For five years the University Departments of Political Science in Sweden have
channeled a large part of the:.r resources into a research program concerned with
local self-government in its broadest sense.

Why a Research Program?

If one is convinced that the social sciences can contribute toward society, then the traditional lack of research planning appears highly unsatisfactory. According to an old humanistic research tradition, each person engaged in research decides for himself what he will work on. Hence, personal and temporary circumstances determine the choice of research projects;. Usually the result is that each person involved in research works indepently on one or two pieces of the large puzzle called society. No one is responsible for completing the puzzle or even assuring that the investigators are working on Ihe same kind of puzzle.

Many, no doubt, cater to the idea that even in the separate special studies, one is in search of the philosophers' stone — i.e., general theories about politics — and that no one can know in advance where the stone is to be found. This appears to be a faulty way of looking at the matter. Even if one hopes that political research will lead to theories which are valid for many cultures and epochs, one must realize from the outset that a vast part of the research is of limited interest as regards time and place. The rapid changes in society make inadequate every analogy with permanent laws of nature which lie waiting to be discovered.

A more reasonable view is that it is incumbent upon each generation of social scientists to make the greatest possible contribution using the intellectual and material resources available. Such a contribution cannot be achieved without a significant degree of co-ordination, though this is not to say that every researcher makes his most worthwhile contribution working in the context of a large, common research project.

If a particular research problem is deemed urgent, it ought to be solved within the time given and with the resources at hand. The results should be evaluated according to resources available and ways in which they are exploited, not on the basis of unlimited time aad an ideal norm.

Why Research on Local Self-Government?

Two main reasons motivated the choice of this research problem. One was that the question is of current and practical relevance for Swedish society since the conditions for local self-government are changing radically. Until the beginning of the 1950'5, there were ebout 2500 local government districts, communes (kommuner), in Sweden. Well ever 100 of those were urban communes; most of the small rural communes coincided with the local parishes of the Swedish state church. As the number of communal tasks increased — largely due to the increasing responsibilities turned over by the state to the communes, it became impossible for the small communes to perform satisfactorily. Through a decision of the Riksdag (the Swedish national legislature), smaller rural communes were combined into larger units, and in 1952 the total number of communes was reduced to less than 1000. Yet the communal tasks have kept growing. Since 1955 the share of local government consumption and investment in gross national product is greater than that of the central government, and a new reform has

been decided upon aimed at successively decreasing the number of communes
to just under 300. Within a period of twenty years, the number of communes will
therefore be reduced to about one tenth of the original number.

The other main reason for choosing this particular research problem is its possibilities to develop methods and theories. The large number of communes with varying characteristics, though functioning within a common institutional framework, provide good prospects for isolating the effect of single variables, and offer a kind of pseudo-experimental attempt: the existing communes can be viewed as the results of an experiment in which a set of independent variables have been changed one by one.

In addition, such a project is ideal for co-operation between different universities. Reasons exist for paying attention to the geographic distribution of the communes, and, furthermore, each university has special ties with its region. Besides, local self-government is hardly controversial: it is generally spoken of with deep respect by politicians.

The Sample of Communes

Some of the investigations can be based directly on the extensive statistics already available, thereby including all the communes. In other instances, such thorough examinations will be necessary that the material must be presented as case studies. Generally however, the investigations are based on a sample of communes.

The following sample procedure was used. To begin with, Sweden's communes were divided into three separate groups. In the first group were communes which are part of the three metropolitan areas (population M); in the second, communes which are undergoing changes — that is, communes which were affected, wholly or partially, by communal mergers 1/1 1967 (population C); and in the third group, the remaining communes — that is, the majority of the communes (population A).

From population A, 36 communes were drawn using a modified "greccolatinsquare-design." As basic selection variables, size and density of population were chosen (see below). Furthermore, the political situation also was considered (communes where the Social Democrates had/did not have a majority according to the 1964 election results), as well as geographical position and location within the block of communes. From the figures below, the commune sample construction appears in relation to the two basic variables. The numbers within parenthesis tell how many communes there actually are within the combined size-density of population classes. As is apparent, six communes have been taken from each such class whenever the real numbers permitted.

