The University of Michigan

Anyone who develops more than a passing acquaintance with Swedish politics is bound to be impressed with the Swedish style of policy-making. For Americans especially, the absence of crises that produce hastily-considered decisions and the lack of many of the crude forms of obstruction - withholding of information, clockstopping to sidestep legal deadlines, or filibusters, for example - are especially striking. Unfortunately, apart from Dankwart Rustow's excellent but limited study,1 we still lack an adequate descriptive statement of policy-making in Sweden. Nor do we have an adequate explanation for the existence of this important but little-understood pattern of behavior. My purpose here is to attempt a better description of Swedish public policy-making and to begin to sketch out some relationships between political culture and patterns of policy determination that may help to account for such patterns. Since these terms are hardly unambiguous, let me begin by indicating how I intend to use them.

Policy development is often studied within the analytic framework of decisionmaking, a framework that typically focuses on a choice, made by a political actor or actors, from among alternatives presented to him or them in some specified situation. Actions taken by the actors leading to the choice are described in fulsome detail, and the choice is typically "explained" by reference to the goals of the various actors.² Because the major elements of this analytic scheme - actor, situation, goal, decision — can be, and usually are, given a variety of meanings by different scholars, case studies of decision-making are frequently criticized for their failure to provide cumulative evidence to support generalization about decision-making behavior. Defenders of such case studies, on the other hand, respond by arguing the merit of providing information that can impart a sense of process — how events move from one point in time to another - and a sense of realism to students of politics.3 Now these seem to me to be worthy purposes, but the confusions built into the above analytic scheme make it virtually impossible to achieve either realism or a sense of the political process.

Consider, first, the concept of "decision" as the thing or event to be explained.
Focusing on such an event assumes that something called "choice" takes place at a
single point in time and that the choice (or "issue" or "problem") can be analyzed

as something separable from both its social context and other similar actions being taken by actors at the same time. One might argue that such assumptions are "useful" from an analytic point of view, but one can hardly claim that they are realistic. For if the best decision-making studies have demonstrated anything, it is that major public policies more often than not result from an accumulation of past commitments, rather than a "choice" made at time x, and that decision-makers are always influenced not only by their past commitments but also by the network of contextual relationships they maintain with other actors and the requirements of other kinds of decisions in which they participate.4 The artificiality of slicing one "decision" or "issue area" out from its contextual setting is best revealed by the frequency with which analysts are forced to bring such factors back into their analyses, in a more or less ad hoc way, in order to increase the power of their "explanations." And if such factors are important enough to bring back into the analysis, we may well question the utility of a scheme which deals with them in an ad hoc, rather than systematic, fashion.

Or consider the conceptual and empirical difficulties created by the "goal" concept. If an actor seeks and achieves some goal in situation A, does the goal then disappear? If it does, then how do we account for the same actor's pursuit of a similar goal in situation B or G? If the answer is that it is the same goal being pursued again in a new situation, the implication is that the goal was not previously achieved, or that the passage of time and the change of situation makes no difference. This, of course, is how the question is usually answered, but it only makes sense if one also assumes that situations can remain the same from one time to another and that people do not change over time - neither of which assumptions strike me as being particularly "realistic." If, on the other hand, the goal does not disappear, but remains to motivate the actor, then the term "goal" seems inadequate to describe something which continues to exist over time, whether or not it is achieved.s If a "decision" is viewed as an event that can be precisely located at one, and only one, point in time, then perhaps the "goal" concept will do. But if, as I have already suggested, decision-making is more complex and less neat than that, then we clearly need another concept which can take account of striving for goals, but striving for them through time.

This need becomes particularly acute if we consider once again, the conclusions reached by some of our better studies of decision-making. One important conclusion is that actors who become involved in making policy frequently have no specific goals with regard to a given issue or problem.6 A change in administrations, an alteration in the external political environment, a new proposal advanced by a powerful group, and a hundred other kinds of changes constantly produce new situations, requiring a public decision of same kind, but for which there are no decisionmaking precedents. In such situations actor's are forced to develop goals for the first time and, apart from the unusual cases in which a single actor can act without regard to what other actors want to do, these goals must be developed in negotiations with other involved actors, none of whom may have thought about the problem before, and all of whom may have different perspectives on what should be done. This,

of course, suggests a second important conclusion, namely that goals change. Particularly in highly organized and bureaucratized societies, in which bargaining and negotiation among political elites is the typical source of public policies, goals emerge as the product of group interaction, in which initial perspectives and disagreements over policy are successively modified until one course of action is produced that is acceptable to all parties.7 Such products have distinct and discoverable patterns, to be sure, but their sources are to be found at least as much in social interactions as in rationalized - and individualized - statements about goals.

