ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Income Inequity in the Utilisation of Inpatient Services

Effects of Publicly Financed Health Insurance Schemes

Vanita Singh ( teaches economics and public policy at the Management Development Institute, Gurugram. Amit Garg teaches in the Public Systems Group at the Indian Institute of Management, Ahmedabad.

Existing literature on the effect of publicly financed health insurance schemes focuses on out-of-pocket expenditure and utilisation of health services, while the effect of PFHIs on equity in health service use remains unaddressed. The data on equivalised household income and inpatient service utilisation from the National Sample Survey Office are employed to analyse the extent of inequity in the utilisation of inpatient services before (2004) and after (2014) the implementation of the PFHI in 2008. This is done by estimating concentration indices and the horizontal inequity indices.

Publicly financed health insurance (PFHI) schemes are a major healthcare reform that have been adopted across developing countries, including India, to achieve universal health coverage (UHC). PFHIs reduce healthcare access inequity by targeting poor people and ensuring need-based access (Kutzin 2012; Roberts et al 2015; Kieny et al 2017). Equity is defined as the inequality that remains after accounting for legitimate factors driving inequality. Legitimate factors are those sources of variation in health or health-seeking behaviour that are “fair” or “just” (Fleurbaey and Schokkaert 2009), such as age, sex, and presence of illness. On the contrary, illegitimate determinants are “unfair” sources of variation such as socio-economic status (Wagstaff and van Doorslaer 2004; Fleurbaey and Schokkaert 2009).

In 2008, the Indian government launched the Rashtriya Swasthya Bima Yojana (RSBY), a government-run health insurance programme for the poor followed by the launch of a National Health Protection Scheme in 2018, an extension of the existing PFHI scheme (Ghosh 2018). PFHI schemes across developing countries target vulnerable groups to achieve horizontal equity in the utilisation of health services (Giedion et al 2013; Roberts et al 2015). Horizontal equity in the utilisation of health services is defined as equal utilisation for equal needs irrespective of socio-economic factors (Culyer and Wagstaff 1993; WHO 2010; Kien et al 2014; Roberts et al 2015; Saito et al 2016).

Thus, it is important to study the effect of PFHIs on equity in health service utilisation (Giedion et al 2013) to assess whether they have achieved this goal. Studies, especially in the Indian context, have not focused on the equity aspect of utilisation of inpatient health services under insurance schemes (Giedon et al 2013; Ghosh 2014b, 2018).

Studies that have focused on the utilisation of health services in India (Ghosh 2014a; Barik and Thorat 2015) as well as in other countries (Kien et al 2014; Saito et al 2016) have reported differences in the utilisation of health services between the rich and poor. However, the existing literature on equality in health services in the Indian context is mainly descriptive and fails to reflect upon the potential association of UHC measures with equity. Moreover, the recent announcement of the Indian government’s intention to expand the existing PFHI scheme under the umbrella of “Ayushman Bharat,” mandates examining the role of PFHI schemes in achieving equity, given the considerable amount of public health funding being directed towards these schemes (Ghosh 2018). Thus, the current study provides timely empirical evidence for policymakers on the role of PFHIs in advancing horizontal equity.

The study focuses on inpatient service utilisation since the main aim of PFHIs in India was to increase access to inpatient services for poor people as inpatient services were previously found to lead to catastrophic health expenditures for poor families (Devadasan et al 2013; Ghosh 2014a; Ghosh and Gupta 2017). The annual hospitalisation rate for each individual is used as a proxy to evaluate the utilisation of inpatient services. Income-related inequalities in healthcare use are widely acknowledged in developed as well developing countries. Although these inequalities have been measured in developed countries, studies focusing on income-related inequalities in health service use are lacking in an Indian context.

