Vertical Fiscal Imbalances and Its Impact on Fiscal Performance: A Case for Indian States

Abstract

Vertical fiscal imbalance (VFI) is defined as the share of sub-national governments’ own spending not financed through own revenues. Theoretical and empirical literatures on VFIs have identified them as an obstacle to sub-national accountability and good fiscal performance. India is a decentralized economy where a marked distinction is made between the spending and revenue responsibilities between the Centre and states. However, spending decentralization has not always complemented revenue devolution, giving rise to huge VFIs in the states. This paper attempts empirically to examine the relation between VFI and fiscal performance for Indian states. It provides stylized facts on size, financing and long-run trend of VFI for 24 major Indian states. Panel data estimation for all those states, covering the period 1995–96 to 2014–15, reveals that on an average the primary deficit of the state governments decline by 15 percentage point of Net State Domestic Product (NSDP) for each percentage point decline in VFI.

Appendix

See Tables 4, 5, 6, 7, 8, 9, 10 and Fig. 7.

Table 4 Developmental functions included in the union lists in the Seventh Schedule of the Constitution

Table 5 Developmental functions included in the state lists in the Seventh Schedule of the Constitution

Table 6 Developmental functions included in the concurrent list in the Seventh Schedule of the Constitution

Table 7 Taxation heads assigned to the union in the Seventh Schedule of the Constitution

Table 8 Taxation heads assigned to the states in the Seventh Schedule of the Constitution

Table 9 Panel Granger non-causality test

Table 10 Result of Granger causality test (state-wise)

Fig. 7

VFI—Evolution overtime for individual states 1. Andhra Pradesh, 2. Arunachal Pradesh, 3. Assam, 4. Bihar, 5. Gujarat, 6. Haryana, 7. Himachal Pradesh, 8. Jammu and Kashmir, 9. Karnataka, 10. Kerala, 11. Madhya Pradesh, 12. Maharashtra, 13. Manipur, 14. Meghalaya, 15. Mizoram, 16. Nagaland, 17. Orissa, 18. Punjab, 19. Rajasthan, 20. Sikkim, 21. Tamil Nadu, 22. Tripura, 23. Uttar Pradesh, 24. West Bengal

Regression-based Fiscal Capacity Approach of Measuring Horizontal Fiscal Imbalances (HFI):

Under this approach, fiscal/taxable capacity is the predicted tax–NSDP ratio estimated from a regression. The estimation procedure takes into account the states’ specific characteristics which include economic, demographic and institutional characteristics.

The basic specification (as given by Le et al. 2008) is

$$ Y = f\left( {\text{GDP, Population, Trade, Agr, Corr, Bureau}} \right) $$

where

Y:

Tax or total fiscal revenue ratio to GDP

GDP:

GDP per capita (constant 2000 $US)

Population: Rate of population growth or age dependency ratio as a share of the total population

Trade: Trade Openness

Agr: Agricultural value added

Corr: Corruption Index

Bureau: Bureaucracy quality.

The problem of extending the methodology at state level in India is simply unavailability of the above-mentioned data for all states and over time.

Hence, this paper employs the following specification:

Y = f (Per capita NSDP_{current prices}).¹⁸

Predicted fiscal capacity $\left( {\widehat{\text{CF}}_{it} } \right)$ is then derived by running a panel fixed effect regression on the above-mentioned specification for the two groups of states, general and special category as:

$$ \widehat{\text{CF}}_{it} = \hat{\beta }\,{\text{NSDP}}_{it} $$

Then, HFI is computed as:

$$ {\text{HFI}}_{it} = \widetilde{\text{CF}}_{t} - \widehat{\text{CF}}_{it} $$

where $\widetilde{\text{CF}}_{t} \,{\text{is}}\,{\text{median}}\,{\text{fiscal}}\,{\text{capacity}}\,{\text{for}}\,{\text{each}}\,{\text{year}}$.

See Table 11; Figs. 8 and 9.