All this suggests that studies which account for "decisions" in terms of actors pursuing goals to solve identifiable problems, over a neat time-span that has a beginning, middle and end (i.e., the decision), are guilty of imposing intellectuality on a process that may have little or nothing to do with intellect. This "intellectuality" involves not only the notion of actors "solving problems" through "goal seeking," but also the notion that the problems "solved" by a "decision" are solved once and for all. Raymond Bauer has recently pointed out that concepts such as decision making ". . . imply a discrete sequence of events with a single definite point of termination. Serious policy making," he continues, "does not involve such decisive resolutions of a problem. A Supreme Court ruling of 1954 did not abolish segregation in public education, nor did the Civil Rights Act of 196!) eliminate discrimination in politics. Each of these events redefined the terms in which an ongoing struggle was conducted. The experienced policy maker knows that as he resolves one issue he is posing others. He realizes that he is frequently not settling an issue but redefining the rules of the game and if, in any meaningful sense, he has been "victorious" he hedges his victory to give himself room for maneuver in the future."⁸ To suggest that case studies of decision making, which seldom reflect these kinds of subtleties, offer an adequate or "realistic" view of process is at best unconvincing and, at worst, confusing.

Quite apart from the issue of whether case studies of decision-making ever "add up" to anything significant then, there are grounds to believe that such studies are weakest precisely where their greatest strength is often said to lie. The difficulty, I think, is traceable to inherent weaknesses in the analytic scheme noted above. If I may recapitulate, those weaknesses consist of:

1) a concept of "decision" or "choice" which artificially focuses attention on an event
that is presumed to take place at a single point in time, without reference to the
context of other events taking place at the same time;
2) a concept of "goal" that is conceptually confused and contradicted by much empirical

evidence;

3) an over intellectualized emphasis on individuals rationally pursuing ends in a process
that appears to have a logically neat beginning, middle and end;
4) a failure to take systematic account of the social processes which condition intellectual

processes; and

5) a failure to give adequate attention to the time dimension — both before and after a
given sequence of decisional activities — within which public policies are worked out.

These difficulties, fortunately, have come to be widely recognized and efforts to
overcome them now appear to be converging on an alternative analytic scheme

capable of handling the complexities of public policy making. To my own effort to develop and apply an alternative scheme several years ago9 we may now add some excellent analytic work, particularly Sir Geoffrey Vickers' The Art of Judgment^ and the recent volume edited by Raymond A. Bauer and Kenneth J. Gergen entitled The Study of Policy Formation.^-¹

If one adopts the view that public policy-making always involves an intellectual process operating within a social process, and that the individuals involved in these twin and inseparable processes are not accidental participants, but persons who maintain relationships with others through time, then the focus of concern shifts from the single decision (whatever that is) to the structure of relationships between participants and the norms which serve to maintain or change those relationships through time. The focus shifts, in other words, from decisions, to systems of decisionmaking. Such systems may be structured as organizations or they may be structured around policy areas, but in either case they are presumed to persist through time in the form of repetitive patterns of behavior. What motivates persistent behavioral patterns it not goals but the relationships such systems seek to maintain with their environments through time. These relationships are expressed positively as norms (which are pursued) or negatively as limits (which are avoided), both of which are built into the expectations that define the roles which constitute the system. Maintaining such relationships is accomplished through the social process of communication, using information that generates "mis-match" signals which reveal divergence from some norm or the approach of some limit. When such signals occur, they produce a response from the battery of responses available to role incumbents ir? the system. Instead of focusing our attention on actors making a decision by pursuing a goal, this scheme asks us to view decision-making as a dynamic process of interaction between a system and its environment, in which the norms and limits governing that interaction are maintained by individuals playing relatively stable roles, using information from the environment and other participants to produce the never-ending series of adjustments necessary to maintain system-environment relationships.12

By stressing the continuous nature of decision-making this scheme avoids the artificiality of the "single choice, single time" study; by emphasizing relationships expressed as norms and limits instead of goals, and by focusing on relatively stable role expectations instead of the heroic actor-individual, it underlines the structural determinants of behavior without denying the rational calculations of individuals in structured situations; and by insisting on the significance of information — communication, it offers a systematic account of the sources of stability and change in systemenvironment relationships. A systems-communication approach, in short, offers the possibility of accounting for decision-making (and decisions) in terms that can have generalized relevance and that can deal with system change through time.

