which funds are now allocated in the form of annual white papers. rks. It has given the Treasury considerably greater control than it had e exercised over the growth of public expenditure. It is now far more han it used to be to sneak in small items with large future expenditure ɔns. Containing the vices of its virtues, as most human mechanisms do, so makes it correspondingly more difficult for old items to be taken out. ake sure that it has an ample margin of safety, the Treasury has adopted devices-the relative price effect and productive potential-to make sure ters do not get out of hand. Based on the questionable assumption that ity in the public sector does not grow at the same rate as in the private e relative price effect leads to a proportion of each year's budget being 7 for safe keeping in case expenditures should go that much beyond ons. Productive potential declares, by fiat, that productivity will not at more or less than something under 3%, thus establishing a small and margin for future increases. Thus PESC works within a margin of with a safety factor added on.*

entalism is nowhere better practiced than in Japan. The Japanese can
mental anyway you come at them, bottom-up, top-down, and seemingly
as well. To begin with, the Ministry of Finance imposed a 125% ceil-
the past year. A quarter of the previous year's budget might seem
re it not for the fact that past expenditures are not challenged at all,
expected that every ministry will get built-in increases amounting to a
f or more of that. In making up its critically important draft budget,
, the Finance Ministry adds both "Natural Increases" (cost of living and
ry items) and "Semi-natural Increases" (those that are not required
›r price increases but which every sensible man knows will have to
). Should the question arise of whether one ministry is to be preferred
ers in obtaining increases, the norm of balance (baransu) is applied
omparable programs and categories receive essentially the same propor-
hat there is to give. By the time Finance holds its own Important Items
n, it is hardly necessary to think at all in putting out its draft budget.
er all this, only a beginning has been made in performing the Japanese
omy on the yearly budget. Concerned lest there still remain grievances
ed, the Ministry of Finance has put away a few percent of the total so
those who feel they have lost out in the original process can participate
are variously called the "resurrection" (literally, to come back from
) or the "revival" negotiations. Within this rapidly diminishing margin,
à increment of an increment, the Ministry of Finance allocates pro-
across-the-board for salary and policy adjustment expenses, so that
p party officials and the Minister of Finance sit down for the final
s, they are not overburdened with numerous hard choices."

apparent that central budget control organs in Britain, France, and
ave largely given up, or retreated from, detailed scrutiny of budgetary
s. They no longer appear anxious to use the budget as a means of im-
heir priorities on the rest of the government. What has happened? Why
ey begun to abandon their ancient powers? What, if anything, have they
eturn? Finance ministries have trimmed their old expenditure powers
better preserve their new authority over economic management.
ending has risen, so have the spenders. Growing ever stronger from their
tories, feeding on the desires for national grandeur or for social jus-
teous in their wrath against those who would deny citizens necessities
they have put their former controllers on the run. Threatened, for in-
by the Liberal Democratic Party in Japan and the spending ministers else-
uch as those who allegedly ganged up on the Treasury's Chancellor of the
er in the 1950s, central control organs have had to consider their
. At the same time they have also been deeply affected by the Keynesian
on in economic thought. Except in the United States, where the Office of
ment and Budget is not in the same organization as the Treasury De-
at or the Council of Economic Advisers, the men whose job it was to con-
enditure are part of a larger apparatus whose tasks include economic
ment. As politician and civil servant alike has grown assured and com-
e chapter on PESC in Hugh Heclo and Aaron Wildavsky, The Private Government
Money: Community and Policy in British Political Administration (London: Mac-
rthcoming).
ohn Creighton Campbell, "Japanese Balanced Budgeting", prepared for the Re-
onference on Japanese Organization and Decision-Making, Maui, Hawaii, Jan. 5-10,
How Powerful is the Ministry of Finance?", University Seminar on Modern East
umbia University, Dec. 10, 1972; and The Japanese Budgetary Process (Doctoral
on, Columbia University, in progress).

B

B

fortable with seemingly simple Keynesian decision rules-spend when ployment is low, cut back when it is high-they have identified their f increasingly with ability to affect the economic situation. Reducing their of calculation has meant not merely considering increments to the budg viewing expenditure as part of economic management. Their interest i expenditures far exceeds the concern they normally have with specific pr that go toward making it up Ministers of finance and their civil service a are not usually selected for their skill at cost-benefit analysis, which mi appropriate for choosing among expenditure projects, but for their abi handle larger economic aggregates. And the politician's interest in holding or manipulating the tax rate is also served by the same macro-econom proach. Hence an implicit bargain has been struck. Ministries of finance ha gun to give up detailed financial scrutiny (made increasingly difficult, i case, by the growing size of government) in return for control over the tot aid economic management.

