34

Table 4. Average annual healthcare expenditure for inpatient by sex and Indian States, NSS, 2004, 2014.

Indian States 2004 2014 Male Female Male Female India state Mean ±CI Mean ±CI Mean ±CI Mean ±CI A & N Islands 1731 ±1765 3574 ±3301 9006 ±5957 2089 ±2121 Andhra Pradesh 10593 ±1731 6409 ±660 15120 ±3049 9111 ±2730 Arunachal Pradesh 2280 ±541 7035 ±4480 5355 ±1370 4495 ±1446 Assam 5813 ±2768 6177 ±1223 8204 ±1850 12306 ±9267 Bihar 9894 ±1811 7670 ±884 11232 ±1668 7667 ±1950 Chandigarh 16916 ±9387 21654 ±14173 37501 ±20709 13406 ±16568 Chhattisgarh 7695 ±1375 5867 ±1211 9606 ±3422 9883 ±3831 Daman & Diu 4017 ±1868 8903 ±4987 6593 ±3597 6421 ±4721 Delhi 15085 ±5436 8476 ±3948 25374 ±7069 19804 ±7657 Goa 4823 ±1767 8057 ±2867 19143 ±10735 18107 ±19870 Gujarat 9274 ±1911 6775 ±1428 13560 ±1914 10012 ±2849 Haryana 12435 ±2539 11592 ±4415 21224 ±4103 12747 ±4966 Himachal Pradesh 10193 ±2228 10053 ±1733 17489 ±5249 11883 ±4430 Jammu & Kashmir 5599 ±1136 6941 ±1538 8523 ±1895 5768 ±2635 Jharkhand 8028 ±1706 4790 ±998 8792 ±1857 8688 ±5591 Karnataka 8652 ±1471 6327 ±944 15270 ±2211 10842 ±2626 Kerala 5964 ±1011 5195 ±1212 17230 ±3037 10125 ±2712 Lakshadweep 11187 ±5161 14161 ±6720 12571 ±6337 9241 ±13127 Madhya Pradesh 6843 ±1080 5603 ±765 16394 ±4607 8169 ±2451 Maharashtra 9762 ±1186 8345 ±1110 17442 ±1729 16327 ±3581 Manipur 5778 ±1045 8423 ±1898 5621 ±809 6383 ±2170 Meghalaya 3931 ±1404 5393 ±2717 4743 ±2109 6123 ±3345 Mizoram 5005 ±4917 3044 ±933 7642 ±2031 8707 ±6244 Nagaland 5682 ±1230 5876 ±1616 7494 ±3671 4857 ±3643 Nagar Haveli 4877 ±2353 4831 ±1526 5157 ±1821 2379 ±1617 Orissa 5972 ±745 4803 ±896 10582 ±1716 8192 ±3489 Pondicherry 5379 ±7333 8470 ±7012 12063 ±5315 5656 ±5315 Punjab 16370 ±2836 15925 ±5688 26428 ±7249 14642 ±5031 Rajasthan 10535 ±1478 8813 ±1002 13208 ±1887 8709 ±3099 Sikkim 5258 ±1705 4576 ±1228 11216 ±3443 7169 ±2791 Tamil Nadu 8144 ±1227 9504 ±2342 16613 ±1945 10931 ±2658 Tripura 3491 ±1276 9760 ±13332 7011 ±2852 3468 ±2913 Uttar Pradesh 11149 ±920 9611 ±749 18579 ±1425 14034 ±3717 Uttaranchal 14007 ±4519 11002 ±3654 9148 ±3126 9514 ±6271 West Bengal 6661 ±1020 6905 ±972 13345 ±2195 9888 ±2119

35

Table 5. Gender disparities in average healthcare expenditure by states during 2004 to 2014, NSS, India

