Hostname: page-component-77c89778f8-m8s7h Total loading time: 0 Render date: 2024-07-16T21:53:28.970Z Has data issue: false hasContentIssue false
  • English
  • Français

Towards A Predictive Theory of Government Expenditure: Us Domestic Appropriations

Published online by Cambridge University Press:  27 January 2009

Otto A. Davis ,
M. A. H. Dempster  and
Aaron Wildavsky
Get access Rights & Permissions [Opens in a new window]

Extract

The project on which this paper reports is aimed not only at increased understanding of the United States federal budget process, but also at predicting government expenditures in total and by bureau with a view to their determination within United States national econometric models. Estimates of likely expenditures using standard econometric techniques are poor, both in absolute terms and in comparison with our own work. Management of the economy should be improved by the use of predictors based on considering budgeting as a political process that is responsive to economic and social conditions. Use of mathematical models in the social sciences should be furthered, not by arguing their hypothetical utility, but by demonstrating that they work. The proof is in the prediction.

Type
Article
Information
British Journal of Political Science , Volume 4 , Issue 4 , October 1974 , pp. 419 - 452
Copyright
Copyright © Cambridge University Press 1974

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 Wildavsky, Aaron, The Politics of the Budgetary Process (Boston: Little, Brown, 1964).Google Scholar

2 Davis, O. A., Dempster, M. A. H. and Wildavsky, A., ‘A Theory of the Budgetary Process’, American Political Science Review, LX (1966), 529–47CrossRefGoogle Scholar; Davis, O. A., Dempster, M. A. H. and Wildavsky, A., ‘On the Process of Budgeting: An Empirical Study of Congressional Appropriation’, Papers in Non-Market Decision Making, I (1966), 63–132Google Scholar; Bowman, G. W., Davis, O. A., Gaillot, H. J. and Hess, A. C., ‘A Note on Supplemental Appropriations in the Federal Budgetary Process’, Papers in Non-Market Decision Making, II (1967), 91–101Google Scholar; Davis, O. A., Dempster, M. A. H. and Wildavsky, A., ‘On the Process of Budgeting II: An Empirical Study of Congressional Appropriation’, Chap. 9 in Byrne, R. F., Charnes, A., Cooper, W. W., Davis, O. A. and Gilford, Dorothy, eds., Studies in Budgeting (Amsterdam: North Holland, 1971).Google Scholar

3 Davis, Dempster, and Wildavsky, ‘On the Process of Budgeting II’.

4 Peacock, A. T. and Wiseman, T., Public Expenditures in the United Kingdom: 1900–1952 (London: National Bureau of Economic Research, 1962).Google Scholar

5 Dhrymes, P. J., Howrey, E. P., Hymans, S. H., Kmenta, J., Learner, E. E., Quandt, R. E., Ramsey, J. B., Shapiro, H. T. and Zarnowitz, V., ‘Criteria for Evaluation of Econometric Models’, Annals of Economic and Social Measurement, I (1972), 291–324.Google Scholar

6 For a justification of their use, see Davis, , Dempster and Wildavsky, ‘A Theory of the Budgetary Process’, §4, p. 540.Google Scholar

7 Davis, Dempster and Wildavsky, ‘On the Process of Budgeting II’.

8 See Davis, O. A., Dempster, M. A. H. and Wildavsky, A., ‘Toward a Predictive Theory of the Federal Budgetary Process’ (mimeo), delivered at the Annual Meeting of the American Political Science Association, 09 1973.Google Scholar

9 See Niskanen, W., ‘Controllability of the Fiscal Variables’, Evaluation Division Report, Office of Management and Budget, 1971.Google Scholar

10 For discussion of estimation problems, see Davis, Dempster and Wildavsky, ‘Toward a Predictive Theory of the Federal Budgetary Process’.

11 See Wildavsky, , The Politics of the Budgetary Process, pp. 16–17, 154–5Google Scholar, for discussion of fair-shares.

12 In our earlier studies we rejected the constant term a priori in the specification of our basic models. In the research reported here, we allowed the constant term to remain in the extended model when it was significant by the appropriate criterion. Significance of the constant term when the model is basically correct could be a result of mispecifications involving omitted variables or non-linearities. Both effects are exacerbated in our models by process aggregation. In our paper, ‘Toward a Predictive Theory of Government Expenditure: III. U.S. Domestic Appropriations Over Two Recent Five Year Periods’, Zeitschrift fur Nationalekonomie (to appear), we report on the positive relation between omission of the constant term and significance of the gaming term (y − x)–1 i.e. the difference between appropriation and request in the previous year. Otherwise the specifications resulting from forcing out the constant term hardly changed.

13 In some cases, however, the estimation bias of known once-for-all events, e.g. Sputnik, has been removed by adding a binary dummy variable to our set for potentially affected agencies, e.g. NASA, NSF, AEC, etc.

14 Of course not all exogenous binary dummy variables which changed over the estimation period did so over the prediction period, so they cannot be used to test the increased explanatory power of the extended model. Changes took place in 3 (HDM), 4 (HLDM), 7 (Pre-el), 10 (Ec. Rec), and 15 (War).

15 See e.g. Dhrymes, Howrey, Hymans, et al., ‘Criteria for Evaluation of Econometric Models’, and Howrey, E. P., Klein, L. R. and Mccarthy, M. D., ‘Notes on Testing the Predictive Performance of Econometric Models’, Discussion Paper No, 173, Wharton School, Department of Economics, University of Pennsylvania, 1970.Google Scholar

16 Davis, Dempster and Wildavsky, ‘On the Process of Budgeting II’.

17 Davis, Dempster and Wildavsky, ‘On the Process of Budgeting II’.

18 This explanation will be tested by a new automatic procedure which we are currently developing and applying to the same sample. The stages of our step-wise procedure were however carefully monitored by hand, see §111 above, and the significant continuous variables make sense when they lead to good prediction, see §VII below.

19 See Davis, Dempster and Wildavsky, ‘Toward a Predictive Theory of the Budgetary Process’ for details.

20 Before analyzing the source of these disturbances, note that the better fit and predictive performance of the appropriation equation is reflected in its ‘explanation’ of the cross-sectional variance of appropriations — about 80 per cent in the worst year versus about 44 per cent for the corresponding figure for OMB requests in the worst of the unexceptional years F1964 and 1965 — both respectable figures. In these years the system also ‘explains’ at worst about 41 per cent of the cross-sectional variance in appropriations, reinforcing the point that good request prediction is the key to good system prediction.

21 Galper, H. and Wendell, H. F., ‘Progress in Forecasting the Federal Budget’, Proceedings of the Economics and Business Section of the American Statistical Association, 1968, pp. 86–98.Google Scholar

22 Cf. Galper, and Wendell, , ‘Progress in Forecasting the Federal Budget’, p. 87.Google Scholar

23 Howrey, Klein and McCarthy, ‘Notes on Testing the Predictive Performance of Econometric Models’.

24 Aidan Vining, a student at the Graduate School of Public Policy in Berkeley, helped prepare this section.

25 In one of the eight cases shown in Table 6, the Interstate Commerce Commission, the ‘extended’ equation was actually the basic request equation (2.1), but fit over the whole post-war period, even though our earlier work showed a significant shift in behaviour in F1955.

26 The ‘extended’ equations of two of the sixteen agencies shown in Table 1, the Agricultural Research Service and the Bureau of Narcotics, were the basic equation (2.4) fitted with an additive binary dummy variable to take account of a non-recurrent event discovered in our previous work, cf. fn. 2.

27 Public Health Service (6), Executive Office of the President (4), NASA (3) and Interstate Commerce Commission (1).