¹ See, for example, Cyert, Richard and March, James, eds., A Behavioral Theory of the Firm (Englewood Cliffs, N.J.: Prentice-Hall, 1963)Google Scholar; Braybrooke, David and Lindblom, Charles, A Strategy of Decision: Policy Evaluation as a Social Process (New York: Free Press of Glencoe, 1964)Google Scholar; Davis, Otto A., Dempster, M. A. H., and Wildavsky, Aaron, “A Theory of the Budgetary Process,” American Political Science Review, 60 (09, 1966), 529–547CrossRefGoogle Scholar; Wildavsky, Aaron, The Politics of the Budgetary Process (Boston: Little, Brown, 1964)Google Scholar; Crecine, John P., Governmental Problem-Solving: A Computer Simulation of Municipal Budgeting (Chicago: Rand McNally, 1969)Google Scholar.

² Some comparative work has begun to appear in the research literature. See, in particular, Aaron Wildavsky. Budgeting (forthcoming): Heclo, Hugh and Wildavsky, Aaron, The Private Government of Public Money: Community and Policy Inside British Politics (London: Macmillan Company; Berkeley and Los Angeles: University of California Press, 1973)Google Scholar; and Campbell, John Creighton, The Japanese Budgetary Process (doctoral dissertation, Columbia University, 1973)Google Scholar; and Olsen, Johan P., “Local Budgeting, Decision Making or Ritual Act,” Scandinavian Political Studies, 5 (1970), 85–113CrossRefGoogle Scholar.

³ Sharkansky, Ira, “Agency Requests, Gubernatorial Support, and Budget Success in State Legislatures,” American Political Science Review, 62 (12, 1968), 1220–1231CrossRefGoogle Scholar.

⁴ Crecine.

⁵ Crecine, pp. 218–219.

⁶ Davis, Dempster, and Wildavsky.

⁷ Davis, Dempster, and Wildavsky, p. 530.

⁸ A fairly comprehensive review and discussion of the postwar economic development in Norway has been published by the Norwegian Central Bureau of Statistics under the title The Norwegian Post-War Economy, SØS 12 (Oslo, 1965)Google Scholar. We have also benefited from publications by the Municipal Office of Statistics, particularly, the Statistical Yearbook for the City of Oslo.

⁹ We have not included the balanced-budget constraint for the Chief Municipal Administrator in the formal decision models set out below; and we see that omission as one of the weaknesses of these kinds of formulations. Nonetheless, we have little reason to believe that systematic bias across agencies or across years is introduced because of the omission. In addition, such a constraint is probably of less importance in the Scandinavian context. Flexibility for some American municipalities is restrained by (1) requirements that tax increases be approved by local referenda, and (2) certain cumbersome methods for approval of municipal loans. The result is that the balanced-budget constraint for the city manager becomes crucial in the American context. While the balanced-budget constraint is undoubtedly important in the Scandinavian context, procedures for obtaining additional revenue—in the event of excessive budgetary requests—are generally less cumbersome. In Oslo, tax increases (within certain bounds set by the national government), as well as rate increases for municipal services, are approved by a simple majority vote in the Council; and loans on the domestic and international market provide an additional option for the Chief Municipal Administrator faced with increasing demands on the local treasury. The result is that the balanced-budget constraint may be less central than would be the case for many American cities. Yet, we do not wish to minimize its continued importance here, and we refer to it informally throughout the paper.

¹⁰ For those familiar with the Scandinavian languages, the following English-Norvegian translations may be helpful; agency chief-etatsjef; department head-rådmann; Chief Municipal Administrator— Finansrådmann; City Council—Bystyret. A suitable translation for finansrådmann proved to be particularly difficult. Some may prefer the translation, Head of the Finance Department or Budget Director. While these terms are closer to a literal translation, they are misleading in terms of the functions actually performed. The scope of the Finansrådmann's activities is quite close to that for the American city manager. In any event, the scope of those activities is definitely broader than would be suggested by either of the alternative terms above. In addition, our translation is the official English translation used by the Finansrådmann's Office.

¹¹ In each case we assume the initial basis of calculation to be consistent with the notion of process in-crementalism—i.e., based initially upon the decision taken immediately prior in the process. This notion of an initial basis of calculation implies that there is one figure which serves as the fundamental basis of judgment. Subsequent alterations in that initial judgment appear as additional variables in the models and contribute to the final judgment for the decision in question.

¹² The equation is presented in Table 1d in a form for which the interpretation is more straightforward, given the kind of behavior it attempts to model. The actual estimation equation is of the form:

¹³ For treatments of estimation methods in regression models, see: Johnston, John, Econometric Methods (New York: McGraw-Hill, 1963)Google Scholar; Christ, Carl F., Econometric Models and Methods (New York: Wiley and Sons, 1966)Google Scholar.

¹⁴ Durbin, J. and Watson, G. S., “Testing for Serial Correlation in Least-squares Regression,” part I, Biometrika, 37 (12, 1950), 409–428Google ScholarPubMed; part II, Biometrika, 38 (06, 1951), 159–178CrossRefGoogle Scholar.

¹⁵ Similar criteria were used in the Davis, Dempster, and Wildavsky work cited above except that the use of tests of significance for the constant term was not indicated in the published report.

¹⁶ These data were taken from the following budget documents published by the Oslo municipal government: Sak 1A: Finansrådmannens Forslag til Driftsbudsjett and Dokument nr. 3: Oslo Bykasses Driftsbudsjett, editions for 1953 through 1971.

¹⁷ A list, by agency, of all estimated equations actually selected may be obtained from the authors upon request.

¹⁸ Sharkansky, , “Agency Requests,” pp. 1230–1231Google Scholar.

¹⁹ See, for example, Fenno, Richard, The Power of the Purse, chapters 10–12 (Boston and Toronto: Little, Brown, 1966)Google Scholar.

²⁰ Davis, , Dempster, , and Wildavsky, , “A Theory of the Budgetary Process,” p. 530Google Scholar.

²¹ Considerable care was taken in the interview setting in obtaining these responses. The respondent was introduced to this series of questions by the following: “We'd like to ask you to recall, in as much detail as possible, the actions concerning your budgetary requests last year. We realize that some points you will recall quite clearly but that others you won't be able to remember. Please feel free simply to tell us if you don't remember certain things.” In this way we hope to distinguish between those elements of the decision structure which lie at the surface of individual cognizance and those which do not. In addition, the timing of the interview was quite important. Respondents were interviewed after they had made their budgetary decisions for the upcoming year but before any final decision had been made on appropriations.

²² It is important to note that, while there is some variation across the six questions, this is by and large an individual phenomenon. That is, respondents found in the don't know category on one question are likely to be found in that category on other questions as well.

²³ See, in particular, Brofoss, Karl-Erik, “Interesseor-ganisasjonens og Budsjetteringsprosessen i Oslo Kommune: En Partiell test av et teoretisk resonnement,” Institute for Political Science, University of Oslo, mimeo, 1975Google Scholar; Cowart, Andrew T., “Partisan Politics and the Budgetary Process in Oslo,” American Journal of Political Science, 19 (11, 1975)CrossRefGoogle Scholar forthcoming; and Cowart, Andrew T., “The Machiavellian Budgeter,” mimeo, 1975Google Scholar.