India state D1(2004) D2(2014) D2-D1 A & N Islands

• 1843

6917 8760 Andhra Pradesh 4185 6009 1824 Arunachal Pradesh

• 4754

860 5614 Assam

• 363
• 4101
• 3738

Bihar 2225 3564 1339 Chandigarh

• 4738

24095 28833 Chhattisgarh 1829

• 278
• 2106

Daman & Diu

• 4886

172 5058 Delhi 6609 5570

• 1039

Goa

• 3234

1037 4270 Gujarat 2500 3548 1048 Haryana 843 8477 7635 Himachal Pradesh 139 5606 5466 Jammu & Kashmir

• 1342

2754 4096 Jharkhand 3239 104

• 3135

Karnataka 2325 4428 2103 Kerala 769 7105 6336 Lakshadweep

• 2974

3330 6304 Madhya Pradesh 1240 8225 6985 Maharashtra 1417 1115

• 302

Manipur

• 2645
• 762

1882 Meghalaya

• 1462
• 1381

81 Mizoram 1961

• 1065
• 3026

Nagaland

• 194

2637 2831 Nagar Haveli 46 2778 2732 Orissa 1169 2390 1222 Pondicherry

• 3091

6407 9498 Punjab 445 11786 11341 Rajasthan 1721 4498 2777 Sikkim 682 4047 3365 Tamil Nadu

• 1361

5682 7042 Tripura

• 6268

3543 9812 Uttar Pradesh 1537 4546 3008 Uttaranchal 3004

• 366
• 3371

West Bengal

• 244

3457 3701 Note: D1=Male-Female(2004), D2=Male-Female (2014)

Out of Pocket healthcare expenditure by type of disease

36

Table 6 shows gender wise average HCE by type of disease for 2004 and 2014 in India. The results demonstrate that HCE had increased for all disease. Average HCE has significant differences by sex and type of disease. The patients suffering from Cancer (2014= INR 35596, 2004= INR 18447), Heart diseases (2014= INR 24887, 2004= INR 16924) and Accidents/Injuries (2014= INR 15846, 2004= INR 18447) were reported with higher HCE, since these diseases have higher treatment cost. The average healthcare cost much higher for patients with heart diseases was higher for males compared to females. Similarly, Average HCE for hypertension disease was reported to be two times higher for males compared to females. The female hospitalized due to undernutrition have higher expenditure; it is a consequence of gender disparity in nutritional intake. Gender disparity HCE is less among the patients suffering from communicable diseases such as Diarrhea, Fever and Respiratory disease. Thus, we may conclude that gender disparity was less for communicable diseases compared to Non-Communicable diseases. Table 6. Average healthcare expenditure by gender and type of disease, India, 2004, 2014 Type of disease 2004 2014 Male Female Total Male Female Total Diarrhea 1891 1722 1808 3435 2988 3213 Fever 3311 2995 3175 5432 4446 4955 Hypertension 7156 5223 6053 12480 4740 8469 Respiratory 4721 4639 4687 7948 7917 7932 Tuberculosis 7935 5619 7135 7751 7030 7490 Bronchial Asthma 4121 4784 4367 9382 7455 8484 Cataract 3133 3412 3287 4481 6828 5883 Diabetes 9544 5463 7438 9913 8323 8997 Under nutrition 4145 2143 3039 4455 6073 5308 Anemia 5190 4156 4572 9323 8355 8720 STD 12070 7124 7957 5330 2946 3947 Heart disease 16737 17206 16924 30195 17172 24887 Cancer 16111 19918 18447 39460 32655 35596 Accidents/Injuries 10443 7809 9653 16717 13366 15846 Others 8356 7251 7786 14116 10697 12231

Gender Disparity in Intra-household healthcare expenditure

37

Descriptive statistics Gender Disparity in Intra-Household healthcare expenditure

Table 8 shows the average intra-household gender disparity in HCE by socio-economic and demographic characteristics of the household heads. The household headed by less than 50 year age (INR 3844) group shows less gap in average HCE of male and female compared higher age group (INR 21086) head of household. The household headed by male (INR 14112) show higher gender disparity than female head (INR 13737). The households which belonged to Muslim (INR 18527) community shows the higher disparity in HCE followed by Other (INR 12154) compared to Hindu (INR 6524) community household. Analysis by social groups demonstrate that the households belonging to OBC caste group shows higher disparity. By economic status, richest (INR 23417) and rich (INR 5669) household shows higher disparity than poorer (INR 2025) and poorest (INR 4637) households. The analysis of gender disparity in intra-household HCE by dependency ratio shows that the disparity was higher among the household having higher dependency ratio. The households which have no (INR 4582) dependency ratio shows less disparity compare to 0.01-0.5 (9962), one and above (INR 6914) dependency ratio. The regional wise analysis shows that North, Central and South region shows higher disparity compared to East and West region. Average expenditure allocation was reported to be higher in Northeast region for females compared to males.

