Resumé

The literature of the last decade indicates two important lessons for future
research. The first is that political scientists and scholars in related fields have
rediscovered the nature and übiquity of politics. In particular, they have realized
that neither public administration nor organization theory can solve political
issues. The second lesson is that the growth of government has enlarged the list of
important political actors with the addition of numerous 'public' groups,
organizations, and institutions. The combined message of the two lessons is that
analyses of the public sector and policy-making should be sensitive to the
importance of political actors in the public sector. This paper is a preliminary
attempt (based on elite questionnaire data) to map the general structure of the
Danish policy-making system by focusing on the interactions among government
bureaucrats, interest organizations, and MPs. These elites are further included in
an analysis of public expenditure problems, with particular emphasis on the (often
neglected) role of political parties and their relationships to various groups and
segments of society. The emergent overall picture of the policy-making system
indicates problems that presumably cannot be 'solved' within the framework of a
democratic polity.

Two Important Lessons of the Last Decade

The mood of political scientists in the West seems to mirror the social and economic conditions of Western societies. The optimism of the 1960's and early 1970's is matched only by the pessimism of the late 1970's and early 1980's. During the last decade most Western societies apparently faced new difficulties and challenges: The public sector grew out of control, policy failures and lack of government authority became visible, badly needed reforms and reorganizations did not work out as planned, taxpayers reacted to the growth of taxation and public expenditures, etc. Referring to parts of the gloomy post-1973 literature Brian Barry commented that several authors were 'grotesquely over-reacting to

the disequilibriating effects of a sudden fourfold increase in the price of crude oil' (Barry 1979, 194). Barry probably was right, but some recent developments nonetheless ought to be mentioned because they indicate two important lessons for future research.

The first lesson is that political scientists and scholars in related fields have rediscovered the nature and übiquity of politics. In particular they have realized that neither public administration nor organization theory can solve political issues. Organizational reforms and rearrangements designed to reduce the public sector or to rationalize policy-making procedures are extremely difficult to implement because they invariably mobilize adversely affected interests. Donald P. Warwick's analysis of the U.S. State Department is a good illustration of this point. Warwick concludes that 'a reorganization plan pegged only to considerations of rationality is doomed a failure. . . . Almost any significant change in the executive bureaucracy touches the interests and self-definitions not only of the employees involved, but also of related congressional committees, constituency groups, and interested publics' (Warwick 1975, 205). This conclusion is quite consistent with findings of U. S. Congressional and policy-making research. Thus, J. Lieper Freeman (1965), Roger H. Davidson (1974, 1977), Randall B. Ripley & Grace A. Franklin (1976), and others used terms like 'political subsystems', 'subgovernments', 'cozy triangles' and 'policy-making triads' to describe the patterns of interaction among congressional committees, bureaucrats, and external interests in various policy areas. It may be that these triads are not 'iron triangles' any more, and several new types of political actors have certainly emerged in a number of fields (Heclo 1978; Davidson 1980; Jordan 1981), but the name of the game is still politics.

Politics is also what ultimately accounts for the 'policy failures' of 'the waning welfare state'. Fritz W. Sharpf notes that a uni_dimensional model of the policy-making process is inadequate and proposes another model in which 'each step in the policy process is regarded as the outcome; or result of interaction process in which a variety of participants with differing perceptions, differing institutional and personal goals and interests, and differing power and resistance potentials are pursuing a variety of strategies in the context of relatively stable interaction "games"' (Sharpf 1977, 348).

Numerous other studies could be quoted to illustrate the politics-isstill-politics' lesson. The crucial point is the following: In a non-dictatorial social and political order various political actors pursue their self-interests in all matters of importance to them by utilizing their resources through the channels of influence available. Whereas the collective or social outcomes

of these endeavours may be tolerable in times of economic growth,
they often create 'problems' in times of less fortunate economic conditions.

The second lesson is that the growth of government has enlarged the list of important political actors with the addition of a number of 'public' groups, organizations, and institutions. As Samuel H. Beer (1976) shows, political actors other than parties and private interest groups are influential in the 'new public-sector polities', in particular the professions and technocrats of the 'professional-bureaucratic complex'. There are no reasons to believe that the new public actors are motivated by considerations fundamentally different from those of 'private' actors. The corollary is that governing becomes more difficult as more problems and tensions in society are represented in government (Rose 1981, 24). A similar idea is forwarded by B. Guy Peters & Martin O. Heisler (1981) who note a paradox: While government has grown as measured by most indicators, its authority may actually have shrunk.