Though the United States has neither institutionalized incrementalism established the equivalent of a Finance Ministry, there are signs it is movi the direction of emphasizing totals in order to make expenditures the serva economic management. What is revenue sharing, after all, but a means of tralizing totals and decentralizing priorities within these umbrella figures? does the creation of a super-secretary in charge of economic affairs with an near the President mean but a desire to better control management of the e omy? What, indeed, is the significance of recent changes in the Office of Man ment and Budget-the split between administration and budgeting and the e parallelism between budget divisions and the policy spheres of the new dome super-secretaries-other than a greater concern with totals and a lesser emph on what goes into them? For the budget arm of the OMB is being transfor into staff assistants to the super-secretaries, much as the previous military d sion participated directly in drawing up the defense budget as an aid to Secretary, except that now there will be less need to pretend they are bac by an independent controller in the person of the Director of the Budget. W then, will exercise budget control if the former controllers will not? Who w look beneath the totals to get at the composition of expenditures?

The change from interest in composition of expenditures to concern about the total amount has alleviated certain anxieties and created others. Now the need be less conflict between spending ministries and hteir financial controlle over who has the expertise in a particular line of policy. Having shored up the positions in this way, however, the central control organs began to worry abo whether they had gone too far, that is, whether anyone was paying attention public policy in the country except those who had an axe to grind because the were the ones doing the work. Conflict had been reduced, authority had bee maintained, but the rationality of expenditures was thrown into greater doub than ever.

Of course I exaggerate; budget controllers still exist and they do on occasion attempt to propel their own ideas about desirable public policy into interna ministerial allocations. But the essential rationale for this type of intervention has been seriously weakened by the growth of public sector activity. The Minis try of Finance can always justify imposition of a total on the grounds that it is the expert on managing the economy and knows how much expenditure it does or does not require. Its budget arm, however, in view of its small size and limited technical capacity, can hardly claim the necessary expertise. Traditional financial controllers, therefore, have become vulnerable to attack from outsiders who demand a share in budgetary decision making on the grounds that they do possess knowledge going to the merits of the issues. And, in their turn, the old boys have responded to these efforts to enter their preserve by sponsoring (or at least tolerating) their own reform under the rubric of program budgeting.

The attack from the outside came from the planners. They claimed a comprehensive rationality that would enable them to determine where the nation should go and how it should get there for years ahead, both in regard to economic policy and to the allocation of resources among sectors and within major projects. Finance ministries met the challenge by bolstering their capability in economic management and by mobilizing the political power to depart from the plans whenever they thought necessary, which was quite often. Whether it was because the planners in Japan did not anticipate so much economic growth or those in England so little or the ones in France missed on inflation, or because the central finance organs refused to implement the plans through the budget, in no case

B

ed targets met or even roughly approximated." What is important for
rposes, however, is not the fate of the plan but the nature of the
ween finance and planning in regard to expenditures.

are spenders. Their raison d'etre is economic growth. Typically they
ate expenditures and overestimate revenues in order to leave room
vestments they believe necessary to accomplish their major purpose.
re natural allies for large spending departments whose projects they
irable to secure economic growth. For the same reasons, planners are
1 enemies of financial controllers who wish to place limits on expendi-
o preserve their own authority in doing so by maintaining the right
allocations within the total. An all-too-brief recapitulation of experi-
eat Britain and Japan must suffice for illustrative purposes.
artment of Economic Affairs and the National Economic Development
hich preceded it in Britain, postulated high rates of economic growth.
there was no theoretical rationale for doing so, spending departments
allies found it convenient to assume that they could be allowed to grow
ate of increase in gross national product. Consequently, when economic
1 below the announced figures, the Treasury found expenditures rising
apidly than incomes with subsequent balance of payments and inflation-
ties. Nor did the Treasury just want a lower publicized rate of growth;
wanted no public rate of growth because it knew that politicians would
themselves to be accused of selling the country short, hence the figure
vitably be higher than experience justified. So it fought the DEA and
secured its abolition.

anese Bureau of the Budget was not at all happy with the publication
1 expenditures years into the future, for it knew that spending minis-
d use them as a floor from which they would try to get additional incre-
e Budget Bureau insisted on treating the plan figures as maximums
h they were free to cut according to the conditions of the time as they
1 them. Power and happiness, however, are not always equivalent.
vas not too difficult for the central finance organs to beat off attacks
outside, without much change in their own activities, they did find it
to accommodate their own internal dissatisfaction with the seeming
telligence devoted to budgetary review.