Comparison of gender disparity in average healthcare expenditure (Inter household) and intra-household expenditure, India, 2014

Table 9 demonstrates the comparison of gender disparity in per capita and intra-household HCE by socio-economic and demographic characteristics. The results shows that gender disparity was two times higher within household compared to per capita HCE. Intra-household disparity was more among the higher age group compared to adult age groups whereas the gender disparity in per capita HCE were higher among the adult age group. Gender disparity, per capita and intra- household HCE were higher among the male household head compared to the female head. By literacy, among literate household head gender disparity were higher compared to the illiterate household head. Intra-household disparity was higher among the literate household head. Religion wise analysis shows that the gender disparity in per capita HCE was higher in other

38

Table 7. Average intra household gender disparities in annual health care expenditure for inpatient by background characteristics, India, NSS, 2014

Variable Categories Mean

• Std. Err.

95% [Conf. Interval] Age of hh head <50 3844 1759 394 7295 50 and Above 21086 4049 13145 29027 Sex of hh head Male 14112 2765 8689 19534 Female 13737 4201 5497 21976 Sector Rural 16041 4015 8166 23916 Urban 10771 2472 5922 15620 Education hh head No education 7174 2179 2900 11449 Up to primary 22572 3950 14825 30319 Up to secondary 615 2550

• 4386

5617 Up to higher secondary 24662 10364 4336 44989 Graduate and above 8924 7513

• 5811

23659 Religion Hindu 6524 1693 3204 9843 Muslim 18527 2969 12705 24350 Others 12154 6653

• 895

25204 Social group SC/ST 17202 3131 11060 23344 OBC 2729 3360

• 3861

9318 General 3852 5940

• 7798

15501 Wealth quintile Poorest 5973 1409 3210 8736 Poorer 4018 3671

• 3183

11218 Middle 11727 10116

• 8114

31569 Richer 13505 2199 9193 17818 Richest 24967 5069 15025 34908 Dependence ratio 9377 2654 4172 14582 0.0-0.5 9260 5279

• 1094

19614 .6-1.0 27735 4617 18679 36791 1.0+ 6770 3079 732 12809 Region North 15490 4937 5808 25173 Central 28293 13414 1984 54603 East 4120 3231

• 2216

10457 Northeast 2265 2498

• 2634

7163 West 3364 3641

• 3778

10506 South 17009 3747 9661 24358 Others 5889 2355 1271 10507

religion community whereas Intra-household disparity was higher among Muslim (INR 18527) followed by Other (INR 12154) religious community. By social group, gender disparity in average per capita HCE was higher among OBC (INR 9245) followed by General (INR 8770) caste group whereas intra-household disparity was reported higher SC/ST (INR 17202) followed

39

Figure 14. Gender disparity in Inter and Intra-household healthcare expenditure by Sector, 2014 Figure 15. Gender disparity in Inter and Intra-household healthcare expenditure by literacy status, 2014 Figure 16. Gender disparity in Inter and Intra-household healthcare expenditure by Economic status, 2104

6815 16041 9611 10771 2000 4000 6000 8000 10000 12000 14000 16000 18000 Inter househoold Intra household Rural Urban 4601 7174 10504 14194 2000 4000 6000 8000 10000 12000 14000 16000 Inter househoold Intra household Illiterate Literate

40

Figure 17. Gender disparity in Inter and Intra-household healthcare expenditure by region, 2014

5000 10000 15000 20000 25000 30000 Inter househoold Intra household Poorest Poorer Middle Richer Richest All