One important aspect of government growth is the increasing number of citizens employed in the public (service) sector. Although a satisfactory distinction between the 'public' and the 'private' sector is yet to be found (Peters & Heisler 1981), it appears that the political behavior of public employees is extremely understudied in political science. Electoral and mass behavior studies rarely introduce private vs. public employment as an independent variable. For that reason it is not possible to tell, for example, whether the potential emergence of a new class of young technocrats and professionals exposing a postmaterialist ideology (Inglehart 1981) in effect is a possible new class of public employees.

Such a possibility may seem a bit far-fetched, but some research indicates that it cannot be ruled out a priori in the Danish case. Ten years ago Jørgen S. Dich published a much debated book, The Ruling Class ('Den herskende klasse', reviewed in the APSR, vol. 71, no. 4 (December, 1977), pp. 1697f.) arguing that the employees in the health, education, and welfare sectors have become a ruling class whereas in the past they were servants of the public. Today, about 1/3 of the Danish labor force is employed by the state and local governments (Westergaard-Nielsen 1980; Jeppesen 1982). It should be mentioned that, apart from public transportation, communication, and utilities, nationalized industries are virtually absent in Denmark, that public service institutions (especially in the fields of education, health, and social welfare) are dominating in terms of the number of employees (Kristensen 1982a; Hibbs & Madsen 1981), and that the annual growth of the number of public employees was about 5 per cent in the latest five-year period (Nordstrand 1981). While no

systematic study of Danish public employees has been published so far,
some scattered findings of relevance can be mentioned.

A research team (Damgaard 1980) has demonstrated that in Denmark public employees are generally more active in politics than their counterparts in the private sector of society: They are more active in party politics, better organized in trade unions as well as more active in such organizations, and more likely to be single-issue activists (in particular those employed in health, education, and welfare) than employees in the private sector. Public employees also tend to be more leftist in their political orientations, more likely to perceive capitalist influence and less likely to perceive bureaucratic influence than employees in the private sector of society.

A study by J. Goul Andersen (1981) further shows that public employees are more supportive of existing social welfare programs than employees in the private sector. They also are less likely to think that welfare programs are abused and that politicians are too lavish with taxpayers' money. Interestingly, a breakdown into policy areas reveals that public employees primarily want increased funds for their own areas and not necessarily for other parts of the public sector.

The combined message of the two lessons is that analyses of the public sector and policy-making should be sensitive to the increasing importance of political actors in the public sector. As a first step in this direction, this paper attempts to outline the general structure of the Danish policymaking system by focusing on the interactions among government bureaucrats, interest organizations, and MPs. Ultimately, the nature and structure of these interactions explain the lack of successful reforms in the public sector.

The analysis is based on data from questionnaires administered to Danish
elites by a research team based at Aarhus University.'

Wherever possible, a distinction is made between the private and the public sectors to examine important similarities and differences. The following three sections focus on the relations between bureaucrats, interest organizations, and MPs in pairwise combinations. Then, under 'Political Actors and Public Expenditures' all elites are included in an analysis of some aspects of public expenditures with particular emphasis on the (often neglected) role of political parties and their relationships to various groups and segments of society. The paper is an attempt to put some pieces of new information together, and the emergent overall picture of an advanced welfare state suggests problems of government with no solutions' in a democratic political order.

Bureaucrats and Organized Interests

As mentioned above, political actors are assumed to pursue their selfinterests in policy-making. The actors may be of a 'private' or 'public' type, bureaucrats and institutions, organizations and groups, MPs, parties, citizens, etc. (Olsen 1978). In a democratic order no single actor is in a position to impose a policy upon the rest of society. Hence, they are all in need of allies for the making of decisions or for preventing undesired policy changes.

That interest groups pursue and protect rather well-defined interests is given by definition. The interests of bureaucrats are assumed to include increases in funding that can be used for organizational and personnel expansion, higher service levels, etc. MPs and parties above all need votes and are therefore likely to promote the interests of those groups and segments of society that can provide the votes. From these assumptions it follows that specific causes and issues tend to create alliances of particularly interested actors which, if effective, produce decisions of mutual benefit to the members of the coalitions and, possibly, at the expense of the society at large (Damgaard 1977,259-282).

Previous research (reviewed in Damgaard 1981; Buksti 1982) provides strong evidence for the existence of sectorization and specialization in policy-making. Extending this research, the analysis attempts to reveal possible similarities and differences in the political processes associated with 'private' and 'public' sectors respectively.