Use of the "role" concept is, of course, a major source of potential generalizations, since this concept suggests relatively stable sets of expectations that are learned by individuals as they move into and out of various social institutions.l3 Description of any given set of decision-making roles thus opens up a whole set of questions whose answers can relate one system to a variety of others: what sorts of individuals are

recruited, from what social strata, for a particular decision-making system? How does recruitment take place? How are new roles learned, etc.¹⁴ Since expectations are mental constructs, they can also be related to other sets of mental constructs, at various levels of generality. To illustrate, I would like to offer a generalized description of decision-making in Swedish politics and relate that pattern of behavior to two hypothesized sets of expectations, one of which can be thought of as the elite political culture, the other of which can be thought of as the popular, or citizen, political culture.

My description is built upon a series of decision-making case histories, some of which I have done myself, and refers to actual patterns of behavior.¹⁵ Though the policy areas are quite varied, remarkable similarities of style and process seem to me to justify an attempt to synthesize the patterns that seem to recur again and again. The popular and elite expectations toward government — what it is, how it should operate and how it should be judged - refer to mental constructs rather than behavior, and they are hypothesized. That is, apart from my own rather limited interview experience, newspaper reports and political histories, there is no good evidence to show that the constructs I will be talking about are in fact held by Swedish citizens and officials. By stating them in a fashion that makes them testable, however, the groundwork can be laid for future work that might add enormously to what we know about Swedish political life.

The Swedish Policy Making Style

Policy making in Sweden is accomplished by highly-specialized roles attached to a dual structure of societal power. On one side there is the governmental structure which, in a small nation, with a long tradition of governmental centralization, is highly integrated around a strong central government.¹⁶ At the apex of this structure stands the cabinet, made up of elected political officials, each of whom bears responsibility for policy-making — but not administration — in some major area of public responsibility.17 On the other side stands what may be the best-organized structure of interest groups to be found in any nation of the world. Virtually all social interests of any significance — from industrial workers to tennis enthusiasts — are organized into local, regional, and national associations, and the most important interests (labor and industry) are further centralized by "super" organizations that represent the interest of all of their associated organizations in national negotiations.18 Because of the high degree of multi-level office holding in the governmental structure (e.g. 70 percent of Swedish Riksdag members hold local office), distinctions between "national," "country" or "local" politics have little practical significance. And because of the extraordinary overlap, at all levels, between governmental and nongovernmental positions (major labor, management and other interest group officials are elected members of the Riksdag), the distinction between "public" and "private" policy is all but meaningless.19

Any and all important social questions are automatically subject to determination by representatives of this twin policy-making structure, which in a real constitutes a cohesive national elite. Which representatives are involved in any given issue depends, first, on subject matter (conservation will be dealt with by conservation experts, incomes by labor and management experts, etc.) and second, on function, that is, new policies will be worked out by groups organized by national ministries, which bear special responsibilities for policy innovation, while interpretation of current policies is a job normally left to the specialized administrative boards. Specialization by subject matter means that "expert" roles are always involved, while specialization by function means that separate role structures exist for the development of new policies. One important consequence of the latter form of specialization is that major policy changes seem to be constantly taking place. At the present time, for example, the educational system is being completely changed,²⁰ the number of local governments is being reduced by a factor of ten,²¹ the Riksdag is being changed from a bicameral to a unicameral structure, a new metropolitan government for Stockholm is being created,²² and a massive new program of government investment in the northern half of Sweden is getting under way.

Activities of this kind are typically initiated by the official experts employed by either the government or the interest group organizations, reacting to information that reveals difficulties in realizing current policies or in the accomplishment of newly-enacted policies. Since Swedish public statistics are probably the most comprehensive and best-kept in the world, opportunities for such expert initiatives arise quite frequently. Once raised, the issue is further refined by the appropriate governmental ministry, with a view toward determining whether official action is necessary and if so, along what lines. Clearly, governmental review of these issues provides plentiful opportunity to avoid action, particularly if the issue is likely to be embarrassing. If action is taken it is in the form of a commission of experts, created to review the problem and propose alternative solutions for consideration by the government and the Riksdag. The fact that one hundred or more such commissions are likely to be in existence during any given year, each with a staff and funds, suggests that relatively few problems over which there is serious organizational concern are likely to be shoved under the rug of governmental inaction.23

Once appointed, the committee works without further governmental direction, though its members are certain to include governmental officials who are in a position to keep their colleagues or superiors informed. If committee membership and staff is not expert enough, outside experts are called upon to develop information. Such committees typically work for two to three years, though many work for as long as five years or more,²⁴ producing thick reports which outline their proposals and support those proposals with massive research documentation that is historical (how the problem developed and was dealt with in the past) as well as analytic. Completed reports are submitted to the appropriate national ministry, which immediately circulates copies to all parties and organizations that may have an interest in the matter - proposals to eliminate nine-tenths of all local government units, for example, were circulated to all local units, as well as country and national agencies

involved in the work of local governments.²⁵ Responses from interested parties are considered by the government, which then determines its position. If legislation is required a bill is drafted for consideration by the Riksdag, whose committees must report all bills assigned to them and whose two chambers must vote on all bills reported from committees.26 Action necessary to implement a legislative decision is typically programmed with similar care and attention to detail: the school reforms coming to fruition now were initiated in 1956; the two legislative acts reducing the number of local governments were put into effect over five and ten-years periods, respectively; the important revision of the Swedish pension system, decided in 1958, was programmed to come into effect in 1963, and so on.²⁷