The construction of the sample will make it possible to analyze the correlations between political and administrative variables connected with communal selfgovernment and the sample variables with the ultimate aim of treating questions of causality. The sample is not, however, suitable for a description of the Swedish communes. One cannot, for example, state with satisfactory precision how common a particular characteristic is in all of the Swedish communes.

The three largest cities — Stockholm, Göteborg, and Malmö — have been added
to the A-sample.

Six communes have been taken from the C-sample.

All of the selected communes, 30 +3+6, in accordance with their geographical position, have been distributed among the Political Science departments of the five universities, each of which is responsible for contact with a number of these communes.

The Research Program

For purposes of presentation, the research program may be divided into three
parts:

1. The Citizens' Influence in Local Government. This includes various investigations of communal elections, the activities of communal assemblies, the participation of elected representatives in communal administration, the position of communal officials, etc.

2. The State and Local Self-Government. This includes studies of the state's general
direction and supervision of kommunal activity, the actual degree of self-government
in some major areas of communal activity, etc.

3. Communal Units in Relation to Communal Functions. The purpose is to study the relation between function and area (or population) and to examine the question of the proper size and type of commune for the performance of different communal functions.

The investigations were initiated in full scale at the turn of the year, 1965/66. Since
communal elections were held in 1966 and the next elections will not occur again
until 1970, the resources had to be concentrated mainly on election studies.

Among the investigations already initiated, the material of which is bein analyzed
at present, we mention the folk wing:

1. A study basedon interviews with voters: a) The main study embraces 50 personal interviews in each of the 36 communes of sample A and 200 personal interviews in each of the largest cities, making a total of 2,400 interviews, b) The study of the merged communes has been made in the 6 merged communes of sample C and includes a total of 23 communes and 900 telephone interviews, c) In one metropolitan region, the Stockholm area, 50 telephone interviews have been conducted in each of the 16 communes chosen with regard to size, population increase, and political majority, which, together with 200 interviews from the city of Stockholm proper, total about 1,000 interviews.

The interview questions deal with, among other things, duration of residence,
experience of other commune types, knowledge about and attitude toward communal
self-government, party activity, and evaluation of communal services.

2. Information and propaganda investigation, a) An investigation of the coverage of communal affairs in the daily press and in local radio programs. This investigation covers a span of three month;? and approximately 70 % of the total daily press editions. Nearly 90,000 items have been coded, b) The organization and extent of election campaigns have been studied with the aid of interviews with all of the election managers in the 36 communes of the A-sample. Detailed information about various campaign activities as well as personnel and material resources of the parties have been obtained, c) The contents of the election propaganda in mass media as well as reports from campaign meetings have been closely studied.

3. The nomination process wit.iin all of the 44 selected communes has been studied, primarily with assistance of, undergraduate students. They have been allowed to participate in the internal nomination meetings held by the parties. They have interviewed leading local politicians and procured source materials of various kinds.

This large amount of material is analyzed mainly to clarify: a) which persons are active in the nomination process and b) how representative the nominees are. A special study c) of the recruiting channels for communal assemblies is basedon a mail questionnaire given to all those nominated in a sub-sample of the original sample (2,700 persons in 18 communes). The major function of the questionnaire is to give information about the background those nominated, i.e. their activities up to the time of their nomination, d) A number of nominations at the county level have been studied.

It should be stressed that obvious advantages are gained by using results from one sub-project in another. An example may be given: In the election survey, the respondent was asked whether he was interested in reading news concerning communal activities in his commune; which newspaper he reads if he feels that this particular newspaper gives sufficient coverage to communal activities, and whether he has confidence in what the paper has to say about such activities. The answers given to these questions can be completed by transferring them to the same tape data containing the results from the information study dealing with how much and what the newspaper mentioned by the respondent usually writes about communal issues in the commune where the respondent lives.

Among studies about to be started, mention should be made of an interview with elected members of the communal councils in the sample communes and studies of the work performed by the communal councils and of the communal councils and boards. A study of the activities of the representative communal organs — their connections with parties and organizations, etc. — poses problems other than those concerning the influence of the citizens. One confronts questions dealing with the administrative aspect of local self-government. The study of representative democracy in action thus forms a bridge to administrative studies.