The organization of the central administrative bureaucracy, and particularly the horizontal distribution of tasks and responsibilities among ministries, may be analyzed in different ways. With Rose (1976) one can talk about 'state defining functions' (activities related to defense, internal order, and mobilization of finance) and other, more recently adopted, functions (mobilization of physical resources and provisions of social benefits). It is also possible to conceive of ministries as reflecting strong interests of different kinds. Apart from the ministries in charge of state defining functions, ministries are usually created as responses to demands from social, economic, and institutional interests. However, once established, all ministries develop a life of their own with distinct and selfish interests. That is why reorganizations not associated with growth or with the creation of new ministries are so difficult to implement.

Although tasks, policy-areas, and interests differ in a number of ways, a rough distinction can be made between private and public sector ministries. Some ministries are mainly in charge of the regulation of the private production sector of society, others are mainly responsible for the production and distribution of public services. Such a categorization is

proposed and used by Jørgen Grønnegård Christensen (1980, 1981) who, in addition, lists ministries with either coordination or general societal regulative functions. Christensen's typology is used with minor modifications in this analysis. The private sector ministries are: Agriculture, industry, fisheries, labor, and housing. The public sector ministries comprise social welfare, education, interior (health), culture, church, defense, and public works. Other ministries are finance, taxation, environment, justice, foreign affairs, and energy. The Prime Minister's Office and the Ministry of Greenland have been left out of the analysis.

Previous studies lead to the expectation that bureaucrats in most ministries indeed have close contacts to interest organizations. But there are two reasons for believing that bureaucrats in private sector ministries tend to be more involved with interest organizations than their colleagues in public sector ministries. One reason is the existence of big and well-established economic organizations with clear interests in the areas covered by the private sector ministries. Another reason is that public sector ministries often have 'clients' and interested parties other than organizations, as, for example, public institutions, specialized groups of state and local government employees, other interested publics, etc.

Table 1 shows that 58 per cent of the top and middle level bureaucrats have regular contacts (i.e. at least once a month) to interest organizations. It also reveals, as expected, that such contacts are more widespread among bureaucrats in private sector ministries than among bureacrats in public sector ministries. However, the contacts of bureaucrats in the two types of ministries presumably involve different kinds of interest organizations.

To test this possibility, interest organizations have been classified as belonging to one of four broad categories (based upon the more refined categorization of Buksti & Johansen 1977, 1979): Labor, white collar, other economic (employers, business, etc.), and non-economic interest

organizations. The first three types of (economic) organizations are primarily based upon the profession and education of their members, while the fourth represents all other interest organizations. The reason for this classification is the hypothesis, suggested by previous research on interest organizations, that private sector ministries have close contacts to labor and 'other economic' organizations, while public sector ministries are closely linked to white collar organizations and possible noneconomic organizations.

Table 2 gives some support to these expectations. The organizational environment of private sector ministries is to a very large part made up of organizations representing private production interests of various sorts while non-economic organizations, in particular, constitute a small part of that environment. The picture is completely different for the public sector ministries. Bureaucrats in these ministries are predominantly in touch with white collar organizations, presumably in most cases organizations representing the interests of public employees, and non-economic organizations in such areas as education and welfare.

All nation-wide interest organizations were asked about their possible contacts with the various ministries (in this respect the project is a follow-up for the Buksti & Johansen study, e.g., Buksti & Johansen 1979). It is therefore possible to look at the bureaucrat-organization relationship on the basis of organizational data as well. A total of 654 interest organizations replied that they were in regular contact (i.e., at least once a month) with at least one ministry. According to previous studies of interest groups (Buksti & Johansen 1979) and the above data on bureaucrats, the distribution of these contacts over types of ministries should vary by type of organization. In particular, the organizations representing private producer interests should be expected to concentrate

their contacts upon private sector ministries, while white collar organizations
and non-economic organizations should be expected to concentrate
upon public sector ministries.

Table 3 shows that these expectations are only partly fulfilled. As expected, white collar and non-economic organizations indeed concentrate their activities around public sector ministries, but 'other economic' organizations do not concentrate their contacts upon private sector ministries. In fact, half of their contacts involve 'other ministries'. It turns out that these ministries are primarily foreign affairs and environment, both of which are important to some aspects of private producer interests (foreign trade and environmental ragulation of business activities). Furthermore, Table 3 indicates that the interests of labor organizations involve a large number of ministries although the core contact ministry is the ministry of labor (Buksti & Johansen 1979).