Publicly Financed Health Insurance Schemes in India

In 2007, Andhra Pradesh (AP) became the first Indian state to launch a PFHI scheme. In 2008, the central government launched the RSBY across all states. The scheme was rolled out in a phased manner and states had discretion in implementing the scheme. The objective was to protect poor families (those below the poverty line [BPL]) from catastrophic health expenditures arising from hospitalisation and ensure improved access to inpatient services for poor people thus increasing improving equity in the utilisation of health services (Palacios 2011; Selvaraj and Karan 2012; Ghosh 2018). The federal structure of India gives autonomy to states to design their state-specific health schemes. In the following years, Tamil Nadu (TN), Karnataka, Kerala, Chhattisgarh, Rajasthan, and Maharashtra either extended RSBY coverage or started their own state-specific PFHI (Selvaraj and Karan 2012). The state-specific PFHIs were similar in design to RSBY, except for their names and the coverage amount. Thus, we expect the impact of RSBY/PFHIs would be similar in terms of improving equity in accessing inpatient services.

For the state-level analysis, we focus on 21 major states (out of 29 states and seven union territories) constituting 98.44% of the population (Table A1 in the appendix, p 47). The smaller states were not included to ensure reliable estimates as the National Sample Survey Office (NSSO 2015) report itself mentions that due to very small sample size, estimates from the smaller states may be unreliable. We track the PFHI/RSBY implementation for these 21 states and found that out of these 21 major states, only 16 had 100% of their districts covered by PFHIs/RSBY by 2013 (Table A1). We focus on these 16 states as the sample data was collected for June 2013 to June 2014. The 16 states are AP, Bihar, Delhi (one of the major union territories), Goa, Gujarat, Haryana, Himachal Pradesh, Jharkhand, Karnataka, Kerala, Madhya Pradesh (MP), Punjab, Rajasthan, TN, Uttar Pradesh (UP), and West Bengal. Out of these 16 states, 13 had implemented the RSBY, while the others had their own state-specific PFHIs. The RSBY was launched in 2008 as a social security scheme by the Ministry of Labour and Employment, and in 2015, was transferred to the health department. The RSBY is a centrally sponsored social programme aimed at providing cashless hospitalisation services amounting up to 30,000 to poor people. The central budget covers three quarters of the cost of the programme and thus states have incentives to join the scheme (Palacios 2011). For each of these 16 states, we compare the horizontal inequity index before and after the implementation of PFHIs (Table 3, p 44). Of the 16 states, Assam, Bihar, Chhattisgarh, Jharkhand, MP, Rajasthan, UP, and Uttarakhand lag far behind in health outcome as well as development indicators (MoHFW 2011). We refer to these as Empowered Action Group (EAG) states.


We use individual-level data (excluding deceased members) from the 60th (Morbidity and Healthcare 2004) and 71st rounds (Social Consumption: Health 2014) of the NSSO. The data for each round are comparable as both rounds collected information on “whether the person was hospitalised in the last 365 days” and recorded the ailment for which treatment was taken. Both rounds collected information on morbidity, particulars of inpatient and outpatient treatment in the previous year and last 15 days, respectively. This information was collected from a nationally representative sample of 3,83,338 individuals in 2004 and 3,33,104 individuals in 2014. Both rounds used a multistage stratified random sampling method. The details of the sampling methodology, questionnaire, definition of variables, and initial findings can be found in reports prepared by the Ministry of Statistics and Programme Implementation (MOSPI 2006, 2015).


We examine the inequality and inequity in the utilisation of inpatient services across income quintiles. The income quintiles were based on the adjusted monthly household consumption expenditure (MHCE). The MHCE was adjusted by the household size based on the methodology proposed by the Organisation for Economic Co-operation and Development (OECD 2008, 2011). The utilisation of inpatient services is defined as the use of any health facility for taking in-house treatment in the previous 365 days. We define the “inequality in inpatient service use” as the difference observed in the mean values of inpatient service use across income quintiles.

We define “inequity in inpatient service use” as the deviation of actual health utilisation from standardised utilisation, that is utilisation driven by the need factors alone with non-need determinants of health service utilisation held constant (Wagstaff and van Doorslaer 2004). We define need based on the individual’s age, sex, and presence of non-communicable diseases (O’Donnell et al 2008; Qian et al 2017; Pan et al 2017). The non-need factors affecting health service utilisation that are included in our analysis are state identifiers, rural/urban residence, literacy level, occupation of the head of household, marital status, health insurance status, and income (proxied by consumption expenditure). The non-need factors are selected in the model of health service utilisation based on Newman– Anderson model of health-seeking behaviour (Anderson and Newman 2005). Wagstaff and van Doorslaer (2004) advocated for the indirect standardisation method for the measurement of horizontal inequity as it does not require grouped data and is computationally easier. Thus, for our study, we have adopted the indirect standardisation method.