• 5000

5000 10000 15000 20000 25000 30000 Inter househoold Intra household North Central East Northeast West South Others India

41

Inter and Intra-household health care expenditure

Table 8 Comparison of gender disparity in Inter and Intra household health care expenditure by background characteristics, India, NSS, 2014

Variables Categories Inter household Intra household Difference Age (in years) 15-49 4106 3844

• 262

50 and Above 12301 21086 8785 Sex of hh head Male 7094.23 14112 7017 Female 14195 13737

• 458

Sector Rural 6815 16041 9226 Urban 9611 10771 1160 Literacy of hh head Illiterate 4601 7174 2573 Literate 10504 14194 3690 Religion Hindu 7539 6524

• 1015

Muslim 7382 18527 11145 Others 10745 12154 1409 Social group SC/ST 3269 17202 13933 OBC 9245 2729

• 6517

General 8770 3852

• 4918

Wealth quintile Poorest 4637 5973 1336 Poorer 2025 4018 1993 Middle 4504 11727 7224 Richer 5669 13505 7836 Richest 23417 24967 1550 Dependence ratio 4582 9377 4794 0.0-0.5 9962 9260

• 703

0.6-1.0 7917 27735 19818 1.0+ 6914 6770

• 144

Region North 11484 15490 4006 Central 9756 28293 18537 East 5449 4120

• 1329

Northeast

• 2532

2265 4797 West 3335 3364 29 South 11325 17009 5684 Others

• 2796

5889 8686 India Total 7731 14075 6343

42

by general (INR 3852) caste group. Analysis by the economic status of the households shows that disparity has been increasing according to the economic status. The disparity was lower among the poorer household despite economic development. Middle, Rich and richest household shows the higher disparity in HCE. By dependency ratio, the household having higher dependence ratio have higher disparity among male and female HCE.

Decomposition Results

Table 10 demonstrates the result of Oaxaca decomposition to examine the contribution of selected background characteristics the male-female difference in HCE. The difference in HCE between male-female was decomposed into three parts : fist part is endowment effect explain the gap due to difference in the distribution of the determinants of HCE between male and female, the second part is the coefficient effect explain the gap due to the difference in the effect of determinants between male and female; and the third is interaction between the both endowments effect and coefficient effect. The analysis shows that the mean value of log OOP expenditure was 8.94 for male and 8.75 female. The difference the mean value log of OOP expenditure is 0.186. The difference in male-female average HCE due to the gaps in endowments (E) was (0.156), due to the coefficient (C) was (0.035) and due to the interaction (CE) was (- 0.006). Furthermore, the decomposition analysis reveals that about 85% male-female difference in HCE is explained by the difference in the distribution of socio-economic and demographic factors. Type of education, type of disease, level of care and hospital duration are contributing towards widening the gender gap in HCE. The contribution of the duration of hospitalization towards widening gap was the highest (about 35 %): it explain that the distribution of duration of stay in hospital is more favorable to males than female. Next was the type of disease (27%), education (26%), Level of care (12%). About 18% of the difference was explained by the effect of determinants (coefficient). The positive contribution by age group, sector, education etc. indicates that effects of these factors are responsible for wider gender gap in HCE. More precisely, the positive contribution of caste (160%) and sector (171%) suggest that effect of OBC and Other caste is more favorable to male than female on HCE. There are a few offsetting factors such as religion of household, type of disease and duration of stay indicating favorable effects of

43

these variable to reduce the male-female gap in HCE. Since the contribution of interaction effect is marginal (about -3%) we did not present the results in detail. Table 9. Oaxaca decomposition: contribution by endowments, coefficients, and interaction to male-female difference in health care expenditure, India, NSS, 2014

Mean prediction high (H): 8.941 Mean prediction low (L): 8.755 Raw differential (R) {H-L}: 0.186 Due to endowments (E): 0.156 (83.87 %) Due to coefficients (C): 0.035 (18.81 %) Due to interaction (CE):

• 0.006 (-3.22 %)

