This thematic Chapter provides an analysis of the expenditure of State governments covering the period 1980-81 to 2009-10. It also attempts to examine whether major categories of expenditure have shown any convergence in per capita terms across the States. The composition of aggregate expenditure shows that revenue expenditure increased steadily during the 1980s and 1990s. However, there has been a modest decline in its share since the beginning of the 2000s mainly on account of a decline in development revenue expenditure. The ever increasing trend in the committed expenditure-GDP ratio declined during 2005-10 mostly owing to a decline in the interest payment-GDP ratio. Reflecting the emphasis of States on quality of expenditure, the capital outlay-GDP ratio has been rising since 2000-01. Deviating from the declining trend till 2004-05, the average share of development expenditure in aggregate expenditure has risen significantly during 2005-10. A State-wise analysis shows that the trend in major categories of expenditure varies across States. An analysis further shows that in per capita terms, the levels of capital outlay, development expenditure and social sector expenditure have shown some convergence across the States.

1. Introduction

6.1 Under the federal system in India, the State governments have a major responsibility with regard to the provision of economic and social infrastructure¹². The State governments accounts for around 60 per cent of the combined expenditure of the Centre and States reflecting the vital role that the States play in the growth and development of the economy. In the following discussion, an attempt is made to examine trends in various categories of expenditure by State governments. An inter-temporal analysis of the expenditure pattern of the States not only signifies the changing policy emphasis of State governments but also helps to assess whether any convergence is taking place across the States in terms of allocation of expenditure. As a special theme for the present study, an analysis of the trend and pattern of States’ expenditure is presented in this Chapter. The theme is aimed at a focused analysis of the expenditure of State governments covering the period 1980-81 to 2009-10. The analysis is structured into the following sections. Section I deals with the overall trend in aggregate expenditure of State governments followed by Section II and Section III discussing the trend and composition of revenue expenditure and capital expenditure, respectively. Expenditure pattern in terms of development and non-development expenditure of State governments is discussed in Section IV. Section V brings out trend in major categories of social sector spending of the State governments. State-wise trend in major categories of expenditure is discussed in Section VI. In Section VII, an attempt is made to examine whether major categories of expenditure have shown any convergence in per capita terms across States.

6.2 Public expenditure plays an important role in achieving goals of growth, development, equity and stability. In the context of developing economies like India, public expenditure assumes importance in order to ensure an equitable distribution of resources. The redistributive powers of the State emanate from the normative arguments in favor of greater equality to be achieved through public expenditure. The level and composition of public expenditure can have macroeconomic as well social implications. Policy makers and some researchers have argued that expenditure on growthenhancing functions could enhance future revenue and justify the provision of ‘fiscal space’ in the Budget. However, there are no simple ways to identify the growth-maximising composition of public expenditure (Semmler et al, 2007)¹³. Given the empirical evidence that a link between public expenditure and growth is contingent upon the nature of expenditure, an assessment of the trends and composition of public expenditure assumes importance.

2. Overall Trend

6.3 Trend analysis shows that the aggregate expenditure of State governments as percentage of GDP accelerated during the 1980s and decelerated during the 1990s. Aggregate expenditure as percentage of GDP moved upward during 2000-05. However, compression in the consolidated expenditure of State governments can be observed during 2005-10 mainly on account of some rationalisation of revenue expenditure during the fiscal responsibility legislation (FRL) period. This is evident from a decline in the RE-GDP ratio from 13.3 per cent in 2000-05 to 12.4 per cent during 2005-10 (Table VI.1 and Chart VI.1).