A comparison of Tables 2 and 3 thus shows somewhat different patterns of bureaucratic-organizational relationships. The contacts between public sector ministries on the one hand and white collar and non-economic interest groups on the other, are basically reciprocal or symmetrical. This is not the case with private producer organizations and private sector ministries, as 'other ministries' are also very important to producer organizations, which, furthermore, figure quite prominently among the organizations that 'other ministries' are in touch with, cf. Table 2. Finally, labor organizations, accounting for a (numerically) small part of the organizational contacts of bureaucrats in all types of ministries, correspondingly report contacts with a wide range of ministries.

MPs and Bureaucrats

Without underestimating the importance of local constituencies (Damgaard 1982, 15 - 22), it seems obvious that party affiliation and policyarea specialization are the most significant characteristics of Danish MPs. The patterns of legislative specialization, in particular sectoral specialization explained by the educational and professional background of MPs, are described elsewhere (Damgaard 1977, 1980a, b). In the present context the purpose is to map the contacts of MPs with ministers and bureaucrats. Previous studies lead to the expectation that MPs' contacts with ministers and bureaucrats are highly specialized, that is, MPs on given (specialized) legislative committees are expected to concentrate their contacts upon ministries the jurisdictions of which correspond to those of the committees. There are no reasons to expect differences between public and private sector ministries in this respect.² However, there are reasons to expect MPs of the governing Social Democratic Party to report closer contacts with ministries than MPs of opposition parties, at least in as far as ministers are concerned (cf. Damgaard et al., 1979, ch. 5, on the informal contacts among Social Democratic MPs and ministers).

For all ministries, MPs were asked to indicate whether they had regular contacts (i.e. at least once a month) with the minister and/or various categories of bureaucrats. An affirmative answer is counted as a specialized contact if the MP in question is a member of the committee with jurisdiction parallel to that of the ministry in question, and as a non-specialized contact if the MP is not a member of the relevant committee. In Table 4 the level of specialized contact (and non-specialized contact) is measured by the average percentage of MPs reporting specialized (or non-specialized) contacts with ministers and bureaucrats.

Table 4 demonstrates, first, that the level of specialized contact with ministers and buraucrats is much higher than the level of non-specialized contact. This holds true for all types of ministries and for both categories of MPs. Second, there are no substantial differences with respect to type of ministry, that is, MPs are in touch with the three types of ministries to approximately the same degree. A third conclusion concerns important differences, however. Thus, MPs, irrespective of the Goverment/ opposition party distinction, have more contacts with the minister than with the bureaucracy. This fits well with the constitutional theory, according to which the minister alone is responsible, but the table nonetheless confirms that MPs have direct contacts with bureaucrats as well. It further appears that MPs of the governing party are slightly more likely to establish such contacts than opposition party MPs, although the latter dominate in absolute terms.

Organized Interests and MPs

The links of organized interests with MPs are less studied than their contacts with ministries, but quite important, as this paper will show. Interest organizations were asked to indicate the frequency of their possible contacts with legislative committees, legislative parties, individual MPs as well as with the Government (and ministries). Table 5 provides the answers for the various types of interest organizations. Percentages are based upon the (large) number of organizations reporting regular contacts (i.e., at least monthly) with ministries in the same series of questions.

Table 5 shows some rather striking differences with respect to the contacts of interest organizations with Parliament and Government. On all four scores, labor organizations are most likely to report regular contacts and white collar organizations least likely to do so. Other economic organizations and non-economic organizations occupy intermediate positions, although it is quite noteworthy that 42 per cent of the non-economic organizations have regular contacts with individual MPs. For all types of interest organizations, the percentages are highest in the case of contacts with individual MPs, but it should be kept in mind that individual MPs are also members of legislative parties and legislative committees (and some Social Democrats even of the Govenment).

It is difficult to explain these variations without further information, because the parliamentary contacts of interest organizations are related not only to subject-matter areas but also to party-political ideologies and interests. Thus, contacts of labor organizations are possibly concentrated upon MPs belonging to the Social Democratic party, which is formally linked to the labor unions, while some 'other economic' organizations presumably have most contacts with non-socialist parties. In the cases of white collar organizations and, in particular, non-economic organizations, the party-political component of contacts would seem to depend very much upon the specific subject-matter.