Conceptual Basis. In a research project of this size, it is hardly possible to direct from the outset all the separate investigations toward proving a few theories or collections of hypotheses. Such a procedure would presume that one had in advance a profound understanding of what is being studied. We do not possess such a knowledge. Even if this prerequisite were a fact, limiting the scope to a few theories would no doubt be considered unsatisfactory and arbitrary by many of the participants in the project. We have chosen intead to proceed from two general and common starting-poits, or conceptual frames of reference — one theoretical and one practical — which will be succesively filled in.

For the theoretical frame of reference, we have chosen an equilibrium model. This simply means that the political system is throught of as continuously reacting to changes which occur within the technical-economic-social system. Furthermore, it seems appropriate to make a basic distiction between a registering or observing and a reacting function in the political system. Changes in society are registered and channeled through parties and pressure groups, through the administration and individuals. In small communes, personal contacts play an essential part. When it comes to reaction, that is, the formulation of the political decisions, one is again confronted with parties, pressure groups and the administration as active participants in the decision-making organs.

Traditionally there is a tendency to overestimate the importance of the first two of the afore-mentioned organizations in relation to that of the administaration. An apparent characteristic of the modern democratic state and commune seems to be that an increasing number of the demands and wishes of the citizens created by the changes in society are anticipated mainly by the administration. Under such circumstances, general elections are less important as an expresion of the will of the citizenry than as an essential instrument of control through which one can determine whether or not the anticipations were correct.

It is possible that an attempt to make a firm distinction between a registering and a reacting function in the political system makes it more difficult to apply a traditional key concept of political science, namely power. We are not, however, certain that this would be a detriment.

Both functional and rational (means-ends) explanations will be used in the various studies. When studying an organization separately and from within, it is natural to use a rational pattern of explanations regardless of whether it concerns a party or a communal council . When cne combines the separate studies and examines how the communal political system works in general, the functional way of analyzing the problem becomes essential.

Values which have been attached to local self-government by the general debate in Sweden form the practical basis for the research program. The central research task is, in this case, to study how the prospects for realizing these values vary according to the type of commune and how they are influenced by communal reorganizations.

The research group has sougnt to specify those particular values which have been attached to local self-government. Hitherto, attention has been paid mainly to those values which can be included under the heading "citizens' influence." Within each sub-project, it becomes a matte:- of operationalizing these specific values, i.e., devising a technique by which it will be possible to measure and study more closely the degree to which the values are satisfied within different kinds of communes.

It should be possible with the indicated outline to combine and summarize the many separate studies in the last stage of the project so that structural changes can be read as indications on various value scales. There are no a priori reasons to believe the optimum positions on all scales will be found in communes of the same type, i.e., that there should exist an optimum commune size when examined from all relevant points of view. Already if a number of different values show a tendency to reach optimum positions within the same type of commune, it should be possible to derive certain practical directions from the investigation. Even if this proves not to be the case, one ought to be able to describe, on the basis of forthcoming results, more exactly what happens as a consequence of communal reorganizations.

Organization of Research. The research project is led by a board of directors consisting of one representative from each of the Political Science departments at the five universities in Sweden. The writer of this paper has functioned as chairman of the board and program leader. For the different sub-projects, there are special project leaders. We have worked toward having each of the departments participate in all of the sub-projects whenever possible. Each member of the board also serves as local project head.

During 1965, the first year of operation, about 15 research assistants were employed, and this year the number will increase to about 25. Before the project has been completed, about 50 graduate students will have been employed for varying periods of time. During the firsi year, nearly 200 undergraduate student have written papers within the framework of the project, each representing at least two months' work. Basically, the first year's experience of research activity and co-operation between the departments has been good.

The project is financed be the Tri-Centennial Fund of the Bank of Sweden. Funds
granted to date amount to roughly 900,000 Sw. Kr. (approximately 180,000 dollars)
annually.

All this may sound rather impressive. But the voyage has just begun and a long
time will pass before the ship is scheduled to reach harbor and come to anchor
sometime during 1970.