6.4 In terms of the expenditure policy, not only the level of expenditure, but the structure also matters. The former reflects the amount of distortions, while the latter has important consequences for the effectiveness of the expenditure policy (Kastele, 2005). Broadly, all expenditure of the government, which does not result in the creation of physical or financial assets, is treated as revenue expenditure. As far as a broad composition of the total expenditure of State governments is concerned, revenue spending showed a steady increase during the 1980s and 1990s. Since the beginning of the 2000s, there has been a modest decline in the share of revenue expenditure to total expenditure. With the concomitant rise in the share of capital expenditure to total expenditure from 16.8 per cent during 1995-00 to 21.2 per cent during 2000-05, there was an increase in the capital outlay by 0.2 percentage points during the same period (Table VI.2 and Chart VI.2). However, the share of capital outlay in aggregate expenditure rose sharply from 9.6 per cent during 2000-05 to 15.4 per cent during 2005-10. Nevertheless, the spending patterns of State governments show the persisting domination of revenue expenditure with marginal significance for their long term growth potential as they are generally considered to be consumption spending of the State governments unlike capital expenditure. Moreover, the rising share of revenue expenditure reflects structural rigidities in expenditure patterns making expenditure management of State governments difficult. Furthermore, the compound annual rate of growth (CARG) in revenue expenditure during 1980-81 to 2009-10 is found to be higher than capital expenditure.

3. Revenue Expenditure

6.5 The composition of aggregate expenditure by State governments in terms of revenue and capital expenditure is reflective of the quality of expenditure incurred. The hypothesis that improvements in the composition of public expenditure have positive repercussions for growth is widely supported in literature. It is generally found that fiscal consolidation achieved through compressing selected revenue expenditures tend to trigger higher growth rates than adjustments based on revenue increases and cuts in more productive spending (Pang et al., 2007)¹⁴. Thus, it is important to examine whether the trend and composition of revenue expenditure has undergone any change over the period. As far as the composition of revenue expenditure is concerned, it continues to be dominated by development expenditure which mainly comprises spending by States on social and economic services. Development expenditure accounted for 71 per cent of the total revenue expenditure of the States during 1980-85. However, its share in total revenue expenditure steadily declined till 2000-05 (54.7 per cent) before rising marginally in subsequent years (58.0 per cent during 2005-10). The share of non-development revenue expenditure in total revenue expenditure witnessed a concomitant increase till 2004-05 and a moderate decline thereafter (Chart VI.3). Development revenue expenditure continues to be dominated by social services. Social services—accounting for 57.5 per cent of the total development revenue expenditure during 1980-85—have witnessed a marginal increase in their share since 1995-2000. In contrast, the average share of economic services recorded a marginal increase during 1985-90 and 1990-95 but declined in subsequent sub-periods (Chart VI.4).

6.6 Development revenue expenditure as percentage to GDP (DRE-GDP) which stood at 7.5 per cent during 1980-85, rose to 8.5 per cent during 1985-90 due to a rise in revenue spending on social as well as economic services. However, the DREGDP ratio has witnessed a secular decline since 1990- 95 mainly due to declining revenue expenditure on economic services as percentage to GDP. Revenue expenditure on social services as percentage to GDP has also declined since 1985-90 albeit at a slower pace compared to economic services. The major categories of revenue expenditure on social services, viz., education, sports, art and culture; medical and public health; and water supply and sanitation witnessed a decline in terms of GDP. Among the economic services, agriculture and allied activities accounted for a major decline in the DRE-GDP ratio over the years. However, the States have gradually increased their revenue spending on the energy sector from 0.1 per cent in 1980-85 to 0.7 per cent of GDP during 2000-05 (Table VI.3). Non-development revenue expenditure as percentage to GDP (NDREGDP) has shown an upward trend over the period of analysis. The NDRE-GDP ratio doubled from 2.9 per cent during 1980-85 to 5.8 per cent during 2000-05 before declining marginally to 4.9 per cent during 2005-10. Revenue expenditure in the form of grantsin- aid and contributions as compensation and assignments to local bodies and Panchayati Raj Institutions (PRIs) witnessed an increase from 0.1 per cent of GDP during 1980-95 to 0.2 per cent during 1995-2005 and further to 0.3 per cent during 2005-10. Since the total own revenue of the local bodies, which were assigned wide expenditure obligations after the 73rd and 74th Constitutional amendments, did not show any perceptible improvement, State governments have been increasingly financing local bodies.