Though vastly over-simplified, the process described here is common enough at all levels of Swedish government to be recognized as distinctive. Swedish policy-making is extraordinarily deliberative, involving long periods of time during which more or less constant attention is given to some problem by well trained specialists. It is rationalistic, in that great efforts are made to develop the fullest possible information about any given issue, including a thorough review of historical experiences as well as the range of alternatives suggested by scholars in an out of Sweden. It is open, in the sense that all interested parties are consulted before a decision is finally made. And it is consensual, in that decisions are seldom made without the agreement of virtually all parties to them. This desire for consensus sometimes means that issues over which there is significant conflict take a long time to resolve - as did the supplementary pension issue - but Swedish officials demonstrate again and again that there is no particular hurry — if a solution cannot be reached now, it may be possible next month, or next year. Given this emphasis on consensus, and since very little is done without thorough prior research anyway, "crises" seldom occur. Most crisis situations are anticipated, and where they are not, it is easy enough to stall until consensus is reached on what to do. The result is decision making which never seems rash, abrupt, irrational, or indeed, exciting.

Let me emphasize that systems of "expert" or "specialist" roles, operating in a deliberative, rational, open and consensual fashion, represent the "normal" in Swedish decision-making. Systems operating in this fashion are found at every level, and for every issue. Though I have used the national decisional process as my example, I might as easily have used examples from lower levels, where similar characteristics prevail and where, as a consequence, the amount of paper documenting and rationalizing policy choices is quite overwhelming. This pattern, in short, represents a national style, which is itself a reflection of Swedish political culture. To appreciate how this political culture structures decision-making, we need to look first at citizen attitudes toward government, and then at elite orientations.

Citizen Attitudes

Among a population of virtually complete literacy, in a country where the consumption of newsprint is quite high, and in which voter participation reaches extraordinary levels - in the recent parliamentary election ballots were cast by 89 percent of eligible voters²⁸ —a high level of citizen awareness and knowledge about government would be expected. I know of no studies which attempt to document this conclusion systematically, but scattered evidence from public opinion polls suggest that it is indeed true. A mid-1967 poll, for example, revealed that 90 percent of a national sample were able to correctly identify Education Minister Olof Palme's party allegiance, while 62 percent of the sample correctly identified Sven Wedén as the newlyselected leader of the Liberal Party (among their party colleagues, the figures for correct identification were 93 and 75 percent, respectively).29 Another poll conducted later in 1967 revealed that 92 percent of the respondents (again a national sample) had "heard about" the European Economic Community and that 59 percent could correctly identify what it was.30 In January of 1968, after several months of lively newspaepr debate over the activities of the Swedish security police, 83 percent had heard about this organization and, of these, 70 percent could correctly identify what it was.³¹ An earlier poll showed similarly high levels of public knowledge about a proposal for a new state investment bank, after a two-month campaign of propoganda organized by national authorities.³² "Knowing" and "understanding" are quite different, of course, but even with this caveat in mind, the available evidence suggests a high level of awareness of government activities among Swedes.

Public knowledge seems to be accompanied by a good deal of affection and support for the political system. Some of this is surely attributable to the King, an 88-year-old scholar and humanitarian, whose accomplishments and humility symbolize widely-held Swedish values. Even more, however, is due to historical events that have conspired to give Swedes a sense of the distinctiveness of their nation and a feeling of pride, bordering on superiority,, over national accomplishments. Though the Scandinavian countries are often treated as one homogeneous mass, the Swedes regard themselves as different from their neighbors - and for good reason. Unlike Norway or Finland, Sweden has been an independent nation for centuries and has never fallen under the control of an invading power as all of her neighbors have. Moreover, for the past 200 years at least, Sweden has not been plagued with the disputes over language that have upset — and continue to upset the internal politics of adjacent nations.33 Reinforced by war-time isolation and criticisms from neighboring nations unable to avoid Nazi force, and reinforced again by a post war policy of neutrality not shared by those nations, this sense of distinctiveness continues to be widely shared by Swedes. In addition, the outstanding success of postwar social and economic policies have transformed Sweden from one of the poorest to one of the wealthiest nations in Europe - all within the memory of a population that includes a large proportion of persons over 50.³⁴ Most Swedes, I believe, view their nation's distinctive history of national unity and social success as a source of great pride and strong positive attachment to their "peoples home."