Indirect standardisation of health variable: The linear relationship between inpatient service utilisation, and need factors and control variables is represented by the following equation:

yi = G (α+∑jβj χji+ ∑kγk zki) + ϵi … (1)

where yi is healthcare utilisation variable; i denotes the individual, α, β, and γ are parameter vectors; χji is the individual values of the J (j=1,…J) confounding variables (need); and zki are individual values of the K (k=1,…K) non-confounding (control) variables. When we use a linear model, the indirectly standardised utilisation (iîs) is given by the difference between actual utilisation (yi) and need-based expected utilisation (iX), plus the mean of actual utilisation ȳ.

iîs = y– yˆi+ ȳ

We analyse the concentration curve and measure the concentration index (CI) of inpatient service use to assess if there is pro-rich or pro-poor inequality in the utilisation of health services. We focus on inpatient service use as PFHIs only cover inpatient services.

The concentration curve: The concentration curve is the bivariate analogue of the Lorenz curve as it plots the cumulative proportion of one variable against the cumulative proportion of the population ranked by another variable (Kakwani et al 1997; Koolman and van Doorslaer 2004; O’Donnell et al 2016). In our study, to visualise the inequality in the utilisation of inpatient services, we plot the cumulative percentage of the inpatient utilisation (on y-axis) against the cumulative percentage of the population, ranked by household per capita monthly expenditure, from poorest to the richest (on x-axis). If everyone, irrespective of their income, has exactly the same value of the health variable, the concentration curve will lie along a 45-degree line, known as the line of equality. If the health variable is more concentrated among poorer (richer) people, the concentration curve will lie above (below) the line of equality.

Concentration index: Concentration indices are commonly used for measuring socio-economic-related inequality in health (O’Donnell et al 2008; O’Donnell et al 2016). The standard concentration index as proposed by Kakwani et al (1997) can be written as:

where N is the sample size, his the health variable for person i, µ is the mean of the health variable, and ri is the fractional rank in the income distribution of the ith person.

Horizontal inequity index: We measure horizontal inequity using the horizontal inequity index (HI) for inpatient services utilisation to assess the effect of PFHIs on equity in inpatient service use. The HI indicates health inequality attributable to illegitimate factors. When using linear models, it is given by the difference between the concentration indices for actual utilisation (Ca) and need-standardised utilisation (Cn) (O’Donnell et al 2008: 182).

HI = Ca–Cn

The HI ranges between -2 and 2, and a value of zero indicates utilisation is according to need, that is, there is no inequity. A positive (negative) value of HI indicates the presence of inequity that is pro-rich (pro-poor) after controlling for need.

Data analysis: Data were analysed using the Stata 15 statistical software package, and estimates were weighted to account for the multistage stratified sampling design (O’Donnell et al 2008). We used bivariate and multivariate linear regression analyses to study the income-based inequity in the utilisation of health services. We compared the means of actual health service utilisation and need-standardised health service utilisation across income quintiles before and after the reform for India, as well as for urban and rural India separately. We used concentration curves and concentration indices (CI) to assess the degree of inequity in the healthcare use and compared these across two time-periods (before and after the introduction of PFHIs). Furthermore, we carried out an interstate analysis to explore the role of the health system in reducing inequity. The sample in each state has been categorised as poor and non-poor based on the Planning Commission’s estimates of statewise urban and rural cut-offs for monthly per capita consumption expenditure (Table A2 appendix, p 47). For each of these 16 states, we assess the effectiveness of targeting under PFHIs by estimating the proportion of poor people (those BPL) covered by PFHIs. As PFHIs target poor families, we expect that the states with effective targeting under PFHIs would have lower inequity in inpatient service use. The access to inpatient services, especially for poor people, also depend on the public health infrastructure in the state. The states also vary in their public health infrastructure, bed to population ratio and doctor to population ratio. To assess the correlation between the proportion of poor covered by PFHIs, health system indicators and HI, we do canonical correlation analysis (CCA). CCA analyses the relationship between the dependent (IMR and HI) and independent (proportion of poor people covered by PFHIs and doctors and beds per 1,000 population) variables that can be linearly combined in different ways (dimensions).