6.7 Interest payments, administrative services and pensions account for a dominant portion of the non-development revenue expenditure. Such expenditure is of a committed nature and has a first charge on the government’s resources. Thus, such expenditure renders the expenditure management process less flexible for the State governments. Committed expenditure as percentage to GDP rose substantially from 2.3 per cent during 1980-85 to 5.0 per cent during 2000-05. This was mainly on account of a sharp increase in interest payments by State governments to service their outstanding debts comprising mainly of loans from the Centre, internal debt, small savings and provident funds. Between 1980-85 and 2000-05, around 66 per cent of the total increase in committed expenditure could be attributed to a rise in interest payments. The debt servicing burden soared with high cost borrowings financing current expenditure amidst growing fiscal imbalances, particularly during 1986- 87 to 1997-98. In fact, the burgeoning fiscal gap fed on itself as the ratio of interest payments to revenue receipts of the States also deteriorated sharply during the 1990s. While expenditure on administrative services as percentage of GDP remained broadly stable during the period of analysis, expenditure on pensions by State governments as percentage of GDP increased significantly (Table VI.4 and Charts VI.5 and VI.6). During 2005-10, committed expenditure moderated owing to a decline in the interest payment-GDP ratio enabling a reduction in the non-development component of revenue expenditure as well as the total revenue expenditure-GDP ratio.

6.8 Before the initiation of the National Small Saving Fund, loans from the Centre were a major source of financing of fiscal deficits of the States till 1998-99. Thus, interest payment on these loans remained a major component in the total interest payments of the States till 2003-04. Thereafter, there has been a significant decline in interest payment on loans from the Centre partly due to the Debt Swap Scheme (DSS) operated during 2002-05 and the Debt Consolidation and Relief Facility (DCRF) recommended by the TwFC (Table VI.5). A similar trend has been observed in interest payment as percentage to revenue receipts (IP-RR). The IP-RR ratio moved progressively from 7.5 per cent during 1980-81 to 26.0 per cent during 2003-04. Subsequently, there has been a considerable decline in the IP-RR ratio to 15.1 per cent during 2008-09 (Chart VI.7). This broadly complies with the sustainability level of below 15.0 per cent prescribed with respect to the IP-RR ratio of the States by the TwFC. Interest payments as percentage of GDP also showed a secular increasing trend till 2003-04 and a declining trend thereafter (Chart VI.8).

6.9 In the light of the Centre’s decision to discontinue Plan loans to the States with effect from April 2005 as recommended by the TwFC, the States had to mobilise resources for funding their GFD mainly through market borrowings and special securities issued to NSSF. Consequently, interest payments on market loans and NSSF loans have gradually risen in the recent period. Since the interest rate for NSSF loans is the highest of all the borrowings of the States, it puts enormous strain on interest payments. However, interest payment on small savings and provident funds as percentage to GDP has remained almost stable in recent years.

4. Capital Expenditure

6.10 Capital expenditure assumes importance as it has a lasting impact on growth than revenue expenditure. If spent efficiently, it also ensures a more productive economy and enhances the government’s net worth arising from augmented revenues due to higher expected GDP in the future. Likewise, an efficient allocation of the government’s capital expenditure can impact private investment provided it does not raise cost of borrowings for private investors. Thus, capital outlay for infrastructure is expected to enhance productivity levels of private investment leading to crowding in of private investment.

6.11 During 1980-2010, the share of capital expenditure in total expenditure of States exhibited a marginal increase and that too mainly in recent years. Capital outlay¹⁵ as a percentage to GDP (CO-GDP), an indicator of investment activities of State governments, began to show a declining trend in the early 1980s which has become pronounced particularly since 1987-88. During this period, the States slashed their investment activities faced with revenue deficit from 1987-88 onwards. There was a slowing down of investments by State governments reflecting a further shifting of State expenditure towards revenue expenditure largely due to continued growth in non-Plan expenditure on account of interest payments, pensions and administrative services. However, this trend has reversed in recent years and the evolving pattern of expenditure, particularly during the post-2002-03 period, indicates a sharp increase in capital-outlay as percentage to GDP. An inter-temporal comparison shows that the CO-GDP ratio declined from 2.0 per cent during 1980-85 to 1.8 per cent during 1985-90. The decline became more entrenched during the 1990s with the deteriorating fiscal situation of the States. However, with the overall improvement in State finances, the States have been able to scale up capital outlay as reflected in the CO-GDP ratio of 2.4 per cent during 2005-10 (Chart VI.9).