Pride and support, however, seem to be focused largely on the outputs of the system, rather than inpust. An extraordinarily functional and service oriented people, the attachment of Swedes to political parties and interest organizations seems firmly rooted in instrumental calculations of interest, rather than democratic ideology or personal loyalty. It is interesting and possibly significant, for example, that no active Swedish political leader was among the top ten individuals recently named by Swedish young people in response to the question: "Which person in the world do you most admire?" Indeed, only seven votes of more than 2,000 cast went to Swedish political leaders - five to Finance Minister Gunnar Sträng, one to Education Minister Olof Palme, and one to Communist Party leader Hermansson.3s Perhaps more interesting is another poll which indicated that four-fifths or better of supporters of each major party (for the Socialists, 86 percent) would continue to support their party if present leaders left.36 It is difficult to imagine that a Robert Kennedy, Eugene McCarthy or George Wallace, with their emotional, vaguely ideological, but always personal followings would ever do very well in Sweden.

Swedish popular emphasis on what the government, parties and interest groups do for them, rather than they how do it, may help to account for the shifts in support given to the bourgeois parties, whose vote totals and legislative representation move up and down as they seek to find programs that can motivate voter and member loyalty.37 Such fluctuations in support, together with the continuous search for program improvement among government agencies, underline what appears to be the highly tentative and conditional nature of citizen involment in governmental support. There is something almost cold-blooded in this, as though citizen support could be withdrawn and would be withdrawn if the government were to stumble even momentarily. This is doubtless a foreigner's exaggeration, and yet the Swedes do not accept myths, are not flagwaving nationalists, and do not invest much emotion in their governmental activities. If service standards go down, what else is there to motivate support?

The question is of more than rhetorical interest in view of the somewhat limited notion of the responsibilities of citizenship held by most Swedes. Although voting turnout is amazingly high - encouraged by permanent registration and the use of Sunday as election day - other forms of civic participation are something less than popular. Apart from parties, citizen political organizations such as Leagues of Women Voters, neighborhood improvement groups, reform organizations, and the like, do not exist, largely, it seems to me, because Swedes do not expect to have anything to say about the operation of government agencies. Swedes elect their representatives to do the work of governing and to not regard it as their obligation to help that work along: government is something done by people who hold official positions and not done by citizens. The fact that government officials historically have enjoyed high status and that their work is highly specialized and thus conducted in the technical language of experts, encourages distant relationships between officials and citizens.3B Moreover, individual membership in party and interest organizations is a formality for most members, who seldom participate in work accomplished mostly by small leadership cadres.³⁹ Organizational control by small cadres is not

peculiar to Sweden, of course, but the attention usually given to the formally-inflated organizational membership lists suggest a beehive of individual action that is far from an accurate picture of Swedish organizational life. As far as interest, political and governmental organizations are concerned, "Sven Svensson" - the average Swede - expects the leaders to do the jobs for which they were selected and does not expect his responsibilities to extend beyond participation in the selection process.

These several characteristics of Swedish popular attitudes toward government - high awareness, positive feelings of pride and support for the system, emphasis on system outputs, but little interest in participation - suggest a variant of what Almond and Verba call "The Subject Political Culture." Within its encompassing framework, the political-governmental elite work in terms of expectations that reflect, as well as shape, public attitudes.

The Elite Culture

An obvious and important consequence of the subject citizen orientation is that it frees political decision-makers from a wide range of day-to-day citizen pressures that might be expected to occur in democratic systems. Citizens don't expect to participate in decision-making and officials don't expect them to try. Moreover, government employment historically has been an occupation of very high social status, reserved originally for the nobility and, more recently, for the minute proportion of the population holding university degrees — indeed, positions in the national administration have been monopolized for a long time by persons holding a certain kind of degree (law).⁴⁰ Expansion of both government employment and university enrollments in recent years may be reducing the selectivity of government positions, but the social status of such positions remains very close to the top of the Swedish status hierarchy. Coupled with freedom from public interference, this recruitment selectivity gives Swedish public officials a selfconsciousness and cohesion that is the mark of an elite.