Results and Discussion

Descriptive statistics: Descriptive statistics are presented in Table 1 (p 42). The mean age of the Indian population increased from approximately 26 in 2004 to 29 in 2014. The economically active population (15–59) has increased from 58% in 2004 to 63% in 2014, while the dependent age group (0–14 years) has considerably reduced from 35% to 29% in 2014 (Table 1). The increased life expectancy is represented by the increase in the proportion of the population aged above 60 years (7% in 2004 to 8% in 2014). The sample age-sex distribution is similar to that reported in the census reports of 2011 by the Government of India (2011), supporting the representativeness of our study sample. The proportion of males is higher in both years (51.2% in 2004 and 51.4% in 2014) and the majority of the population (75% in 2004 and 70% in 2014) resides in rural areas, although this reduced somewhat by 2014. The persons reporting the presence of non-communicable diseases (NCDs) have shown significant increase from 3.2% to nearly 6% during 2004 to 2014. The increased reporting of NCDs is suggestive of the increase in the need for inpatient care as we have used the presence of NCDs as a proxy for the need for inpatient care. This increase can be attributed to increased awareness about NCDs, increased access to diagnostics and shift in disease pattern from communicable to NCDs. The variables that have shown significant increases between the periods include the proportion of persons with health insurance (increasing from 0.6% to 15.1%), the proportion completing secondary education (increasing from 7% to 46%) and the proportion of salaried persons in the sample (increasing from 10% to 18%). All these variables positively affect access to healthcare and thus we do not make any causal claims for PFHIs and inequity reduction in our study. In the analysis of equity differences between the two study periods, we control for all these differences while estimating horizontal inequity index for both the years.

Income inequality in the utilisation of inpatient services: The annual inpatient rate (defined as the percentage use of inpatient services at the individual level over last one year) has almost doubled from 2.4 (in 2004) to 4.4% in 2014. Figure 1 displays the proportion of the Indian population reporting inpatient service use by income status in 2004 and 2014. The annual inpatient rate is highest for the richest 20% of India’s population for both the years (2004: 3.7%; 2014: 5.9%). However, the inequality seems to be reducing as the percentage increase in inpatient service use is higher for the poorest quintile relative to the richest quintile. This result is indicative of increased inpatient service use among poorer people.

The decline in inequality is also reflected in our analysis of concentration curve and concentration indices (CIs) (Table 2, p 43). Figure 2 (p 43) compares the concentration curves of actual inpatient service use for 2004 and 2014. It reflects a reduction in inequality, although utilisation remains pro-rich as the curve lies below the line of equality. A dominance test (O’Donnell et al 2008) confirms that the distribution of the inpatient service use is less pro-rich in 2014 than it was in 2004. These findings are further supported by the analysis of CI for 2004 and 2014. For both the study periods, the CI values are positive (Table 2), suggesting that the inpatient service use is concentrated in richer quintiles in India. However, it became less positive in 2014. In 2004, the CI for actual utilisation was 0.165 for India (mean of actual utilisation: 0.024), which reduced to 0.121 (mean of actual utilisation: 0.044) in 2014, suggestive of reduced, but not eliminated, pro-rich inequality. When we compare rural and urban India, we find that rural areas have higher inequality (higher CI values) in both the years but the reduction in inequality is higher in rural areas. In 2004, the CI for rural areas was 0.165 (mean = 0.022) that significantly reduced to 0.136 (mean = 0.042) while for urban areas the CI reduced from 0.082 (mean = 0.03) to 0.069 (mean = 0.043). These CI values suggest that the proportion of people utilising inpatient services tends to be less concentrated amongst the rich over time and the reduction in concentration is larger for rural areas. Figure 3 clearly suggests that although urban areas are more equitable (the curve is closer to line of equality for both the years), the reduction in inequality is higher for rural areas. At the same time, the mean of inpatient service use has also increased, and the increase is higher for rural areas. Analysing these two changes together, we can say that the increased utilisation over time has disproportionately benefited the poor in rural areas relative to urban areas. Studies that have looked at the impact of PFHIs/RSBY in India also reported positive impact of PFHIs on health service utilisation for rural areas but not for urban areas (Azam et al 2016).