Variables Gap in endowments ( E) Gap in coefficient ( C) Gap arise from endowments and coefficient (EC) % % % Age group

• 0.003
• 1.92

0.0460 131.43

• 0.001

16.67 Sector 0.003 1.92 0.0600 171.43

• 0.001

16.67 Education 0.040 25.64 0.0370 105.71 0.007

• 116.67

Caste 0.000 0.00 0.0560 160.00 0.000 0.00 Religion

• 0.001
• 0.64
• 0.0060
• 17.14

0.000 0.00 Household size 0.000 0.00 0.0190 54.29 0.000 0.00 Economic status 0.002 1.28

• 0.0210
• 60.00

0.000 0.00 Type of disease 0.042 26.92

• 0.0400
• 114.29
• 0.003

50.00 Level of care 0.019 12.18 0.0410 117.14 0.000 0.00 Duration of stay 0.055 35.26

• 0.0480
• 137.14
• 0.008

133.33 Const. 0.000

• 0.110
• 314.29

0.000 0.00 Total 0.156 100 0.035 100

• 0.006

100 % explained (V/R): 84.1 18.81

• 3.22

Discussion and Conclusion

There has been a resurgence of literature of interest in recent years in the issue on gender differentials in health outcomes, such as mortality and nutrition, to a certain extent, among women in reproductive age groups. It is found that health expenditure is growing faster for families than overall expenditure (Mohanty et al. 2016. India, going through the end of the third stage of demographic transition implying the proportion of aged population is increasing continuously due to fertility decline and increasing life expectancy. This leads to a higher

44

burden of the chronic disease older age of life. Due to the changing age and disease pattern, share of non-communicable burden of diseases is increasing. It is important to note that these diseases are more expensive in terms of treatment. Using recently available nationally representative data, this study examines nature and extent of gender difference in HCE by demographic and socio-economic characteristics in India. This study examines gender disparity in average and in Intra-household HCE. Gender disparity in average health care expenditure Finding of this study suggests that there is a huge difference in the average HCE between male and female. Average HCE significantly lesser for female compared to male in 2004 and 2014. The difference in average HCE increased in 2014. Figure (18) shows the ratio of average HCE for male and female in 2004 and 2014. Figure 18. Ratio of male and female healtheare expenditure by age group, 2004, 2014 Figure 19. Average annual healthcare expenditure by sex and age group during 2004 and 2014

45

As figure 18 illustrates, the disparity was higher in child age group in 2004, and it reduced in

• 2014. Figure 19 shows that during 2004 the difference was higher in 10 -25 and 35 to 40 year

age group, and again it has increased in older age group. Whereas during 2014, average HCE was in the favor of girls for 0-5 age group and the difference was higher 20-40 year age group. Among the older age group difference between male-female HCE shows a drastic increase in the favor of the male. This indicates that among older age group female are facing discrimination in health care which leads to the poor health status of older women. One more explanation for higher disparity older age group is that in higher age group main causes of hospitalization is chronic diseases that have a higher cost of treatment. So due to gender discrimination in HCE female incurred less in higher age group. Figure 6 shows the gender disparity in HCE by type of disease. It shows that disparity between male and female is less for communicable disease and compare to non-communicable and other disease because their high cost of treatment. Gender disparity in Intra-household health care expenditure Gender difference in Intra-household HCE shows the clear picture of gender discrimination in resource allocation within the household. A study done by Asfaw (2010) shows that parents are less likely to be hospitalized girls child using onerous sources as a health finance. Parents dig more deeply into their pockets to hospitalize their sons than their daughters. Studies across India

5000 10000 15000 20000 25000 Male_04 Female_04 Male_14 Female_14

46

have found that boys are much more likely than girls to be taken to a health facility when sick (Das Gupta, 1987). Gender disparity in average HCE for males is around seven times higher in 50 and above age group compared to less than 50 age group. It is higher in rural areas compared to urban areas. The difference in the amount of average HCE of male-female was higher in economically batter off households. Findings of this study are consistent with those study conducted in India and some Asian countries (Sen, 2002; Pandey, 2002; Maharana, 2014; Ladusing, 2013; Batra 2014, Song 2014; Iyer 2008.) Gender difference in HCE for communicable (INR 63 in constant price) and non- communicable (INR 4418 in constant price) estimated in this study is significantly higher than the average difference between male-female for short-term morbidity (INR 57 and major morbidity (INR 2781) estimated by Saikia et al. (2016). The reason for the difference is that our study based on inpatient care but the study done by Saikia at al. used both inpatient and

• utpatient.