For members of this elite, political and governmental activity is defined as work, rather than as game. Though salaries of government administrators are high, and a considerable number of politicians hold such jobs, there are few other rewards, material or symbolic, for public activity. Salaries for national legislators are low (about $ 5,000) and little supplemented, while per diem allowances for local legislators are virtually insignificant. Strict standards of official honesty, enforced by constitutional requirements that all official transactions be made available to the public immediately, reduce opportunities for graft almost to the vanishing point. Top political executives are well-paid (the prime minister earns $24,000 per year), but enjoy few other benefits: not until four years ago were limosines made available to Cabinet members (on a shared basis), nor is there a special full-time residence for the prime minister. Symbolic gratifications are equally scarce. Organizations, rather than individuals, are the bedrock political units in Sweden, and few politicians other than

party leaders and cabinet ministers are well-known to the public, which, as noted earlier, views them in less than heroic terms. And young politicians can confidently expect to wait a long time before achieving even that degree of notoriety,41 since political organizations are strictly hierarchical and not wont to change their leaders very often. Television may alter this situation, but so far the tube has not created any new political personalities. In Sweden, one does not go into politics to make money, gain fame, or for enjoyment. One enters politics to work at problems, either because of special competence, or special interest - preferably the former.

The "politics is work" orientation gives the Swedish political elite a strong action commitment. Government is established to do things, not talk about doing things or think about doing things. Thus political debate in Sweden is almost totally devoid of ideological or philosophic symbolism, focusing instead on specific proposals to deal with specific problems — usually economic in content. Should the children's allowance be raised to 1100 or 1200 crowns? Should rent controls be abolished for some or all of Stockholm's apartments? Should one or two subway lines be extended to the new city being built on the outskirts of Stockholm? Should x or y million crowns be invested in the north? In this context, the use of Royal Commissions, noted above, are the perfect expression of the work orientation. On the one hand, they are created to provide answers to a specific action question: what to do about x problem. And on the other hand, since there is "nothing to do" until those answers are provided, nothing is done - for periods that can run as long as five years, and sometimes more.⁴² By the same token, once the appropriate action has been determined, Swedish officials waste no further time debating the issue. Disagreements that remain are put aside in order to get on with implementation of the decision.

Obviously, the absence of citizen agitation over issues that are being studied makes it possible for decision-makers to defer action at very low cost themselves. But the Swedish political elite is also extraordinarily patient in terms of policy.43 The temporal dimensions within which they work are long term, stressing the continuity of any given problem over many years and implying its continued existence into the indefinite future. Within these dimensions, problems are not said to be solved once and for all; they are merely temporized until such time as they require attention again. Viewing themselves not as "decision-makers", but as participants in a long stream of historical events, Swedish politicians do not interpret "crises" in terms of immediacy. And because actions they take will be credited more to their organizations than to them as individuals, incentives to "make a record" on which to launch a political career are minimized. Indeed, one of the surest methods of injuring a political career is to offer proposals that have not been carefully thought through, researched, and discussed.

But if benefits arising from policy making are attributed to collectivities (such as parties or governmental departments), harmful consequences are frequently attributed to individual officials. Swedish law spells out official responsibilities in great and minute detail. Furthermore, officials who act beyond their authority or whose actions reveal failure to use their authority properly are individually responsible for whatever evils result - with penalties that include removal from office.⁴⁴ Officials are therefore

encouraged to search out the implications of proposed action for a long time, and to consult widely, if only to ensure that the proposed policy does not jeopardize their own position. Patience, careful research and consultation, therefore, are not merely intellectual traits. They are socially useful mechanisms of self-preservation.

I have already suggested that the lengthy search for alternatives and widespread
consultation normally results in agreement among the officials involved. Two habits
of mind, in particular, seem to me to encourage official agreement. One is relativism

- a characteristic trait of officials for whom there are few, if any, absolutevalues or principles. Principles are adhered to strongly in Sweden, but no principle is beyond the reach of an argument that demonstrates the social utility of abandoing that principle in favor of some other. But, among an action-oriented elite, very few issues ever get defined in terms of "principle", for a second habit of mind discourages such definitions. That habit is particularism: the definition of issues in terms of their specific, rather than generalized significance. Again and again, issues which are raised in terms of one principle or another are gradually resolved into increasingly specific problems to be solved, thus avoiding the disagreeable task of reconciling divergent opinions.₄₅ If such a divergence occurs, the normal strategy is simply to postpone action until some specific program is found that all can accept. There are, of course, rare cases when neither particularization nor postponement avoids a conflict over principle. Perhaps the outstanding recent case of this kind was the government's passage (in 1958) of a supplementary pension plan that was compulsory and publiclycontrolled. Though conservatives fought hard against this plan, and even attempted to repeal it after it was passed, they very quickly came to support it once it was clear that theirs was a minority position.46 Where particularization fails to work, in other words, relativism operates to reduce or eliminate the distance between opposing positions.