Although most Danish parties receive at least some votes from almost all major groups in the electorate, they tend to get particularly strong support from certain subsets of the population (Andersen & Glans 1980). The Social Democrats thus receive more than half of the labor vote and almost half of the votes cast by retired people. The Liberals (Venstre) obtain about two-thirds of the farmers' votes, the Conservatives get about 30 per cent support among self-employed in urban industries - as does the Progress Party of Mogens Glistrup - and about one fourth of the upper white collar vote. Half of the students vote for left wing parties, in particular the Socialist People's Party and the Left Socialists. The left wing parties are, furthermore, strongly supported by public employees while the Liberals and Conservatives do better among white collars in the private sector. The support for the Social Democrats is more balanced in terms of public/private occupation. An analysis of electoral developments in the 1970's (Andersen & Glans 1981) shows quite significant gains of the left wing parties and the Social Democrats among publicly employed white collars, particularly within the health and education sectors.

Such and similar differences in party support are probably reflected in
the behavior and policy attitudes of MPs. These differences are thus expected
to manifest themselves in the patterns of MP-contacts with interest

organizations. MPs were asked to report on the kinds of organizations
with which they had regular contact (i.e. at least monthly), cf. Table 6.

Some of the speculations arising from the attempt to interpret Table 5 above obviously get a solid base in Table 6. There are indeed striking differences in the organizational contacts of MPs, and most of them are quite as expected. Social Democratic MPs are in close contact with labor and white collar organizations while members of the left wing parties have such contacts to a somewhat lesser degree. The picture is; completely reversed for other economic organizations with which the Liberals and Conservatives, in particular, have close contacts. In the cises of employers' organizations and organizations of industry, etc., the Conservatives report the highest level of contact, while the Liberals are leading in the areas of agriculture and fisheries. Collectively, the four small center parties occupy intermediate positions. They have above-average contact with white collar and most other economic organizations, but not at levels, as high as those of one or two other parties. Finally, the Progress Party MPs seem to be rather isolated from the well-established economic interest groups.

The contacts of MPs with non-economic organizations are, as expected, somewhat more complex in terms of party affiliation. But some main contours emerge from Table 6. First, left wing and Social Democratic MPs appear, in contrast to all other parties, to have well-developed connections with organizations of environment protection, consumers, welfare, health, and women. The same parties plus the center parties similarly have close contacts with organizations in the areas of culture and energy. The Social Democrats and the center parties further have above-average contact with educational organizations.

Second, in all remaining non-economic areas one or more of the right and center parties (L, C, PP) report contacts with organizations at levels above the average as well as with one or two of the parties (or group;; of parties) just mentioned. The only exception to this is the unique position of the Liberals with respect to local governments which reflects the strong position of that party in rural parts of the country. The combination of parties differs a great deal, however. In some areas, such as housing and radio listeners, the pattern presumably reflects the existence of opposing or competing organizations with different party-political connections. In other areas, such as defense, the pattern probably reflects a cause with rather wide support among the parties.

MPs are thus not only in contact with bureaucrats but also with various
interest groups in private and public sectors of society. The connections
of MPs with organized interests are probably important for their attitudes

toward public expenditures in specific policy areas, and hence for
the problems of controlling the growing public sector.

Political Actors and Public Expenditures

The patterns of interaction described in the previous sections to some extent reflect coalitions of different types of actors with compatible self-interests in certain areas. There are good reasons to believe that some such structure of'political-bureaucratic decision-making' is a cause of governmental growth (Kristensen 1980). The following attempts to marshal some empirical evidence from the Danish elite surveys.

The bureaucrats were asked about the level of funding for their own area of the administration. The questions included a choice among three options phrased as statements of facts, the respondents bei ig asked to select the most correct one:

1. Moderately increased appropriations for the area would have beneficial
effects clearly justifying the increase of expenses.

2. The level of funding is adequate.

3. Moderate cuts within the area would yield a reduction of benefits so
marginal that it would be a reasonable price to pay for the value of the
cuts.

In Table 7, the answers of top and middle level bureaucrats are presented.

As Table 7 shows, roughly half of the bureaucrats think that an increase of funds would yield benefits clearly outweighing the costs, one-third consider the level of appropriations to be adequate, and only about one-sixth think that cuts could be made without serious problems. The balance of opinion is thus strongly in favor of increased appropriations, and, as it appears, mostly so within the public sector ministries. But bureaucrats in all types of ministries are really quite receptive to the idea of increasing their own funds. If the arguments of this paper are valid, one should expect that the bureaucrats would be able to find some allies among interest organizations and MPs.

