CHAPTER VII INCOME ELASTICITIES OF GOVERNMENT EXPENDITURE

Transcription

1 CHAPTER VII INCOME ELASTICITIES OF GOVERNMENT EXPENDITURE In this chapter an attempt is made to measure the relationship between the growth of state government expenditure and state domestic product. The chapter begins with a brief review of previous attempts at measuring relationship between government expenditure and income in terms of elasticity coefficient. In general terms, these researches have concentrated on the measurement of the elasticity (E) of government expenditure (G) with respect to some measures of income, (1), EG.I' that is, the percentage change in government spending associated with a one percent change in income. All these studies have basically tried to test the validity of Wagner's Law. Gandhj1 has pointed out that there are basically five different interpretations of Wagner's Law. First, Peacock and Wiseman's conclusion implies that the elasticity of G with respect to GNP, EG.GNP must exceed one if Wagner's Law is valid. Thus Gandhi concludes that the symbolic statement according to this version appears to be that E=f (GNP), where E stands for the level of government expenditure and GNP stands for gross national product. and that the elasticity of public expenditure with respect to Gross National Product (ee.gnp) is greater than unity. Seiendly, Hook, Emi, 1 Ved P. Gandhi, "Wagner's Law of Public Expenditure: Do Recent Gross Section Studies Confirm It?", Public Finance No. 26, pp Alan T.Peacock and Jack Wiseman, ''The Growth of Pubic Expenditure in the United Kingdom", George Allen and Unwin, London, 1967, p

2 Veverka, Andie and Veverka and Pryor have computed elasticity of government expenditure with respect to national income (Y) rather than GNP 3 In this case EG.v must be > 1 for the law to be valid. Goffman has used output per capita (GNP/POP) in the elasticity measurer and for the validity of law the condition to be satisfied is EG.(GNPtPoP?1. 4 Fourth, Michas, has used per capita government expenditure with respect to per capita output, or E(GtPOP). (GNPtPOPl' to calculate elasticity. 5 Finally, Musgrave, Pryor, and Wagner and Weber have public expenditure with respect to gross national product (ee.gnp) is greater than unity. Secondly, Hook, Emi, Veverka, argued that the correct elasticity is E(GtGNP>.(GNPtPOPl' the elasticity of government expenditure relative to output with respect to output per capita 6 Table 7.1 summarises 15 empirical studies. The table very clearly indicates the fact that we cannot make any generalized conclusion regarding relationship between government expenditure and income. Elasticities for 12, out of the 20 reported, exceed one. 3 Hook, Erik, "The Expansion of the public Sector - A study of the Development of public Civilian Expenditures in Sweden during the years ", Public Finance (N0.4, 1962), pp , Koichi Emi, "Government Fiscal Activity and Economic Growth in Japan, " (Tokyo: KinoKuniye Bookstore Co., 1963); Hindrich Veverka, ''The Growth of Government Expenditure in the United Kingdom since 1790", Public Expenditure; Appraisal and Central, eds. A.T. Peacock and D.J. Robertson (Edinburgh:Oiiver and Boyd, 1963), pp ; Suphan Andie and Jindrich Veverka, ''The Growth of Government Expenditure in Germany Since the Unifiction", "Finanzarchiv (January, 1964), pp ; Frederich L. Pryor, "Public Expenditures in Communist Nation" (London: George Allen and Unwin, 1968). 4 Gottman, I.J. and D.J. Mahar, "The Growth of Public Expenditures in Selected Developing Nations: Six Cariffean Countries, ", Public Finance (NO. 3, 1968) pp Michas, A.S. "Wager's Law of Public Expenditure : What is the Appropriate Measurement for a Valid Test?" Public Finance (No. 1, 1975) pp Musgrave, op.cit. pp 73-74; Pryor, op.cit. p.63, and Richard E. Wagner and Warren Weber, 'Wager's Law, Fiscal Institutions and the Growth of Governmenf', National Tax Journal (No. 1, 1971 ), pp

3 Table 7.1 Summary of Fifteen Studies of Wagner's Law Study 1. Martom and Lewis (1956) 2. Williamson (1961) 3. Hook (1962) E(G.Y) 4. Emi (1963) 5. VEVERKA (1963) 6. Andie and Veverka (1964) 7. Thorn (1967) 8. Gottman and Mahar (1968) 9. Gupta (1968) 10. Pryor (1968) 11. Lall (1969) 12. Musgrave (1969) 13. Henning and Tussing (1974) 14. Wagner and Weber (1977) 15. Pluta (1979) Elasticity Estimated and Data Base E(G/GNP).(GNP/POP), Various countries cross-section data E(G/GNP).(GNP/POP), Various countries, cross-section data Sweden E(G. Y), Japan, E(G.Y),United Kingdom, E(G.Y) Germany, E(G.GNP). (GNP/POP), Various countries, Cross-Section data E(G.GNP), Six Caribbean Countries, ( ) E(G/GNP), (G.GNP/POP) Various countries, cross section data E(G/GNP),(GNP/POP), Various Countries Cross Section data Developing E(G/GNP),(GNP/POP), 46 Countries Cross Section data E(G/GNP),(GNP/POP), 40 Countries Cross Section data E(G/GNP),(GNP/POP), United States E(G/GNP),(GNP/POP), 34 Foreign nations E(G/GNP),(GNP/POP), Taiwan, Expenditure Elasticity Findings > 1 for all functions > 1 for all economic categories > 1 by economic category > 1 by function > 1 by function > 1 by economic category > 1 by economic category Source James T. Bennett and Manueal H. Johnson. "The Political Economy of Federal Government Growth " The Center for Education and Research in Free Enterprise, Texas A and M University, College Station, Texas, 1980, p