udget merely registers the constellation of prevailing political forces,
o need for any special technique. It is sufficient if each actor plays his
cole so that neither spenders nor guardians get their way entirely. But
ere a way, outside of (or apart from) the political process, for deter-
e merits of proposed expenditures, central controllers would have a new
They could better justify their own activity to themselves. This was
se of program budgeting. Instead of dealing with budgetary inputs, such
nel and maintenance, for purposes of narrow control, the budget would
zed around categories emphasizing outputs, so as to improve choice.
ling quantitative indicators of inputs and outputs, and ranking pro-
cording to merit in achieving these objectives, resource allocation would
à a more rational way.

ain difference between the old and the new forms of budgeting was
could be done and the other could not. The burden of calculation im-
PPBS was literally beyond comprehension. Failing to create this simul-
equation of society in the sky, understandably unable to come up with
connecting inputs to outputs across the widest range of public policies,
budgeting everywhere fell of its own weight. While it is also true that
re significant bureaucratic objections from those who felt they bene-
e under the old ways or who could not comprehend the new ones, and
objections from participants like American Congressmen who looked
■m budgeting as a sort of mumbo-jumbo designed to exclude them, no-
as an ideal (or even a better) budget created. It was one thing to talk
cusing on future results, considering a wider range of alternative, and
getting better information, and something else to actually perform
e could remotely specify the required cognitive operations.
ephen Cohen, Modern Capitalist Planning: The French Model (Cambridge: Har-
ersity Press, 1969) and Aaron Wildavsky, "Does Planning Work?", The Public
24 (Summer, 1971), pp. 95-104, for France. See Isamu Miyazaki, "Economic
n Post-War Japan", The Journal of the Institute of Developing Economics VIII :4
-, 1970) and Wildavsky, "If Planning is Everything, Maybe It's Nothing", Policy
forthcoming, 1973) for Japan. See Heclo and Wildavsky, The Private Government
Money and the citations therein for Britain.

"

ron Wildavsky, "Rescuing Policy Analysis from PPBS," Public Administration XIX:2 (March/April, 1969).

From the outset, the central budget agencies were ambivalent, divided tionally between more traditional types and program budgeters who, thems were unsure of how far they could or should go. In Japan, for instance, pro budgeting had to fit in with pervasive norms mandating equity among ticipants. If program budgeting could be somehow shown to be more scie so that its results were both more desirable and more acceptable, that v take care of the problem. But the sponsors had the usual trouble figurin what to do and the usual difficulty in getting anyone who was disadvantag agree that what they did was good.

The federal government of the United States, which had first introduced gram budgeting, was also the first to abandon it. Being initially the ambitious, it experienced the most extreme difficulties in negotiating the I ways between knowledge and power. The Office of Management and Budget neither able to produce information from program categories appropriat the level of the user nor to impose such preferences as it did have upon recalcitrant and unwilling. The spending agencies played along because thought PPBS was for the OMB, whose favorable opinion was useful to th When they discovered that the OMB was as much in the dark about what do as they were, and that program budgeting was supposed to be for them, t used the program structures as the latest in the long series of devices for s advertisement in the struggle for funds. It was a good two years after PP was disestablished, in the midst of a circular commenting on the reduction unnecessary paperwork in the federal government, before the OMB finally beg to receive inquiries from people who noticed something was missing. Inform tion on the French Rationalizacion du Choir Budgetaires is sparse, but c versations with a few people who are in a position to know suggest that is still an ephemeral part of French budgeting.

Great Britain has been saved for last because the evolution of its thought a practice is most instructive. When all of the publicity surrounding progra budgeting reached Britain, the Public Expenditure Group in the Treasury w well served by its customary caution. Observers, sent to America, came ba enticed but wary of the innumerable pitfalls they faced. They proceeded pra matically in two directions: they encouraged a few departments with intereste personnel to try out the creation of program structures and categories, an they gave part of the time of one of their bright young men to work with loca government officials in actually introducing program budgeting at that level Let them struggle. Thus when neither group was able to advance because the could not figure out how to make the system work, the Treasury had no sub stantial investment to abandon.