Income inequity in the utilisation of inpatient services: Figure 2 illustrates the difference between actual and need-standardised utilisation across income quintiles. Ideally, we would like the two utilisations (actual and standardised) to be as close as possible, as the actual use should be dependent on need alone and independent of non-need factors. The graphs coincide for middle-income quintiles but not for the poorest and the richest quintiles in both the years. For the poorest quintiles, the need-standardised use is higher than the actual use while for the richest quintiles the need-standardised use is lower than the actual use in both the years. These results are suggestive of “illegitimate (income-based) inequality” or inequity in inpatient service use for both the years. The poorer quintiles fall short of accessing care based on their needs while the richer quintiles are advantaged as there is an indication of over-utilisation of health services. Furthermore, we explore the CI values for actual and need-standardised inpatient service use (Table 2).

The CI values can measure inequality but to comment on inequity we need to account for the inequality of need-standardised use (O’Donnell et al 2008) and instead consider the horizontal inequity index given by the difference of actual and need-standardised CIs. The CI values for need-standardised utilisation are closer to zero which indicates that the need-standardised use is more equally distributed across income quintiles. The inequality in actual inpatient service use is pro-rich, although reduced in 2014. However, the inequality in need-standardised inpatient service use has become more pro-rich. This may be due to higher reporting of NCDs by richer quintiles as compared to poorer people. The reporting of NCDs may also suggest greater access to diagnostics and healthcare for those reporting NCDs (here, richer people).

The horizontal inequity analysis (Figure 4, p 44) for India suggests that inequity in inpatient service use has become less pro-rich post PFHI implementation (2004 HI: 0.158; 2014 HI: 0.112), and the reduction in inequity is greater for rural areas (2004 HI: 0.157; 2014 HI: 0.119). The actual use and need-standardised use show that inequality in urban areas, as measured by the CI, has remained largely unchanged with no statistically significant difference detected between 2004 and 2014. Since need is being proxied based on age, sex and the presence of NCDs, the pro-rich concentration of need-expected use suggests that the reporting of NCDs has increased among richer quintiles in 2014. It is possible that if need were represented by some other disease category (say communicable disease), the results of our analysis may differ. However, we chose NCDs as WHO (2015) reports that 61% of the mortality in India is attributed to NCDs and it is mainly due to a lack of access to health services.

Interstate analysis: Next, we compared inequality across states and across years using concentration indices for actual and need-standardised use for each state in 2004 and 2014 (Figures 5a and 5b, respectively) The need-standardised use is more or less equally distributed across income quintiles as the CI for need-standardised use is much closer to zero for both years and for all the states. However, the actual-use CI varies considerably across states in both years. In 2004, the actual use was pro-poor (negative CI) in Goa and Kerala, while it was almost equal (CI approaching zero) in Delhi and Himachal Pradesh (Figure 5a). For all other states, pro-rich inequality is observed in 2004. The inequality became more pro-rich in 2014 for Delhi, Haryana, Jharkhand, and up, while for all other states, the inequality reduced in the post-PFHI implementation period. To comment on the inequity, we estimate the horizontal inequity across states. Table 3 reports the horizontal inequity index across the study states before and after PFHI implementation. For most of the states, the inequity index shows significant reduction in the post-implementation year, with the exceptions of Delhi, Haryana, and Himachal Pradesh.