4 Bird argues that due to changes in expenditure base one cannot derive meaningful inferences about the future from observed past income elasticities. Therefore, according to him, the notion of the income-elasticity of public expenditure is singularly barren of fruitful hypotheses "explaining" in any meaningful sense the.growth of pubic expenditure 7 Income Elasticities of Government Expenditure for Gujarat State We here provide the results of our own study of the relationship between government expenditure and income. As mentioned earlier, most of the results pertaining to elasticities in the earlier studies were derived only as a test of the Wagner's Law of increasing state activities with development. However, our intention is not of testing the Wagner's hypothesis and to make any predictions about future relationship between expenditure and income on the basis of past trend. This is because as Bird has rightly pointed out, one has to conduct a more detailed analysis of the changes in the "expenditure base", which we have not been able to do here. However, this can be taken as one of the areas for future research. 7 Bird, Richard, M. "The Growth of Government spending in Canada", Canadian Tax Papers, July 1970, Canadian Tax FoundatiOIJ, 100 University Avenue, Toronto 1, Canada, pp

5 The following equation has been used to calculate the elasticity: log G = a + b log Y (1) Where G = Government Expenditure Population Y= SOP Population Taking derivative of equation (1) with respect to Y we get. d(log G) d(log Y) = b (2)* Which shows that coefficient 'b' in eq (1) represents elasticity. Tables 7.2 and 7.3 give estimates of elasticity of expenditure for economic and functional categories respectively. Here, d(log G) d(log Y) = Y dg G dy 131

6 Table 7.2 Estimate of Income Elasticities of Expenditure for Economic Categories { to ) Variables Elasticity** T-values+ T-values+ coefficient (b=1) (b=o) 1. Wages and salaries * 58.75* 2. Goods and Services * 3. Total Consumption Expenditure * 36.08* 4. Interest Payments * 24.06* 5. Subsidies * 17.38* 6. Transfer to Local Bodies * 7. Other Current Transfers * 44.54* 8. Total Transfers * 9. Total Current Expenditure * 10. Gross Capital Formation * * 11. Capital Transfers to Local Bodies * 12. Other capital Tran sfers and * Investments 13. Other Loans and Advances * * 14. Capital Transfers * 28.87* 15. Total Capital Expenditure * 21.21* 16. Total Expenditure * 52.18* ** Refers to the regression coefficient of per capita expenditure at current prices on per capita SOP at current prices. Significant at 1% level. + T-values are worked out for two sets of null hypothesis, (i) b=o, (ii) b=1 132

7 The above table shows that the income elasticity of expenditure for all the economic categories but one (goods and services), is both numerically as well as statistically either equal to one or greater than one. This implies that for most of the items of expenditure (statistically 9 out of 15 results), one percent increase in income (State Domestic Product) leads to greater than one percent increase in expenditure, thus showing high responsiveness of various these expenditures heads to changes in state domestic product. In other six cases the estimated elasticities are not significantly different from one. Table 7.3 Estimate of Income Elasticity of Expenditure for Functional Categories ( to ) Variables Elasticity** T-values+ T-values+ coefficient (b=1) (b=o) 1) 2) 3} 4) 5) 6) 7) 8) 9) 10) 11} General Services * * Education * * Health * * Agriculture * * Industry and Mining * * Transport and Communication * * Other Economic and Social Services * Economic and Social Services * * Miscellaneous * * Interest Payment * * Total * * ** * + Refers to the regression coefficient of per capita expenditure at current prices on per capita SOP at current prices. Statistically significant at 1% level. T-values are worked out for two sets of null hypothesis, viz. (i) b=o, and b=1. 133

8 Table 7.3 shows the income elasticity of expenditure for functional categories. The income elasticity of expenditure for 'health' is numerically as well as statistically speaking less than one indicating that one per cent change in SOP leads to less than one per cent change in expenditure on health. In other expenditure categories the estimated elasticities are greater than one. However, statistically speaking the elasticity coefficient for 'Other Economic and Social Services ' category is not significantly different from one. In all other categories as well as for total expenditure income elasticities are statistically speaking significantly different from unity. The highest income elasticity of expenditure is for interest payments. 134 '