Yet the Treasury could not sit still. Its own concern about the adequacy of program review was exceeded by that of the Conservative Party and its associated businessmen who were concerned to give the Prime Minister and Cabinet more adequate tools for making expenditure decisions. Instead of rushing into program budgeting across-the-board, however, the Treasury persuaded the businessmen to join with them in creating a mechanism for Program Analysis and Review (PAR) that was aimed at encouraging departments to analyze a few major programs each year with the knowledge that they will be subject to Cabinet review. PAR is still in its infancy. It has hardly had time to make an impact. But it exists. Its viability may depend in part on whether another new apparatus. The Central Policy Review Staff. (CPRS), which the Conservative businessmen designed to aid the Prime Minister and Cabinet in reviewing policy and adhering to party programs, will also be able to maintain itself in the heretofore closed world of top level British political administration. Thus the Treasury now faces competition, not from planners with grandiose general schemes, but from analysts who are interested precisely in having an effect on the choice among major alternatives in the annual allocation of resources.

The very success of the Treasury in creating the PESC procedure for controlling expenditures has enabled others to concentrate on improving programmatic choices. In the Public Expenditure Survey Committee the British have the only major innovation in budgeting practice that has proven successful among the rich nations of the world. PESC does join knowledge and power. The Treasury does issue explicit instructions that departments can follow. Between the carrot of securing greater safety for vulnerable activities and the stick of its existing powers over expenditure, the Treasury has energetically sought and received substantial departmental cooperation. Departments and the Treasury

See Jeanne Nienaber and Aaron Wildavsky, Budgeting and Evaluation in Recreation Policy, or Money Doesn't Grow on Trees (New York: Basic Books, 1973, forthcoming).

agree to disagree, but they produce a five-year projection of existtures that has become the actual mechanism for allocating resources government. The lesson is not that successful budgetary innovation ms, but that it is based on making calculations manageable. The least effort in terms of theory and data has had the most practical effect. he political question is asked-what is in it for us-all participants. nd department alike, can find affirmative answers.

ig recommendations for Congress I shall use lessons learned from rative survey of budgetary practices. But Congress is not like the s in these other countries. Congress is still powerful. Legislatures ly exercise power independent of the Executive over a wide range s have becofe a rarity in the contemporary world. Members of the Diet act on the budget not as members of the legislature, or even as of the party in parliament, but by virtue of their association with arliamentary Liberal Democratic Party. Members of the House of have little influence over expenditures in any capacity and most of do have comes from belonging to the majority party. Deputies and n France may trade on their extra-parliamentary positions as civil r mayors to importune for favors before the executive budget is made urse of action required to convince the Executive to accede on one another is so involved that few try and less succeed. It is important, in appraising recommendations for change, we take account of special characteristics.

r choice under congressional conditions

orm proposals failed because they gave inadequate consideration to ites for decision under Congressional conditions. Whatever is recomust be applicable to a large, heterogeneous, and independent body berating within a political system where power is fragmented and There is no sense in treating Congress as if it were made up of (or y) a small, cohesive group of men easily able to agree or to impose on others. To avoid these traps it will be helpful if we state the stics of the Congressional environment in the form of criteria for ong alternative budgetary procedures.

y committees, not one. If you have a committee which makes a budget not altered by others, a committee without rivals, a steering committee islature, that is called a cabinet on the British model. Congress is a titution; it has many centers of power depending on the time and To be effective, Congrss has to make use of its dispersed, fragmented, 1 nature, and not fight against it. A good way to use pluralism is to pecialization.

e specialization, not less. Without specialization, there is no knowledge, ut knowledge there is no power. When executives wish to emasculate es, they break up existing committees and prevent formation of new alized enough to look into specific areas of policy and small enough n dispatch. It is no good railing at last year's failures or the previous lisasters. Unless the legislature can keep up with the Executive, it e an auditor after the fact rather than the budgeter ahead of the italize old committees rather than create new ones. The budget must every year. There is no time for a new committee to learn; it must ell the first time or it (worse still, its activity) will be held in disrecommittees have the virtue of age; they have long been there and expect them to be around. Because they perform other essential appropriations and finance) they can survive initial disappointment to set a total or having it exceeded. Since they cannot compel other to follow their lead, they also need the advantages of existing power ting the matter another way, it is too much to ask a new committee a new function; in this case new wine will age better in old bottles. d committees be willing to learn new budgetary tricks? That depends le to answer the always relevant political question of incentives. ething for all, not nothing for most. If budgeting is a game, all players ve a chance to score. There must be something in it for all-bureauutives, central control organs, legislators (if relevant)-or they will lay seriously. Ministries could not enhance their control over totals, e, unless they were prepared to concede enhanced discretion over within those amounts to spending departments and their ministers.