Decomposition results suggest that about 84% gender difference in HCE is due to male-female difference in socio-economic, demographic and healthcare-related factors. Type of education, type of disease, level of care and duration of stay in hospital are contributing towards widening the male-female gap in HCE. And 18 % difference in male-female HCE is due to the effect of these factors on HCE. The positive contribution by age group, sector, education etc. indicates that effects of these factors are responsible for wider gender gap in HCE. Thus the contribution by coefficient represents the genuine role of gender in OOP expenditure, i.e., less is spent on female health because of the notion that female health is not as important as male health. Results of this study are consistent with those of studies conducted in India. Saikia et al. (2016) found that even after eliminating the role of demographic and socio-economic factors, female HCE is less than that of male. This study also reveals that nearly half of the gender difference in OOP expenditure is due to male–female differences in demographic, socio-economic, and health care-related factors; the rest is due to the effects of these factors on OOP expenditure. Our study, however, contradicts this finding and concludes that nearly 85 percent of the gap in HCE is explained by the differential distribution of socio-economic, demographic and disease-related

• variables. This may be due to the restriction of our analysis to the in-patient care expenditure.

Also, unlike the study by Saikia, Moradhvaj and Bora (2016), we included all the non-

47

communicable and communicable disease expenditure in this analysis. Ladusingh and Pandey shows (2013) that the inpatient care OOP expenditure for females is less than that for their male counterparts when the other effects of other correlates such as age, type of healthcare facility, type of disease and duration of hospitalization are adjusted and the difference is statistically

• significant. This reflects the discriminatory social practices in intra-household resource

allocation. Gender disparity in HEC is a critical challenge to improve the women health status in India. To reduce the gap between male-female OOP expenditure, we need to economically empower the women through improving education status and changes in gender attitude. Out of pocket HCE in India constituted 12.2 % of household consumption expenditure and 21.7 % of household non- food expenditure in 2004-05 (Ladusingh and Pandey, 2013) this is one of the highest OOP expenditure in the world. There is a need to reduce the gap between male-female OOP expenditure through improving the quality of care and providing subsidies in public hospitals to targeting gender discriminatory behavior against women. Gender disparity in HCE is varied among the all-India states. Northern states such as Punjab Haryana, Uttar Pradesh and Bihar and Southern states Kerala and Karnataka shows higher gender disparity whereas some in states such as Assam, Chhattisgarh, Jharkhand, Manipur Meghalaya, Mizoram and Uttaranchal shows less disparity between male-female HCE for

• hospitalization. Figure 20 shows the relationship between coverage of Rashtriya Swasthya Bima

Yojana (RSBY) among poor people and gender difference in average HCE. The figure clearly indicate that with the high coverage of RSBY states shows less disparity in HCE between male and female compare to less coverage states (correlation=-0.6). Figure 20. The Relationship between gender disparity in healthcare expenditure by states and the percentage coverage of RSBY.

48

The present study examines the effect of gender on household resource utilization for inpatient care in India. Further research can be done to examine the gender disparity in Intra-household HCE by type of disease that is major determinants of expenditure. This study found a huge difference in HCE between male-female. So, there is a need to much more work to understand the gender difference by type of illness.

