Purpose of Study: Regional disparities and inequality continue to be a feature of Indian economy even after seven decades of independence. Many of its social indicators need much improvement. Some states are particularly more backward with large proportions of their population being officially poor while some others are comparatively in better position. Such inter-regional disparities have compounded policy challenges of the governments in the poorer states. Against this background, the present study aims to study the dimension of inter-regional disparity for select less advanced states in India.

Methodology: A double log model was used in this study to analyze government expenditure's impact on development projects or schemes. Health-wise, less advanced states, viz. Bihar and Odisha are chosen for this purpose. The study uses the actual data on government expenditure in the social sector, mainly on health. The data on a per capita basis is used for each state to analyze the impact of the per capita government's expenditure on select social indicators. The analysis is done separately for both states.

In a developing country like India where significant part of population are poorer and living under miserable conditions and have to struggle daily for their livelihood, so it is not possible for them to access health care, education and other social services at their own. So, it becomes the duty of the government to provide effective social services at a very reasonable cost. According to (Gupta, 2002), "Health care services have high level of externalities rather than curative services, a minimum package of these services provided by the government would reduce mortality rates". Since, governments in developing countries always have scarcity of funds, so it is necessary to ensure that the funds are used effectively and the desired results are attained at social front. So, it is also important to check the effectiveness of government expenditure on the improvement of social indicators. Further, government's spending is also important to uplift the living standards of the poorer people in the society. As Gera, in her studies also found that government investments in education, health and in the provision of infrastructure can have direct effect on moving household out of poverty (Gera, 2007). Further, Ranjan and Sharma (2008) examined the effect of government development expenditure on economic growth and they discovered a significant positive of government expenditure on economic growth. A study found, educational attainment at basic levels (secondary level) and low infant mortality rates have been shown to have a positive effect on economic growth also (Barro and Lee, 1993). Studies on both developed and developing countries have indicated that sufficient amount of government spending on education and health improves human development and lessens poverty burden as well (Barro and Lee, 1997;Swaroop, 1996). However, it is also necessary to mention that the solely the increase in public spending is not sufficient but the quality of expenditure with good public policies also required. As stated, a government could increase the public spending by a large amount but this does not ensure that it would have desired result on economic and social development as the quality of this spending also matters (Bussato and Brunori, 2011).

Despite the importance of government spending and its role on improvement of social sector, there are not sufficient number of studies have been done in India to evaluate the impact of government spending on social indicators. Thus, present study is an attempt to evaluate the impact of government spending on some selected social indicators and further it will also make a significant contribution to the present literature. As the number of social indicators are very large, so it is not feasible to assess every indicator given the time and data constraint. Hence, the study has selected four indicators i.e., Infant Mortality Rate (IMR), death rate, birth rate and total fertility rate as indicators of health. The study has chosen Bihar and Odisha states.

The following social indicators have been selected for the present study. Anderson al.et (2000), revealed that the USA spent more on health care as compare to other countries. USA spent 14% of GDP on health care in 1998 while OECD median was8% of GDP and results also suggested that Americans enjoys better health care system than other OECD countries. Shenggen Fan et al. (2002) 2012) revealed that government expenditure on health has a significant positive effect on health status while, expenditure on education has no significant impact on either primary or secondary school enrolment. Maitra, B., and C.K. Mukhopadhyay (2012) shown that impact of education and health spending on growth is not an instantaneous but with gestation lags. Initially, expenditure on education and health improves human capital which manifests itself in the form of economic growth. Further, it is found that the gestation lag of education spending was longer than that of health-care spending. Sava? Çevik, M. & Okan Ta?ar (2013) found that government health spending has significant impact on under-5 child mortality rate and on infant mortality rate. Study also concludes that composition of government health expenditure also matters not only the size of expenditure. Tae Kuen Kim and Shannon R. Lane (2013) shown a negative relationship between public the health expenditure and the infant mortality rate (IMR), while positive association between public health expenditure and life expectancy is found. Thus, the study concludes that expanding public health expenditure improves overall health condition. Bhakta, R. (2014) shown that public expenditure on Supplementary Nutritional Program has positive impact on health status of children which also has indirect positive impact on education. Study also concludes that public expenditure on elementary education has direct impact on the enrolment rate. Virupakshapp a D Mulagund (2015) suggested that public health expenditure in India have increasing trend during this period. Further, study concludes that public health expenditure has positive impact on health indicators i.e, it resulted in fall in maternal mortality rate (MMR), infant mortality rate (IMR), fall in total fertility rate (TFR) and improves life expectancy. Wong Sing Yun and Remali Yusoff (2015) indicated there is a unidirectional causal relationship from GDP to education expenditure and from GDP to health care expenditure. Thus, study concludes that GDP affect both the education and health care expenditure. However, reverse causal relationship is not found between them. K. P. K. S.

1. To evaluate the impact of government expenditure on selected social indicators in less advanced Indian states. 2. To suggest policy implications for better utilization of public expenditure on social sectors.

For the purpose of determining the impact of government's expenditure on social indicators, the study has applied log-log or double-log model. In case of Log-log models, the coefficients are used to determine the relative impact of independent variable(s) on relative impact of dependent variable. Here, the independent variable is government expenditure and the social indicator(s) chosen are the dependent variables. The coefficients in a log-log model represent the elasticity of dependent variable with respect to independent variable. Therefore, log-log model presents the empirical interpretation in elasticity term i.e., percentage change in dependent variable due to one percent change in explanatory variable.

