TABLE 11.-ANNUAL GROWTH RATES OF TOTAL GOVERNMENT EXPENDITURE IN CONSTANT PRICES Source: January 1979. "Economic Report of the President," table B-77, p. 272. In real terms, total government expenditure rose at annual rates well below those of nominal expenditure. The median rate-6.2 percent for the full period— advanced from 5.0 percent in the Fifties to 6.9 percent in the Sixties, then declined to 6.2 percent in the Seventies. In 1977 real expenditure, in the median case, was only 3.8 percent above that of 1976. Among individual countries the annual growth over the full period ranged from a low of 3.2 percent (United Kingdom) to a high of 7.3 percent (Netherlands). Greece, which had one of the lowest growth rates in the Fifties, experienced the highest annual growth rate of real expenditure after 1960. West Germany, with a growth rate close to the median for the full period, had the highest rate of increase in the Fifties, the lowest in the Seventies, and a relatively low rate in the Sixties. As was noted in the preceding section, a large part of the increase in nominal expenditure is attributable to price and cost increases. Deflator values for total government expenditure are displayed in Appendix Table A-3 but annual rates of increase were calculated only for the full period. These have not been formally correlated with the growth rates of real expenditure, but inspection reveals that the former are inversely associated with the latter. The United Kingdom, for example, which had the lowest growth rate of real expenditure, experienced the second highest rate of increase in the deflator; and several countries with relatively high growth rates of real expenditure had relatively small increases in deflator values. Changers in real size of the public sector The ultimate objective of this study is a realistic measure of public-sector size. Simple ratios of government expenditure to GDP, which make no allowance for relative price changes, are only a first approximation. Deflation of total expenditure by a price index which fits only a part of the total is equally misleading. Separate deflation of the major components has enabled us to estimate the deflated value of total government expenditure which, when expressed as a percentage of real GDP, measures real size of the total public sector. Appendix Tables A-5 and A-6 show the complete set of public sector ratios in current and constant prices. Percentages for the initial and terminal years of the study are reproduced in Table 12. To accommodate Denmark and Ireland, which have not supplied full data for 1977, results are shown for 1976 as well as 1977. In current prices government spending, as a percentage of GDP, rose sharply over the full period. From 21 percent in 1950 the median percentage rose to nearly 40 percent in the final year. In two cases-Sweden and the Netherlands-total government spending amounted to more than half of aggregate spending for GDP.22 22 The reader is reminded that, although transfer expenditure is part of the total, it is not part of GDP. In comparative studies of public expenditure, GDP is now the preferred base against which government spending is measured despite occasional arguments in favor of alternative bases. 54-267 0- 8029 TABLE 12.-TOTAL GOVERNMENT SPENDING AS PERCENTAGE OF GROSS DOMESTIC PRODUCT Source: January 1979, "Economic Report of the President," table B-77, p. 272. Inspection of Appendix Table A-5 reveals that the median percentage has risen to nearly 26 percent by 1960, to 33 percent by 1970, and to 40 percent by 1975. The OECD study mentioned at the beginning of Part III, using approximately the same definition of total expenditure, found that on average government spending had reached 41.4 percent of GDP in the mid-1970's, up from 34.5 percent in the late 1960's.23 In almost all of these countries pressures have emerged to restrain the growth of public expenditure, but the OECD report predicts "that the most likely outcome for future public expenditure is that it will tend to increase its share of resources in the long run". (p. 37) The forecast is appropriately hedged, and this writer agrees that the ratio of government spending to GDP may continue to rise; it is unfortunate, however, that the OECD report identifies as increase in total spend ing with an increase in government's share of resources. Column 5 of Table 12 shows public-sector size in 1977 in real terms. Total government spending at 1950 prices, as a percentage of real GDP, ranged from a low of 24 percent (Greece) to a high of 48.3 percent (Netherlands). These two countries also occupied the extremes of the range in 1976 when the median per centage was slightly less than in 1977. Appendix Table A-6, which displays real ratios for the intervening as well as terminal years, tells us that the gap between nominal and real size of the total public sector widened significantly after 1970. Since 1975 the median ratio of G to Y, calculated from data in 1950 prices, has been relatively constant-at about a third of GDP. Once again, the reader is reminded that an increase in real size of the public sector does not necessarily mean an increase in government's claim on resources In seven of the thirteen countries the ratio of G. to Y, in real terms, declined over the study period; and in the median case it rose hardly at all-from 12.1 percent in 1950 to 12.7 percent in 1977. Transfer outlays, which do not involve the flow of factor inputs to the public sector, accounted for nearly all of the 12.5 percentagepoint increase in real size of the public sector (median case, 1950 to 1977). Among the six countries in which the ratio of G, to Y (in real terms) rose over the full period, several experienced little or no growth since 1970. In Canada, for the 1977 ratio was 11.6 percent. The decline since 1970 was even more pronounced in the United States, where a rise from 12.1 percent (1950) to 16.8 percent (1970) countries since 1970 parallels that of the seven countries with G. elasticity cowas followed by a decline to 13.4 percent (1977). The experience of these two efficients below unity for the full period (Table 1, column 2). averages are generally consistent with the findings of this study. 23 Public Expenditure Trends (June 1978) table 2, pp. 14-15. No data are shown for 19 50, but the OECD TABLE 13.-INCOME ELASTICITY OF GOVERNMENT EXPENDITURE IN CONSTANT PRICES Source: United Nations, "Yearbook of National Accounts Statistics," 1977 and 1964; also, replies to questionnaire for 1978 Yearbook (unpublished). Elasticity of total government spending For a given country expansion of the public sector over the study period is accurately portrayed by the initial and terminal ratios. For cross-national comparisons, however, we need a measure that is independent of public-sector size at the beginning of the period. The elasticity of government spending with respect to GDP, which measures the percentage change in government spending relative to the percentage change in GDP, facilitates comparison between countries. A full set of elasticity coefficients, for major components as well as total government expenditure, is displayed in Table 13. Columns 1 to 4 present the record of these countries for total expenditure. In the 1950s four countries-those with coefficients below unity-increased real expenditure by a smaller percentage than the percentage increase in real GDP. In the other nine countries the percentage increase in real expenditure exceeded that of real GDP. In the 1960s and 1970s (columns 2 and 3) all coefficients are greater than unity, signifying expansion in real size of the public sector. In Sweden and Switzerland the unusually large expansion in the 1970s is attributable to the relative growth of transfer outlays, as a comparison of their coefficients in column 11 with those of column 7 will show. In the median case the percentage increase in total government spending over the full period exceeds that of real GDP by 80 percent. The excess rose progressively over the 27-year period from 23 percent in the 1950s to 61 percent in the 1960s to 130 percent in the 1970s. The median values for government consumption expenditure also rose over the three sub-periods (columns 5 to 7), but were less than unity prior to 1970. That is, between 1950 and 1970 the percentage increase of real expenditure for government consumption fell below that of real GDP. This was true also for the full period. Finally, the median value of "transfer" coefficients was substantially higher in the 1960s than in the 1950s, but dipped slightly in the 1970s. For the period as a whole the median coefficient signifies that the percentage increase of real transfer outlays was nearly three times that of real GDP. The results of this study cannot be readily summarized. They do emphasize the importance, however of (1) translating nominal expenditure of government into real expenditure; and (2) distinguishing between consumption expenditure, most of which serves the entire community, and transfer outlays, which are directed at specific groups of beneficiaries. Over the study period the dominant trend in these countries was a re-allocation of funds away from the former and toward the later. |