6.12 Another encouraging trend that has emerged in recent years is the rising share of capital outlay in total capital expenditure of State governments. The share of capital outlay in total capital expenditure increased from 44.4 per cent in 1980-81 to 68.6 per cent during 2009-10. This reflects an increasing role of State governments in generating productive capacity and enhancing their growth potential. Although growth in capital outlay has shown fluctuating trends over the years, CARG during 1980-81 to 2009-10 was found to be higher than that of capital expenditure of State governments (Chart VI.10).

6.13 Capital outlay mainly comprises spending on developmental activities pertaining to social and economic services. Developmental capital outlay as percentage to GDP (DCO-GDP) persistently declined from 2.0 per cent during 1980-85 to 1.3 per cent during 1995-2000. However, with increasing focus of State governments on economic services pertaining to rural development, irrigation activities, energy and transport in subsequent years, the DCO-GDP ratio rose to 1.6 per cent during 2000-05 and 2.4 per cent during 2005-10. Developmental capital outlay on economic services as percentage to GDP rose from 1.1 per cent during 1995-2000 to 1.9 per cent during 2005-10. The development capital outlay of State governments on the transport sector as percentage to GDP has witnessed a considerable increase, particularly since the beginning of the 2000s. Similarly, an increase in development capital outlay on the energy sector as percentage to GDP from 0.15 per cent to 0.30 per cent during 2009-10 reflects the State governments’ focus on meeting their energy requirements. Furthermore, developmental capital outlay on social services as percentage to GDP also increased from 0.3 per cent to 0.5 per cent during 1995-2000 and 2005-10. Within the social services, capital outlay was mainly allocated in the sectors, viz., water supply and sanitation followed by education, sports, art and culture and medical and public health (Appendix Table 22).

6.14 Capital expenditure towards repayment of loans to the Centre has surged, particularly since 2002-03. Under the Debt Swap Scheme (2002-03 to 2004-05), the States had to pre-pay their high-cost debt to the Centre through additional market borrowings and proceeds from small savings. In contrast, the capital expenditure of State governments through loans and advances for development and non-development purposes has shown a persistent downtrend during the period of analysis. In particular, loans and advances for developmental purposes dwindled from 1.39 per cent of GDP during 1980-85 to 0.34 per cent during 2005-10.

Development and Non-development Expenditure¹⁶

6.15 Another way to analyse the State governments’ expenditures is to classify them under: ‘development’ and ‘non-development’ expenditures. Development expenditure has a beneficial impact and leads to economic and social development. Non-development expenditure, on the other hand, captures administrative expenditure and interest expenditure. Thus, such a classification of expenditure is important to capture the qualitative changes taking places in the expenditure patterns of State governments. Since development expenditure relates to both economic and social services, its trends broadly capture the evolving role of State governments in the economic and social development of the economy.

6.16 A composition of aggregate expenditure by State governments shows that it has been largely spent for developmental purposes (Chart VI.11). Development expenditure as percentage of GDP (DEGDP), in general, showed a declining trend during 1987-88 and 2004-05. However, the DE-GDP ratio rose thereafter. While development capital outlay as a percentage of GDP has shown a significant rise during 2000-05 and 2005-10, development revenue expenditure as percentage of GDP continued to show a declining trend (Table VI.6 and Chart VI.12).

6.17 As can be seen from Chart VI.13, the share of economic services in aggregate development expenditure of the States was higher than that of social services during the 1980s and 1990s. During the subsequent period, however, the average share of economic services in the aggregate development expenditure of the States was lower than that of social services.