These are the states where the proportion of poor covered by PFHIs is almost nil. Studies that have analysed the performance of RSBY/state-run PFHIs found that targeting is weak, as the list of BPL households is either not updated or manipulated by socially advantaged people (Ghosh 2018). This could also explain the persistent pro-rich inequity in the utilisation of inpatient services post RSBY/PFHI. This study highlights the role of effective targeting and responsive public health system to achieve the goal of equity in accessing healthcare.

Canonical correlation analysis: The CCA reports that these variables are significantly linearly related on a single dimension with an effect size of (0.71) suggesting 71% of the variation in the variables explained by that dimension (Table 4a). The canonical correlation results (Tables 4a and 4b) suggest that states having lower inequity indices along with lower IMRs tend to have higher bed to population ratios and greater proportions of poor covered by PFHIs. Plotting the canonical variate scores for each variable (Figure 7), we can identify states performing well along both dimensions. Quadrant 1 (bottom-left) represents the states with lower IMR and lower inequity with better health infrastructure and higher coverage of poor people under PFHIs. All the states that lie in quadrant 1—Goa, Kerala, TN, West Bengal, Gujarat, and Karnataka—are high-performing states (as per MoHFW 2011). AP, the pioneer of PFHI in India, has the highest coverage of poor people but it is still far from being among the best performing quadrant as its public health infrastructure is relatively poor, as represented by its lower doctor to population and bed to population ratio. On the contrary, Rajasthan, despite having a very high IMR, representative of a poor performing health system, can be at par with AP as the proportion of its poor covered by PFHIs is high (33%) and its public health infrastructure is far better than most of the poor performing states. Under PFHIs, people can access both empaneled private as well as public health facilities but it has been found that the use of public health facilities is concentrated among poor people (NSSO 2015; Nandi et al 2017; Pandey et al 2017). Thus, for the CCA, we focus on public health infrastructure instead of total or private health infrastructure.

Limitations: Our study is not without limitations. First, the need-standardised utilisation is based on self-reported morbidity and the utilisation itself may have been affected by the perception of need. There could be bias in the measurement of inequality due to differences in the conception of illness across income levels. However, researchers have found that poor people report morbidity less often when compared with the rich (O’Donnell et al 2008), suggesting our estimates of the degree of pro-rich inequity may be conservative. Second, we have used two time periods, 2004 and 2014, to capture the association of inequity with PFHIs. Post 2004, there have been many reforms in sectors other than healthcare, which coupled with economic growth can also explain the reduction in inequity. Therefore, we do not make any claims of causality, and such claims may best be supported through natural experiments.


PFHIs were launched with the objective of increasing access to inpatient services and reducing inequities in the utilisation of health services (Palacios 2011; Ghosh 2014b; Ghosh and Gupta 2017; Karan et al 2017). Our findings indicate a positive effect of PFHIs on the equality of utilisation of inpatient services, with the pro-rich inequality declining more for rural areas in the post-implementation phase. There is an overall increase in the utilisation of inpatient services which is higher for the poorer quintiles and rural areas, consistent with a positive effect of PFHIs targeting poor people to help them access inpatient services when in need. The higher healthcare access inequity has been found to be associated with poor public health infrastructure, represented by bed and doctors’ density in public health system and poor performance of public health system, indicated by higher IMR. The lower levels of inequity are associated with higher coverage of poor population under PFHIs. This study has highlighted the role of effective targeting and responsive public health system in reducing inequity. There is need for sustained efforts to reduce healthcare access inequity by ensuring access to poor people either through PFHIs or through strengthened public health system. Perhaps the recent launch of Ayushman Bharat by the Indian government is one step towards sustaining the efforts towards inequity reduction, the results of which are yet to be seen. India’s latest PFHI, the Pradhan Mantri Jan Arogya Yojana (PMJAY), is an extension of the RSBY. Its improved design features in terms of who is covered, what is covered and how much is covered might ensure greater equity reduction. The higher coverage amount (5 lakh) would ensure that poor people are not deterred from seeking care due to fear of out-of-pocket expenditure. It covers both secondary and tertiary care, while the RSBY was restricted to secondary care. Another important feature of the new scheme is that it is an entitlement-based scheme. The beneficiaries are already defined based on the Socio-economic and Caste Census survey of 2011 list; while in the RSBY poor people as identified by their respective states had to enrol for the scheme with an amount of 30. The enrolment was valid for one year and every year there was need for renewal. It was found that many poor people failed to renew their cards in subsequent years. The improved design features of the PMJAY suggest that government has planned the scheme to serve the poor people, provided, it is implemented in the same spirit which would help in reducing inequity.