What all this suggests, I hope, is an elite culture in which a highly pragmatic intellectual style, oriented toward the discovery of workable solutions to specific problems, structures a consensual approach to policy making. This approach is in part a function of intellectual habits that minimize opportunities for conflict, in part a function of the desire for administrative self preservation, and in part a function of the peculiar Swedish distaste for open conflict. Viewing politics as work, and following practices that encourage widespread consultation in decision-making, it is natural for the Swedish elite to think of politics in terms of cooperation rather than conflict. And it is functional for them — given individualized legal responsibility - to avoid conflict situations if they can, to deny that conflict exists (even when it is evident), and above all, to avoid publicizing the conflict that occasionally breaks out within the confines of the elite structure. To publicize serious conflict, after all, is to risk intervention by institutions such as the Ombudsman, with potentially disastrous consequences for officials who have acted improperly. But all official actions must be publicized once they are taken, and as a consequence, they are not taken until agreemnt has been reached informally. Once a decision is made it is presented as a unanimous product, with extreme care taken to avoid identification of the "most" or "least" influential participants. "We had a meeting and came to an agreement"

is about as much as outsiders ever learn from participants in Swedish decisionmaking. Such statements., extraordinarily frustrating for scholars,⁴⁷ say little about the presence of conflict, but a great deal about Swedish elite distaste for open disagreement.

Conclusion: The Impact of Culture on Policy-Making

The range of hypotheses advanced here concerning elite and popular attitudes toward government in Sweden offers a way of understanding policy making that moves beyond the individual case, with all of its built-in intellectual and social artificiality, and toward a more general level of understanding. Individual case studies have value only to the extent that we know how they are related to, or illustrate, more general patterns of behavior. And the value of knowing something about observable patterns of behavior is enhanced if we can relate those patterns to the belief systems in whose terms they are structured. Thus the elite expectation of freedom from citizen pressure, its work orientation and action focus can help to explain the deliberative, rationalistic and consultative pattern of decision-making in Sweden. Similarly, the consensual aspects of that pattern become understandable in the light of elite intellectual traits and social relationships which encourage unanimity among decision makers. Such explanations, morover, are not independent of other aspects of social structure. Indeed, they lead directly to questions about recruitment, communication patterns and institutional relationships that are largely hidden from view in the case study approach. Emphasizing systems of decision-making, composed of roles, structured in terms of cultural patterns of belief and evaluation, thus moves the study of policy making away from its pre-occupation with the unique, and toward a level of analysis that fits more comfortably within the province of social science.

---

NOTES

1 Dankwart The PoliticsPrinceton: Princeton1955.

2 Among the many studies of this kind one of the best, in my view, is Edward Banfield's Political Influence. Glencoe, Illinois: The Free Press, 1961.

3 See, for a sophisticated and still-relevant defense of case studies, Harold Stein (ed.), Public Administration and Policy Development: A Case Book. New York: Harcourt, Brace and World, Inc., 1952, pp. xx-xxx. Stein's argument is distinguished by his recognition of many of the problems to be discussed here, though his conclusions are rather different from my own.

4 Perhaps the best expression of this line of argumentation is Charles E. Lindblom, The Intelligence of Democracy. New York: The Free Press, 1965.

5 This point is discussed in more detail in Sir Geoffrey Vickers' excellent study, The Art of Judgment. New York: Basic Books, Inc., 1965, pp. 31—35.

6 Robert F. Kennedy's account of the Cuban missile crisis makes this point in particularly dramatic terms, as does Glenn D. Paige's study of the American decision to intervene in Korea. See Kennedy, Thirteen Days. New York: W. W. Norton, 1969, and Glenn D. Paige, The Korean Decision. New York: The Free Press, 1968.

7 For a study which makes this point in a Swedish setting, see my "Politics and Planning in a Swedish Suburb," Journal of the American Institute of Planners (July, 1969).

---

8 Raymond A. Bauer, "The Study of Policy Formation: An Introduction," in Raymond A. Bauer and Kenneth J. Gergen (eds.), The Study of Policy Formation. New York: The Free Press, 1968, p. 18.

9 Thomas J. Anton, The Politics of State Expenditure in Illinois. Urbana and London: The University of Illinois Press, 1966.

10 See note 5.

11 See note 8.

12 This formulation is heavily indebted to Sir Geoffrey Vickers' previously noted work.

13 Use of the role concept to structure investigation of decision-making behavior is illustrated in my work, cited in note 9 above. See also N. Gross, W. S. Mason and A. McEachern, Explorations in Role Analysis. New York: John Wiley, 1958.

14 An attempt to relate recruitment patterns to the Illinois system of financial decisionmaking is provided in Thomas J. Anton and Joseph P. Pisciotte, "The Education of Illinois State Officials," Illinois Government (Urbana: The Institute of Government and Public Affairs, September, 1967).