Social Sector Expenditure

6.18 Social sector expenditure is called for as most of the sectors falling under this have large externalities or spillover benefits in areas such as education, public healthcare and water and sanitation. In the Indian context, the State governments are entrusted with higher responsibilities by the Constitution (Seventh Schedule, Article 246) with respect to social spending such as on health, education and family welfare. As far as allocation to social sector expenditure by State governments as percentage to GDP (SSE-GDP) is concerned, it shot up sharply during the second half of the 1980s and showed some moderation during the 1990s. It has often been pointed out that social sector expenditure was the most vulnerable to reduction in the total budget even before India’s reforms began (Tsujita, 2005)¹⁷. A number of studies found that social sector expenditure declined under pressure to reduce fiscal deficits following economic reforms during the 1990s. Although, there was some pick up in the SSE-GDP ratio during the terminal years of the 1990s, on an average basis, it remained at 5.5 per cent which was lower than that in 1990-95. During the first half of the 2000s, the SSE-GDP ratio was broadly stable at 5.5 per cent. However, social sector expenditure as percentage to GDP accelerated during 2005-10 (Chart VI.14).

6.19 A major chunk of the social sector expenditure has been in the form of revenue expenditure rather than capital outlay. Nevertheless, the share of capital outlay in total social sector expenditure showed some improvement during 2000-05 and 2005-10. At present, capital outlay forms approximately 10 per cent of the total social sector expenditure. This indicates that only 10 per cent of the total social sector expenditure (i.e., capital expenditure comprising of capital outlay and loans and advances) is allocated towards investment in social infrastructure. The fiscal priority given to the social sector (i.e., SSE to total expenditure ratio) which averaged at around 31.6 per cent during 1980-85 increased to 38.0 per cent during 1985-90. However, it moderated somewhat during the 1990s and declined significantly during the first half of the 2000s albeit with an upward trend since 2005-06.

6.20 During the 1980s, key social sectors like rural development, education, sports, art and culture, water supply and sanitation and welfare of Scheduled Castes, Scheduled Tribes and Backward Classes seemed to have received particular attention. Although there was some moderation in the SSE-GDP ratio during the 1990s as the deteriorating fiscal position of the States led to resource gaps during and spending on some social sectors were adversely affected. Despite this, key social sectors, viz., education, sports, art and culture, medical and public health and rural development continued to receive a relatively greater allocation of expenditure (Appendix Table 23).

5. State-wise Analysis of Expenditure

6.21 State government expenditure as percentage to GSDP (SGE-GSDP ratio) is found to be significantly higher in the special category States, viz., Sikkim, Arunachal Pradesh, Mizoram, Manipur and Jammu and Kashmir. The role of State government expenditure in the economic activities of these States appears to be significant. One of the reasons for the higher SGE-GSDP ratio in special category States could be higher resource transfers by the Central Government as compared with other States. However, the SGE-GSDP ratio is found to be lower than 20 per cent in States like Maharashtra, Gujarat, Haryana, Kerala, Tamil Nadu and West Bengal (Statement 49).

6.22 In case of States with per capita income below the average, viz., (Sikkim, Jammu and Kashmir and Bihar), the SGE-GSDP ratio rose sharply during 1980-85 and 2005-10. In contrast, during the same period, the SGE-GSDP ratio showed a decline in States, viz., Tamil Nadu, Meghalaya, Gujarat, Maharashtra, Kerala, Manipur, Haryana and Nagaland (Chart VI.15). This indicates that the size of the economy of these States (i.e., GSDP) grew faster than the size of spending by the respective State governments. However, an analysis shows that the average change in the SGE-GSDP ratio during 2005-10 over 1985-90 was not statistically significant at 5 per cent¹⁸. Thus, in general, the size of State governments has not changed significantly over the period of analysis.

6.23 Most of the special category States spent more in the form of per capita expenditure than the national average during 2005-10. In contrast, the high per capita income States, viz., Punjab, Haryana, Gujarat, Kerala and Tamil Nadu spent less than the national average during the same period. The rank correlation between the State-wise SGE-GSDP ratio and per capita aggregate expenditure for 2005-10 turns out to be 0.49 which is statistically significant at 1 per cent. This shows that, in general, the States with high a SGE-GSDP ratio also lead in terms of per capita aggregate spending.