Andersen, R and J F Newman (2005): “Societal and Individual Determinants of Medical Care Utilization in the United States,” Milbank Quarterly, Vol 83, No 4, pp 1–28.

Azam, W, E van de Poel and P Panda (2016): “Analyses of Enrolment, Dropout and Efectiveness of RSBY in Northern Rural India,” MPRA Paper No 70081, Paper No 70081, viewed on 29 May 2022.

Barik, D and A Thorat (2015): “Issues of Unequal Access to Public Health in India,” Frontiers in Public Health, viewed on 29 May 2022.

Culyer, A J and A Wagstaff (1993): “Equity and Equality in Health and Healthcare,” Journal of Health Economics, Vol 12, No 4, pp 431–57.

Devadasan, N, T Seshadri, M Trivedi and B Criel (2013): “Promoting Universal Financial Protection: Evidence from the Rashtriya Swasthya Bima Yojana (RSBY) in Gujarat, India,” Health Research Policy & Systems, Vol 11, No 1, pp 1–8.

Fleurbaey, M and E Schokkaert (2009): “Unfair Inequalities in Health and Healthcare,” Journal of Health Economics, Vol 28, No 1, pp 73–90.

Ghosh, S (2011): “Catastrophic Payments and Impoverishment Due to Out-of-pocket Health Spending,” Economic & Political Weekly, Vol 46, No 47, pp 63–70.

— (2014a): “Publicly Financed Health Insurance for the Poor: Understanding RSBY in Maharashtra,” Economic & Political Weekly, Vol 49, Nos 43–44, pp 93–99.

— (2014b): “Equity in the Utilization of Healthcare Services in India: Evidence from National Sample Survey,” International Journal of Health Policy and Management, Vol 2, No 1, pp 29–38.

— (2018): “Publicly Financed Health Insurance Schemes,” Economic & Political Weekly, Vol 53, No 23, p 17.

Ghosh, S and N D Gupta (2017): “Targeting and Effects of Rashtriya Swasthya Bima Yojana on Access to Care and Financial Protection,” Economic & Political Weekly, Vol 52, No 4, pp 61–70.

Giedion, U, E Andrés Alfonso and Y Díaz (2013): “The Impact of Universal Coverage Schemes in The Developing World: A Review of the Existing Evidence,” World Bank: Washington, DC.

Kieny, M P, H Bekedam, D Dovlo, J Fitzgerald, J Habicht, G Harrison et al (2017): “Strengthening Health Systems for Universal Health Coverage and Sustainable Development,” Bulletin of the World Health Organization, Vol 95, No 7, p 537.

Kakwani, N, A Wagstaff and E van Doorslaer (1997): “Socio-economic Inequalities in Health: Measurement, Computation, and Statistical Inference,” Journal of Econometrics, Vol 77, No 1, pp 87–103.

Karan, A, W Yip and A Mahal (2017): “Extending Health Insurance to the Poor in India: An Impact Evaluation of Rashtriya Swasthya Bima Yojana on Out-of-pocket Spending for Healthcare,” Social Science & Medicine, Vol 181, pp 83–92.

Kien, V D, H V Minh, K B Giang, L Weinehall and
N Ng (2014): “Horizontal Inequity in Public Health Care Service Utilization For Non-communicable Diseases in Urban Vietnam,” Global Health Action, Vol 7, No 1, p 24919.

Koolman, X and E Van Doorslaer (2004): “On the Interpretation of a Concentration Index of
Inequality,” Health Economics, Vol 13, No 7, pp 649–56.

