5.1 Trend of Public Expenditure on Health for the States of North East
Government of India has continuously increased public health expenditure over the years which indicate that increased importance is given to uplift the health care sector (Berman and Ahuja, 2008; Ahamed and Honakeri, 2012). With the launch of NRHM in 2005, it was planned to increase the health expenditure by 2–3% of GDP. Even though expenditure as a percentage of GDP has increased but is less than the targeted value, as seen from Table 1.
Table 1
Public Health Expenditure as a percentage of GDP for India
Year
|
Percentage of GDP
|
2005-06
|
1.13
|
2006-07
|
1.11
|
2007-08
|
1.10
|
2008-09
|
1.16
|
2009-10
|
1.22
|
2010-11
|
1.16
|
2011-12
|
1.18
|
2012-13
|
1.18
|
2013-14
|
1.29
|
2014-15
|
1.41
|
Source: World Health Organization Global Health Expenditure database |
It is found that compared to many other countries India spends very less in health (Rao and Choudhury, 2012). Countries like Canada, Austria, Afghanistan, Brazil etc. which spend 10–11 percent of the GDP on health care (World Heath Statistics, 2009). For the development of a region it is very important to develop the health sector. Focusing on the scenario of North east, it is seen that the health expenditure for the states of North east has increased over the years which can be considered as a positive sign of development. Looking at the growing population of the states is it a real necessity to increase the expenditure on health.
Considering health expenditure in terms of constant prices taking 2004-05 to be the base year we find an upward rising trend for all states over the years 2005-06 to 2014-15. Significant investment is observed for the state of Assam when compared with other north east states, which could be due to the fact that Assam has the highest population among all other states. Therefore keeping in mind the population of Assam, it is mandatory to make more expenditures compared to other states. Hence it is very difficult to directly interpret about the adequateness of public spending on health for states with differences in population, area and NSDP. To get a clearer picture ratio of health expenditure to NSDP is considered.
Figure 2 shows the relative position of health investments of the eight states. Fluctuations are seen in all the states. It means that Assam may be investing more on health sector in absolute terms because of its higher population and greater land area compared to other states whereas in relative terms health investment in Assam is similar to that of its neighbouring states like Meghalaya, Nagaland, Tripura etc. and much less than others like Mizoram, Manipur. Mizoram spends more compared to other states which has increased to around 6% of NSDP on health in 2009-10. Highest Compound Annual Growth Rate (CAGR) was found in Manipur (11.43%) followed by Arunachal Pradesh (7.88%). Finally after analysing considering the various factors, it can be concluded that health expenditure by the government has increased in the years for all states of north east.
5.2 Effect of health expenditure on the health indicators for North Eastern states
Health expenditure is increased with the hope that increasing expenditure will improve the health scenario of the region (Bhatia, 2013; Novignon et.al, 2012). As health expenditure is found to be increasing over the years, it becomes very important to know whether this funds are properly used. Therefore this study uses dynamic panel model to find the relationship exist between increasing health expenditure and health status of north east. Dynamic Panel models are popular in many areas of economic enquiry as it provides new insights (Piper, 2014). This model includes lag of the dependent variable as one of the independent variable on the right hand side of the regression equation which represents the possibility to determine the influence of the past. Use of such models in health sector is prevalent as it includes correlation with history of the model and health indicators are very much dependent on the past scenario.
Let us assume the Dynamic panel data model to be –
𝒀𝒊𝒕 = 𝛃𝟎 + 𝛃𝟏 𝑿𝒊𝒕 + 𝛃𝟐 𝒀𝒊𝒕−𝟏 + 𝑼𝒊𝒕
For i = 1...n, representing the cross section i.e. the states of north east (n = 8). Again, t = 1… T represents time period which is annual data of 10 years in the model. 𝒀𝒊𝒕 is the dependent variable; 𝛃𝟎 is the unobserved individual and time specific effect; 𝑿𝒊𝒕 is the independent variable; 𝒀𝒊𝒕−𝟏 is the explanatory variable which is the lagged value of the dependent variable; 𝑼𝒊𝒕 is the error term.
For the present analysis, three models have been formulated-
Model 1: 𝑰𝑴𝑹𝒊𝒕 = f (𝐇𝐞𝐚𝐥𝐭𝐡 𝐄𝐱𝐩𝐞𝐧𝐝𝐢𝐭𝐮𝐫𝐞𝒊𝒕, 𝑰𝑴𝑹𝒊𝒕−𝟏)
Model 2: 𝐁𝐢𝐫𝐭𝐡 𝐑𝐚𝐭𝐞𝒊𝒕 = f (𝐇𝐞𝐚𝐥𝐭𝐡 𝐄𝐱𝐩𝐞𝐧𝐝𝐢𝐭𝐮𝐫𝐞𝒊𝒕, 𝐁𝐢𝐫𝐭𝐡 𝐑𝐚𝐭𝐞𝒊𝒕−𝟏)
Model 3: 𝐃𝐞𝐚𝐭𝐡 𝐑𝐚𝐭𝐞𝒊𝒕 = f (𝐇𝐞𝐚𝐥𝐭𝐡 𝐄𝐱𝐩𝐞𝐧𝐝𝐢𝐭𝐮𝐫𝐞𝒊𝒕, 𝐃𝐞𝐚𝐭𝐡 𝐑𝐚𝐭𝐞𝒊𝒕−𝟏)
Arellano- Bond Serial Correlation test has been performed to check the correlation between the model errors. The test result showed that there is no autocorrelation. For dynamic panel model the variables should be non-stationary at level but after first differencing it should be stationary. To check the stationary of IMR, Birth Rate, Death Rate and Health expenditure, Panel unit root test was performed. It is seen that the variable were non stationary at level but transforming them to 1st order difference, they became stationary.
Table 2
Results of Dynamic Panel Analysis
|
Model 1
|
Model 2
|
Model 3
|
Variable
|
Coefficient
(Probability)
|
Coefficient
(Probability)
|
Coefficient
(Probability)
|
Constant
|
0.002772
(0.1766)
|
0.000363
(0.2498)
|
0.000418**
(0.0291)
|
Health Expenditure
|
-3.65E-08**
(0.0295)
|
-1.74E-09
(0.6971)
|
-3.25E-08**
(0.0434)
|
Lag Variable
|
0.947112***
(0.0000)
|
0.972157***
(0.0000)
|
0.943965***
(0.0000)
|
R2
|
0.969870
|
0.998082
|
0.988124
|
Adjusted R2
|
0.968997
|
0.998026
|
0.987780
|
From all the 3 models of our analysis it is found that there exist an inverse relation between health expenditure and health status indicators considered like IMR, Birth rate and Death rate. It implies that when expenditure increases all the three indicators fall which is considered to be good for the economy. Health status increases with the increase in the expenditure on health care. Also it is found that significant relationship exist in case of the health indicators like IMR and Death rate and also the lagged value of the health indicators which imply that the indicators are dependent on the past history. But insignificant relationship exist in case of Birth rate this could be because Birth rate is dependent on many other factors such as education and employment opportunities of women, social customs, religious beliefs, average age at marriage and use of contraceptives etc. rather than only depending on expenditures made on healthcare. Further, the R2 values of all the estimated models are reasonably high. High R2 value implies models are statistically significant with reasonably high explanatory power.