Revenue Expenditure

6.24 State-wise analysis shows that revenue expenditure as a percentage of GSDP was generally higher in special category States than that in non-special category States. For instance, the RE-GSDP ratio was the highest in Sikkim at 92.8 per cent during 2005-10, followed by Arunachal Pradesh (64.5 per cent) and Mizoram (58.0 per cent). In contrast, the Revenue Expenditure-GSDP ratio was lower than 20 per cent in all the nonspecial category States (except Bihar) during the same period. Among the non-special category States, Gujarat, Maharashtra and Haryana had the lowest RE-GSDP ratio during 2005-10. An intertemporal comparison shows that the two States, viz., Maharashtra and Nagaland witnessed a decline in the RE-GSDP ratio during 2005-10 as compared with that in 1980-85. However, 15 out of the 28 States recorded a decline in the RE-GSDP ratio during 2005-10 over 2000-05. This shows that there has been some rationalisation in revenue expenditure across States during the FRL period (Statement 50). In the case of non-special category States, the RE-GSDP ratio, albeit lower during 2005-10 than that in 2000-05, was still statistically significantly higher than that in 1980-85.

6.25 Within revenue expenditure, interest payment for servicing of outstanding debt is one of the major committed expenditures of State governments. Interest payment as percentage to GSDP declined during 2005-10 over 2001-05 in all the States, except Madhya Pradesh, Himachal Pradesh Mizoram and Uttarakhand. The decline during 2005-10 over 2001-05 was statistically significant at 1 per cent level of significance. This could be attributed to the Debt Swap Scheme (DSS) operated during 2002-05 and the Debt Consolidation and Relief Facility (DCRF) recommended by the TwFC which facilitated the States in reducing expenditure on interest payments.

Capital Outlay

6.26 Most of the State governments seem to have realised the need to contain unproductive expenditures and reorient spending towards developmental purposes. Capital outlay as a percentage of GSDP (CO-GSDP) recorded an increase across all the States during 2005-10 over 2000-05. Among the non-special category States, Bihar and Uttar Pradesh recorded a significant rise in CO-GSDP ratio while among the special category States, the rise was sharpest in Arunachal Pradesh, Manipur, Mizoram and Uttarakhand. The overall rise in CO-GSDP ratio across States during 2005- 10 over 2000-05 was statistically significant at 1 per cent (Statement 51). One interesting fact that emerges from this analysis is that among the nonspecial category States, the CO-GSDP ratio was considerably higher in Orissa and in some of the underdeveloped States, viz., Bihar, Rajasthan, Madhya Pradesh and Uttar Pradesh during the 1980s. Furthermore, these States witnessed a significant increase in the CO-GSDP ratio during the 1990s moving in tandem with most of the other States. Even during 2005-10, Bihar, Madhya Pradesh, Jharkhand, Rajasthan and Uttar Pradesh had the highest CO-GSDP ratio among the nonspecial category States. However, these States have lower per capita capital outlay as compared with other States. This may have implications for their level of per capita income as well.

6.27 As far as the trend in per capita capital outlay across States is concerned, a few States seemed to have lagged behind others in terms of growth. During 1980-2010, the compound annual growth rate in per capita capital outlay was below 10 per cent in Kerala, Madhya Pradesh, Orissa, Arunachal Pradesh and Mizoram. In contrast the States, viz., Tamil Nadu, Maharashtra, Andhra Pradesh, Karnataka and Assam recorded the highest CARG during the same period reflecting their increasing focus on building infrastructure for boosting growth and development. Per capita capital outlay is generally higher in special category States than in the non-special category States (Statement 52). The rank correlation at 0.71 between the CO-GSDP ratio and per capita capital outlay is statistically significant at 1 per cent. Thus, it can be inferred that in general States having a higher CO-GSDP ratio also spent higher amount in terms of per capita during 2005-10.