Kutzin, J (2012): “Anything Goes on the Path to Universal Health Coverage? No,” Bulletin of
the World Health Organization
, Vol 90, No 11, pp 867–68.

MoSPI (2006): “Morbidity, Health Care and the Condition of the Aged,” NSS 60th Round (2004), Ministry of Statistics and Programme Implementation, Government of India, New Delhi.

— (2015): “Social Consumption: Health,” NSS 71st Round (2014), Ministry of Statistics and Programme Implementation, Government of India, New Delhi.

MoHFW (2011): “Annual Report to the People on Health,” Ministry of Health and Family Welfare, Government of India, New Delhi.

Nandi, S, H Schneider and P Dixit (2017): “Hospital Utilization and Out-of-pocket Expenditure in Public and Private Sectors under the Universal Government Health Insurance Scheme in Chhattisgarh State, India: Lessons for Universal Health Coverage,” Plos One, 17 November,
, viewed on 29 May 2022.

O’Donnell O, E van Doorslaer and A Wagstaff (2008): “Analyzing Health Equity using Household Survey Data: A Guide to Techniques and Their Implementation,”
, viewed on 29 May 2022, Washington DC: World Bank.

O’Donnell, O, S O’Neill, T Van Ourti and B Walsh (2016): “Conindex: Estimation of Concentration Indices,” Stata Journal, Vol 16, No 1, pp 112–38.

OECD (2008): Growing Unequal? Income Distribution and Poverty in OECD Countries, Paris:
Organisation for Economic Co-operation and Development.

— (2011): “Adjusting Household Incomes Equivalence Scales,” Organisation for Economic
Co-operation and Development,, viewed on 29 May 2022.

Pandey, A, G B Ploubidis, L Clarke and L Dandona (2017): “Horizontal Inequity in Outpatient Care Use and Untreated Morbidity: Evidence From Nationwide Surveys in India between 1995 and 2014,” Health Policy and Planning, Vol 32, No 7, pp 969–79.

Palacios, R (2011): “A New Approach to Providing Health Insurance to the Poor in India: The Early Experience of Rashtriya Swasthya Bima Yojana,” India’s Health Insurance Scheme for the Poor: Evidence from the Early Experience of the Rashtriya Swasthya Bima Yojana, R Palacios, J Das and C Sun (eds), New Delhi: Centre for Policy

Pan, B, S D Towne Jr, Y Chen and Z Yuan (2017): “The Inequity of Inpatient Services in Rural Areas and the New-Type Rural Cooperative Medical System (NRCMS) in China: Repeated Cross Sectional Analysis,” Health Policy and Planning, Vol 32, No 5, pp 634–46.

Qian, Y, Z Zhou, J Gao, Y Wang, X Yang, Y Xu and Y Li (2017): “An Economy-related Equity Analysis of Health Service Utilization by Women in Economically Underdeveloped Regions of Western China,” International Journal for Equity in Health, Vol 16, No 1, p 186.

Roberts, M J, W C Hsiao and M R Reich (2015): “Disaggregating the Universal Coverage Cube: Putting Equity in the Picture,” Health Systems & Reform, Vol 1, No 1, pp 22–27.

Saito, E, S Gilmour, D Yoneoka, G S Gautam,
M M Rahman, P K Shrestha and K Shibuya (2016): “Inequality and Inequity in Healthcare Utilization in Urban Nepal: A Cross-sectional Observational Study,” Health Policy and Planning, Vol 31, No 7, pp 817–24.

Selvaraj, S and A K Karan (2012): “Why Publicly Financed Health Insurance Schemes are Ineffective in Providing Financial Risk Protection,” Economic & Political Weekly, Vol 47, No 11, pp 61–68.

Wagstaff, A and E V Doorslaer (2004): “Overall Versus Socio-economic Health Inequality: A Measurement Framework and Two Empirical Illustrations,” Health Economics, Vol 13, No 3, pp 297–301.

WHO (2010): World Health Report, 2010: Health Systems Financing the Path to Universal Coverage, Geneva: World Health Organization.

— (2015): “World Health Statistics,” World Health Organization,











Updated On : 11th Jul, 2022
Back to Top