Log-log model is represented as:

In Yi= In ?1 + ?2 In X i + ui

(1) Where In= Natural log (i.e., log to the base e, and where e = 2.718)

Equation ( 1) is thus:

The coefficients are estimated by OLS regression. Six equations will be fitted/estimated for each selected state.

The if the value of explanatory variable is increased by 1 percent, then the value of dependent variable decreases by 0.10 per cent. From the analysis table we can see the R-squared value is 0.9556 which tells 95.56 percent of variation in dependent variable birth rate is explained by independent variable. The p-value is 0.0000 being less than the significant level of 5% percent which shows that the explanatory variable is statistically significant and, therefore, the null hypothesis that the coefficient of explanatory variable is zero will be rejected. It means we can say that the per capita public health expenditure on health has impact on birth rate. Table 3 provides the results regarding the impact of government's expenditure on health on infant mortality rate (IMR). Here, infant mortality rate is a dependent variable. The squared-R is 0.97 which tells that around 97 percent of the variation in dependent variable is explained by the independent variable. As we can see that the p-value is 0.0000 being less than the significant level of 5% percent which shows that the explanatory variable is statistically significant and, therefore, the null hypothesis that the coefficient of explanatory variable is zero will be rejected. Apart from this, the negative symbol with explanatory variable shows that there is negative relationship between the dependent variable and explanatory one. The explanatory coefficient value is -0.336 which indicates that 1 percent increase in per capita may lead to 0.336 percent fall in IMR. 4.8d provides the results of analysis between per capita health expenditure and total fertility rate (TFR). Here, the total fertility rate is dependent variable while the per capita expenditure on health is independent variable. From the table we can see that the coefficient has a negative sign with value of -0.14 which tells there is an inverse relationship between health expenditure and the TFR i.e., an increase in per capita health expenditure results in 0.14 percent fall in TFR. The R-squared value is 0.911 which tells 91.1 percent of variation in dependent variable TFR is explained by independent variable per capita expenditure on health. The p-value is 0.0000 which is appearing against the explanatory variable is statistically significant because the p-value being less than the significance level of 5 percent (0.05), hence the null hypothesis of that, the explanatory variable is statistically insignificant and being rejected. Table 6 gives the results relating to the impact of government's expenditure on health on death rate in Bihar. Here the death rate is dependent variable. From the table we can see that the per capita health expenditure coefficient has a negative sign which tells there is an inverse relationship between health expenditure and the death rate i.e., an increase in government expenditure on health causes fall in death rate. The coefficient has -0.081 value which means 1 percent increase in per capita health expenditure causes 0.081 percent fall in death rate. The R-squared value is 0.6031 which tells 60.31 percent of variation in dependent variable death rate is explained by independent variable. Further, we can see that the p-value is 0.0007 which is appearing against the explanatory variable is statistically significant because the p-value is being less than the significance level of 5 percent (0.05), hence the null hypothesis of that the explanatory variable is statistically insignificant and being rejected here also. The results of this study are consistent across all variables considered for the study. Our principal conclusion can be summarized as per capita government expenditure on health helps to reduce infant mortality rate, birth rate, death rate and total fertility rate in Bihar and Odisha states. These results indicate that the government should increase its budgetary allocations on health and family welfare as well. These results are also important in considering the fact that there should be the commitment of more funds health. Although only commitment of funds to social sector is not sufficient, better utilization of funds right direction in effective manner is most important. Thus, it is also essential for the government to look after the efficiency and transparency of its budgetary allocations to ensure that these funds are fully utilized (Yun and Yusoff, 2015). Thus, analysis of this study can pave way in determining the optimal mix of It indicates that increase in government spending results in fall in IMR, Birth Rate, Death Rate and TFR. Therefore, the government should further increase its expenditure in health and family welfare. However, merely increasing the allocation of funds to the social sector is not sufficient, effective utilization of funds also necessary. Thus, it is also essential for the government to look after the efficiency and transparency of its budgetary allocations to ensure that these funds are fully utilized. Therefore, policy-makers should address other important factors also apart from allocating public expenditure like the effectiveness of the government schemes in health and family welfare, and proper implementation of such schemes.

From various studies, it can be intuitively explained by the fact that because of extreme poverty and deprivation in India the welfare of the society can be increased by greater involvement of government. At the policy level, the present study recommends that public expenditure should increase further to have a balanced and improved human development of the concerned states. So, an increase in social sector expenditure should also be considered as one of the priorities to promote efficiency in growth and development. Hence, sufficient amount of government funds is recommended to provide support to policies and programs necessary to achieve welfare, growth and development of these states in particular, and the country in general. Therefore, the study is an attempt to analyze the relationship between the public spending on health sector and the selected health indicators in Bihar and Odisha. The study has used the state -level data for the selected states to estimate the direct and indirect effects of government's expenditure on social indicators. The findings clearly indicate that government expenditure does have impact on selected social indicators. The results of the study shows that per capita expenditure on health is inversely related with all the four selected health indicators i.e., increase in per capita expenditure leads to fall in Birth Rate, Death Rate, Infant Mortality Rate (IMR) and Total Fertility Rate (TFR) in both states, however, the amount of decrease will depend on their respective coefficient values.