Development Expenditure

6.28 During 2005-10, development expenditure as a percentage to GSDP (DE-GSDP) has shown a significant increase over 2000-05 across a majority of the States. In a few States, however, the DE-GSDP ratio was still lower than that during 1980-85 (Statement 53). The DE-GSDP ratio is found to be higher among the special category States as compared with non-special category States. Among the non-special category States Bihar, Madhya Pradesh, Uttar Pradesh and Orissa —all with per capita income below the national average—seemed to have focused more on development spending as their respective DEGSDP ratios rose during 2005-10 over 2000-05. In contrast, some rich States like Punjab, Haryana, Gujarat, Goa, Maharashtra and Tamil Nadu witnessed a decline in the DE-GSDP ratios during 2005-10 over that during the 1980s and 1990s. With regard to the share of development expenditure in the total expenditure of States during 2005-10, there was a statistically significant decrease as compared with 1980-85. However, the share of development expenditure in total expenditure by State governments turned out to be significantly higher than that in 2000-05 (Statement 54).

6.29 Per capita development expenditure shows a distinct trend between special and non-special category States. Special category States tend to spend more on development expenditure per capita. No distinct trend, however, is discernible in terms of CARG in per capita development spending across special and non-special category States. During the period of analysis, CARG in per capita development spending by the State governments was found to be higher than 10 per cent in all the States, except Madhya Pradesh, Mizoram and Nagaland (Statement 55). The statistically significant rank correlation of 0.50 between the DEGSDP ratio and per capita development expenditure across States shows that, in general, States with a high DE-GSDP ratio also allocate more in terms of per capita development expenditure.

Social Sector Expenditure

6.30 Social sector expenditure as percentage to GSDP was found to be generally higher in special category States as compared with non-special category States. A State-wise trend in social sector expenditure as percentage to GSDP shows that all State governments increased their spending on social sector development during the second half of the 1980s. However, there was a decline in the SSE-GSDP ratio across a majority of the States during 1990-2005 (Statement 56). In recent years, there seems to be more emphasis on social sector spending by the State governments as an average rise in the SSE-GSDP ratio across the States was statistically higher than that during 2000-05. In most of the States, the per capita social sector spending of State governments recorded a CARG of above 10 per cent (Statement 57). As in case of development expenditure, rank correlation between SSE-GSDP ratio and per capita social sector spending is also found to be statistically significant.

6.31 Once the trend in the share of public expenditures is explored, it is of great interest to examine whether the States have harmonised the functional distribution of their public expenditures. An exercise on the composition of expenditure indicates convergence with respect to major development expenditures across the States (Box VI.19 and Chart A to F).

Box VI.1 : Convergence in the Composition of Major Categories of Expenditure across States

There are many definitions of convergence in literature; there are two convergence indicators that have been widely used: β- convergence and α convergence (Barro and Sala-i-Martin, 1992). The former takes place if it is found that when a group of crosssections (States in this case), starting out the sample period with below-average (variable) tend to grow faster than other groups of cross-sections that start with above-average levels, whereas the latter is found when there is a decline in the dispersion of levels over a period of time.

In order to examine convergence of major categories of expenditure, an exercise covering the period 1980-81 to 2008-09 (RE), is attempted to estimate β convergence by estimating the following equation:

E represents the level of per capita expenditure of a State where 0 and T represent the initial and final years respectively, and u_it is an error term.

The above equation is estimated for major categories of expenditure, viz ., revenue expenditure, capital outlay, development expenditure and social sector expenditure.

It can be observed from Table 1 that the coefficient of ä is negative across all the equations tested for beta convergence. It can be inferred that the levels of per capita capital outlay, development expenditure and social sector expenditure have shown some convergence across States as their respective coefficients of ä are negative as well as statistically significant. However, within social sector expenditure, the coefficient of ä in equations tested for per capita expenditure on rural development is negative but not statistically significant indicating convergence not supported by statistical confidence. Per capita spending by States on other categories of social sector expenditure, viz., (i) health; and (ii) education, sport, art and culture have shown significant convergence across States. Among the broad expenditure categories, the rate of convergence as denoted by beta shows that per capita social sector expenditure had the highest rate of convergence across States during 1980-81 and 2008-09 followed by per capita development expenditure. These empirical findings support the trend shown by weighted coefficient of variation (WCV), weighted by the population share of each State, which reflects a changing degree of dispersion in per capita expenditure of select categories. The weighted coefficient of variation is calculated as:

Where E is the relevant expenditure category in per capita terms, Pi/P is share of the State in total population and N is the number of States.