Table 11 Regression result on relationship between VFI and fiscal performance for 23 major Indian states

Fig. 8

Coefficient of variation of HFI 1/ across states over the period 1995–96 to 2014–15

Fig. 9

Trends in Average HFI 1/

Alternate definitions of VFI and fiscal performances:

• 1. VFI Measure:

• (a) VFI = the ability of the state to finance their current expenditure from their own revenue source

$$ {\text{VFI}} = \frac{{{\text{Own}}\,{\text{Revenue}}\,{\text{Source}}}}{{{\text{Current}}\,{\text{Expenditure}}}} $$

Refer Rao and Singh (2002), Table 2 in Appendix and page no. 9

• (b)
  $$ {\text{VFI}} = \frac{{\frac{{{\text{Own}}\,{\text{Tax}}\,{\text{Revenue }}_{\text{SNG}} + {\text{Non}}\,{\text{Tax}}\,{\text{Revenue }}_{\text{SNG}} }}{{{\text{Combined}}\,{\text{Revenues}}}}}}{{\frac{{{\text{Revenue}}\,{\text{Expenditure }}_{\text{SNG}} }}{{{\text{Combined}}\,{\text{Revenue}}\,{\text{Expenditure}}}}}} $$

Refer Muddipi (1991).

Given VFI is detrimental to fiscal performance, the coefficients of VFI is expected to be negative.

• 2. As indicators of fiscal performance, the following measures are applied:

• (a) Gross fiscal deficit as a proportion of total expenditure of the states.

This measure is used as a component index of fiscal performance index as recommended by the Twelfth Finance Commission (Dholakia 2005).

• (b) Ratio of revenue deficit to revenue expenditure of the states

It reveals the proportion of current expenditure to the government financed by the current deficit.

• 3. HFI Measures:

HFI refers to differential fiscal capacities of the sub-national economics. It occurs when sub-national governments have different abilities to raise funds from their tax bases and to provide services. Given non-concurrence on any particular definition of HFI, the following alternate definitions are used.

• (a)
  $$ {\text{HFI = }}\frac{{{\text{State}}\,{\text{per}}\,{\text{capita}}\,{\text{NSDP}}_{{{\text{Current}}\,{\text{prices}}}} }}{{{\text{Average}}\,{\text{of}}\,{\text{highest}}\,{\text{three}}\,{\text{per}}\,{\text{capita}}\,{\text{NSDP}}_{{{\text{current}}\,{\text{prices}}}} }} $$

Given the definition of HFI used in the main paper, it can be categorized as high or low. Thus, we use dummy variable approach for HFI.

• (b) When HFI = Average of highest three per capita NSDP_{current prices} − State per capita NSDP_{current prices}

$$ \begin{aligned} {\text{HFI Dummy}} & {\text{ = 1, when Average of highest three per capita NSDP}}_{\text{current prices}} \\ &\quad > \,{\text{State per capita NSDP}}_{\text{current prices}} \\ & = 0 , {\text{ otherwise}} \\ \end{aligned} $$

• (c) When HFI = Difference between the fiscal capacity of each state and median fiscal capacity

$$ \begin{aligned} {\text{HFI}}\,{\text{Dummy}} & { = 1,}\,{\text{when}}\,{\text{fiscal}}\,{\text{capacity}}\,{\text{of}}\,{\text{each}}\,{\text{state}}\,{ > }\,{\text{median}}\,{\text{fiscal}}\,{\text{capcity}} \\ & = 0 ,\,{\text{otherwise}} \\ \end{aligned} $$

Recording HFI as a dummy variable makes the regression results more interpretable. It captures the influence of VFI on the fiscal performance of the states, in the presence of differential fiscal position among the states.

Robustness Analysis Results:

See Tables 12, 13, 14, 15, 16, 17, 18 and 19.

Table 12 Regression result on relationship between VFI (alternate measures) and fiscal performance for 14 major general category states

Table 13 Regression result on relationship between VFI (alternate measures) and fiscal performance for 10 major special category states

Table 14 Regression result on relationship between VFI 1/ and fiscal performance (alternate measures) for 14 major general category states

Table 15 Regression result on relationship between VFI 1/ and fiscal performance (alternate measures) for 10 major special category states

Table 16 Regression results—sensitivity analysis

Table 17 Regression results—sensitivity analysis

Table 18 Regression results with HFI as interaction term

Table 19 System dynamic panel data estimation results 1/