15 My own work includes "Politics and Planning in a Swedish Suburb," loc. cit., and "Incrementalism In Utopia: The Political Integration of Metropolitan Stockholm," Urban Affairs Quarterly (forthcoming, September, 1969). In addition, Donald R. Niemi's "Sweden's Municipal Consolidation Reforms" (unpublished Ph. D. Dissertation, University of Chicago, 1966) offers a wealth of information about several related policy-making sequences, as does Björn Molin's Tjänstepensionsfrågan. Göteborg: Akademiförlaget, 1965.

16 An excellent brief discussion of Swedish government may be found in Nils Stjernquist, "Sweden: Stability or Deadlock?", in Robert A. Dahl (ed.), Political Oppositions in Western Democracies. New Haven and London: Yale University Press, 1966, pp. 116-146.

17 The distinction between the policy planning function of cabinet ministries and the routine administration function of administrative boards is made clear by Pierre Vinde, The Swedish Civil Service. Stockholm: Ministry of Finance, 1967, pp. 7—18.

18 See Nils Elvander, Intresseorganisationer i Dagens Sverige (Lund, 1966), for a study of Swedish organizational politics.

19 See Nils Andrén, Modern Swedish Government. Stockholm: Almquist and Wiksell, 1961, pp. 56—57 for figures on the representation of local officials and interest groups in the Riksdag.

20 See The New School in Sweden. Stockholm: National Board of Education Series 65, 1963.

21 Niemi, op. cit.

22 Anton, op. cit.

23 Statistics on the number of commissions appointed annually, membership and length of work period can be found in Hans Meijer's excellent study, Kommitté politik och Kommittéarbete (Lund, 1956), pp. 17-27. Meijer's study covers the period 1905-1954.

24 Ibid., p. 27.

25 cit.,162.

26) Andrén, op. cit., pp. 73, 86

27 See note 15, above.

28 The exact percentage was 88.7. Final election returns, including absentee ballots, were reported in Dagens Nyheter, September 29, 1968, p. 3.

29 SIFO. Indikator. Tune. 1967.

30 SIFO. Indikator. Seotember. 1967.

31 SIFO, Indikator, January, 1968.

32 SIFO, Indikator, September, 1967.

33 Finland's recent dispute over the elimination of Swedish as the official "second language" is a case in point.

34 For English readers, a useful discussion of Swedish population and social characteristics may be found in Anna-Lisa Kalvesten, The Social Structure of Sweden. Stockholm: The Swedish Institute, 1965.

35 Dåsens Nyheter, April 14, 1968, p. 6.

36 SIFO, Indikator, September, 1967.

37 A good analysis of fluctuation in bourgeois party vote totals is Bo Särlvik, "Political Stability and Change in the Swedish Electorate," Scandanavian Political Studies, Vol. I (1966). pp. 188-222.

38 Nils Herlitz. Sweden: A Modern Democracy on Ancient Foundations. Minneaoolis: TV« University of Minnesota Press, 1939, remains useful as an introduction to Swedish bureaucracy against a background of constitutional development.

---

39 Frederic Fleisher, an intelligent observer of modern Sweden, has written: "Even though Swedes often refer to their country as 'Organization Sweden,' most of them have only a hazy picture of the vital role of organizations ... Most citizens do not really knowhow to approach the organizations to which they belong." See his study, The New Sweden. New York: David McKay Co., Inc., 1967, p. 41.

40 Herlitz, op. cit., p. 60, writes: "In the central administration and in the [county] governors' departments most high offices are occupied by lawyers who have behind them the same university studies as judges and barristers." Work now under way in the Stockholm University Department of Political Science is attempting to document this conclusion and update it in terms of more recent patterns of recruitment.

41 There are, of course, significant exceptions to this conclusion, notably Education Minister Olof Palme, who is in his early forties, or 31-year old Per Ahlmark, a prominent member of the Liberal Party.

42 Meijer, op. cit., p. 27.

43 This paragraph draws heavily on my previously-cited work "Incrementalism in Utopia . . „" loc. cit. See note 15, above.

44 Vinde, op. cit., p. 16.

45 I have emphasized this point in my review of the development of metropolitan governmental agencies for greater Stockholm. See note 15, above.

46 Molin, op. cit., pp. 120-121.

47 Andrew Schonfield, a student of Swedish labor market policy., has nicely captured the essence of this frustration: "I recall a British trade union leader after an organized visit to Sweden - there were several such visits undertaken by the British trade union movement in the early 1960's in an attempt to discover the secret of Swedish labor'« success - expressing his frustration over the whole business. The secret was either too banal or too opaque to yield intelligent investigation. 'All they can tell you when you ask them how they do it," he said, describing some particularly difficult decision which involved the concerted action of competing interest groups, is: '"We has a meeting." We has a meeting! I'd like to see how they'd make out with our blokes over here.'" See Andrew Schonfield, Modern Capitalism. New York and London: Oxford University Press, 1965, p. 199.