A panel of charts shows that a declining trend is discernible in WCV of broad categories of expenditure (in per capita terms), viz., capital outlay, development expenditure and social sector expenditure across 22 States during 1980-2009. Within social sector expenditure, a declining trend in WCV was observed in per capita health expenditure. However, per capita rural development expenditure across States has shown no convergence during 1980-2009 as was also seen in the beta convergence analysis. While the beta convergence analysis supported increasing convergence in per capita education expenditure across States, the trend in WCV does not seem to support this finding (Charts A to F).

As far the impact of major categories of expenditure on GSDP growth is concerned, a short exercise based on pooled least square method shows that one percentage increase in the capital outlay-GSDP ratio impacts GSDP growth by 0.2 percentage points in the succeeding year. However, the impact of development expenditure and social sector expenditure with one year lag on GSDP growth appears to be lower than that of capital outlay perhaps due to the fact these are dominated by the revenue expenditure component (Table 2). It is found that the impact of capital outlay and development expenditure on GSDP growth of States is statistically significant.

6. Conclusion

6.32 Trend analysis shows that aggregate expenditure of the State governments as a share of GDP increased during the 1980s but fell during the 1990s. Aggregate expenditure as percentage of GDP moved upward during 2000-05. However, compression in consolidated expenditure of State governments can be observed during 2005-10 mainly on account of some rationalisation of revenue expenditure during the fiscal responsibility legislation (FRL) period. This is evident from a decline in the RE-GDP ratio from 13.3 per cent in 2000-05 to 12.4 per cent during 2005-10. The analysis indicated a convergence of the major components of the development expenditure across the States.

¹² The Union List consists of 97 subjects (e.g., defence, atomic energy, railway and telecommunication and insurance). The State List consists of 66 subjects (e.g., police, local government, public health and taxes on agricultural income). There are 47 subjects in the Common or Concurrent List (e.g., education, forests, vital statistics including registration of births and deaths and economic and social planning). The unspecified or residuary powers rest with the Central Government. Any change in these lists can be made only by an amendment to the Constitution.

¹³ Semmler, Willi, et al. (2007), ‘Fiscal Policy, Public Expenditure Composition, and Growth Theory and Empirics’, World Bank Policy Research Working Paper 4405.

¹⁴ Pang Gaobo, Pinto Brian and Wes Maria (2007): ‘India Rising - Faster Growth, Lower Indebtedness’, Policy Research Working Paper Series 4241, The World Bank, Washington.

¹⁵ Capital outlay includes capital expenditure on social and economic services.

¹⁶ Development expenditure comprises expenditure on social services (e.g. education, sports, art and culture, medical and public health, family welfare, water supply and sanitation, housing, urban development, welfare of Scheduled Castes, Scheduled Tribes and other Backward Classes, social security and welfare) and economic services (e.g, agriculture and allied activities, rural development, special area programmes, major and medium irrigation and flood control, energy, industry and minerals, transport, communications, science, technology and environment and general economic services. Non-development expenditure includes expenditure on general services including organs of the State, fiscal services, interest payments and servicing of debt, administrative services, pensions and miscellaneous general services.

¹⁷ Tsujita, Yuko (2005), “Economic Reform and Social Secot Expenditures: A Study of Fifteen Indian States 1980/81-1999/2000”, Discussion Paper No. 31, Institute of Developing Economies - Japan External Trade Organisation, Japan.

¹⁸ Based on Difference of Means Test, i.e., T= (Mt-Mt-1)/S.E. Where Mt and Mt-1 imply the average SGE-GSDP ratio across States during 2005-10 and 1980-85 respectively. S.E. is standard error.