Reference

Anderson, S., & Ray, D. (2009). Missing women: age and disease. The Review of Economic Studies, 77(4), 1262-1300. Anderson, S., & Ray, D. (2012). The age distribution of missing women in India. Economic and Asfaw, A., Klasen, S., & Lamanna, F. (2007). Intra-household gender disparities in children's medical care before death in India. Institute for the Study of Labor (IZA) Discussion Paper, (2586). Asfaw, A., Lamanna, F., & Klasen, S. (2010). Gender gap in parents' financing strategy for hospitalization of their children: evidence from India. Health economics, 19(3), 265-279. Batra, A., Gupta, I., & Mukhopadhyay, A. (2014). Does discrimination drive gender differences in health expenditure on adults: evidence from Cancer patients in rural India. Indian Statistical Institute Discussion Paper, (14-03).

• 10

10 30 50 70 90

• 6000
• 4000
• 2000

2000 4000 6000 8000 10000 12000 14000

Gender diff (OOP exp) % RSBY

49

Borooah, V. K. (2004). Gender bias among children in India in their diet and immunisation against disease. Social Science & Medicine,58(9), 1719-1731. Canudas-Romo, V., Saikia, N., & Diamond-Smith, N. (2015). The contribution of age-specific mortality towards male and female life expectancy differentials in India and selected States, 1970-2013. Asia-Pacific Population Journal, 30(2). Mohanty, S. K., Ladusingh, L., Kastor, A., Chauhan, R. K., & Bloom, D. E. (2016). Pattern, growth and determinant of household health spending in India, 1993–2012. Journal of Public Health, 1-15. Pandey, A., Sengupta, P. G., Mondal, S. K., Gupta, D. N., Manna, B., Ghosh, S., ... & Bhattacharya, S. K. (2002). Gender differences in healthcare-seeking during common illnesses in a rural community of West Bengal, India. Journal of Health, Population and Nutrition, 306-311. Saikia, N., Jasilionis, D., Ram, F., & Shkolnikov, V. M. (2011). Trends and geographic differentials in mortality under age 60 in India. Population studies, 65(1), 73-89. Saikia, N., Moradhvaj, Bora, J. K. (2016). Gender Difference in Health-Care Expenditure: Evidence from India Human Development Survey. PLOS ONE, 11(7), e0158332. Sudha, S., & Rajan, S. I. (1999). Female demographic disadvantage in India 1981–1991: Sex selective abortions and female infanticide.Development and change, 30(3), 585-618. Behrman, J. R. (1988). Intra-household allocation of nutrients in rural India: Are boys favored? Gupta, M. D. (1987). Selective discrimination against female children in rural Punjab, India.Population and development review, 77-100. Gupta, M. D. (2005). Explaining Asia's “missing women”: a new look at the data. Population and development review, 31(3), 529-535. Sauerborn, R., Berman, P., & Nougtara, A. (1996). Age bias, but no gender bias, in the intrahousehold resource allocation for health care in rural Burkina Faso. Health Transition Review, 131-145.

50

Seeley, J. A., Kajura, E. B., & Mulder, D. W. (1995). Methods used to study household coping strategies in rural South West Uganda. Health Policy and Planning, 10(1), 79-88. Skarbinski, J., Walker, H. K., Baker, L. C., Kobaladze, A., Kirtava, Z., & Raffin, T. A. (2002). The burden of out-of-pocket payments for health care in Tbilisi, Republic of

• Georgia. Jama, 287(8), 1043-1049.

Sood, A. K., & Nagla, B. K. (1994). The extent and pattern of utilisation of health services by rural women: A study in District Rohtak, Haryana.Indian Journal of Preventive and Social Medicine,25(3&4), 110-117. Sudha, S., & Rajan, S. I. (1999). Female demographic disadvantage in India 1981–1991: Sex selective abortions and female infanticide.Development and change, 30(3), 585-618. United Nations Development Program, (1995). Human Development Report 1995, UNDP New York, USA Waldron, I. (1995). Contributions of changing gender differentials in behaviour to changing gender differences in mortality in D. Sabo and D. Gordon. Men's Health and Illness: gender, power and the body. London: Sage.

• WHO. (1998). Gender and Health. A Technical paper. http://www.who.int/frh-whd/index.html

Wilkes A, Hao Y, Bloom G, Xingyuan G. (1997). Coping with the costs of severe illness in rural

• china. IDS Working Paper 58. Institute of Development Studies, Brighton.