All interest organizations reporting regular contact to ministries were asked whether - in connection with such formal or informal contacts or bargainings within the last year - they had supported or opposed certain proposals. In Table 8 the affirmative answers are reported for the questions of whether the organization had:

1. supported proposals from the ministry implying increased expenditures

2. opposed such proposals

3. supported the ministry in fighting concrete cutting proposals
4. supported cutting proposals

Overall, one-third of the interest organizations report that they have supported proposals increasing expenditures while only 13 per cent have opposed such proposals (Table 8, cols. 1 and 2). Similarly, interest organizations support ministries in fighting cuts more frequently than they support cutting proposals (cols. 3 and 4). There are some differences among the types of interest organizations, however. Thus, labor and non-economic organizations are most supportive of expenditure proposals while other economic organizations are least supportive. Still, bureaucrats meet allies more often than adversaries, the only exception being that other economic organizations are not very likely to support ministries in fighting cuts. A detailed analysis requires investigations into concrete cases of support/opposition which cannot be undertaken here, but the information of Table 8 can be used in combination with Table 2 to reveal some possible differences among the types of ministries.

Table 2 showed the distribution of bureaucrats' contacts with the various types of interest organizations. The propensity to spend public money in the organizational environment of the three types of ministries can be measured by an index which a) multiplies the percentages in Table 2 by the percentage differences between col. 1 and col. 2 of Table 8, and then b) adds up the products for each of the four types of organizations that each type of ministry is in touch with. In the hypothetical case where all organizatons support (and no one opposes) spending proposals, the value of the index would be 100. The opposite extrenie value is - 100.

The actual figures are, of course, positive and far smaller than 100, as the percentage differences of cols. 1 and 2 (Table 8) indicate. The propensity to spend in the organizational environment of the four types of ministries turns out to be:

Private sector ministries: 15

Public sector ministries: 20

Other ministries: 18

All ministries: 19

These differences are not dramatic, but the organizations surrounding public sector ministries seem to be more spendthrift than those surrounding private sector ministries. The tentative conclusion therefore is: Not only are bureaucrats in all ministries, and particularly those in public sector ministries, generally in favor of increased funds; also the organizations with which they have contact, and particularly those surrounding public sector ministries, are more in favor of spending than of cutting proposals. But how do MPs fit into this piaure?

The importance of specialization in legislative work was emphasized previously, cf. Table 4. MPs were asked the same set of questions as was put to bureaucrats concerning the level of expenditure (Table 7). It is therefore possible to compare the attitudes of MPs and bureaucrats in this very important respect. It should be noted that MPs were asked to relate their answers to the policy-area they knew best from legislative work (and hence, presumably, were most interested in or, perhaps, even the area in which they were most influential).

The overall percentage of MPs favoring increased appropriations is roughly equal to the corresponding figure for bureaucrats. Each policy area might therefore be thought to have a configuration of actors (bureaucrats, organizations, and MPs) favoring increasing spending over budget cuts. But Table 9 also shows that Parliament, in contradistinction to the bureaucracy, contains quite a large group favoring cuts in appropriations. This difference is clearly related to the paity variable. Left wingers are extremely happy with increased public spending in several areas (not including defense). The Social Democrats are also strongly in favor of public spending with a balance of opinion at a level roughly equal to that of the bureaucrats. Within the Center Parties and among Liberals, opinions are more divided or balanced while Conservatives and Progess Party MPs are strongly in favor of budget cuts. However, some MPs in all parties would like an increase of appropriations for certain policy-areas (Damgaard 1982, 83-85). For example, a member of the

Progress Party, a Liberal and an MP belonging to one of the center parties want increased defense expenditures, a Conservative a larger budget for the police, some Liberals more money for culture, agriculture, housing, etc. Hence, even among parties which in general favor cutting of public expeditures there are possible coalition partners for those who want to increase public expenditures in specific areas. In the last analysis, this is probably one of the main reasons for the failure of more than a decade's reform efforts to control the growth of government expenditures. As Grønnegård Christensen (1982) has shown, improvements of budgetary and planning instruments have not changed the behavior of parties in relation to cutting plans, and not prevented public expenditures from surpassing planned spending limits.

Most Danish MPs think that government expenditures should be cut, and they also agree, according to the questionnaire data, that the reason for the lack of success in this respect is that they do not agree on where to cut. As illustrated in the previous section, all the parties have special groups to cater for. The patterns of contacts with interest organizations (Table 6) and the information in Table 8 on the activities of organizations in relation to spending/cutting proposals nevertheless suggest some differences among the parties. It seems, but cannot be documented in this paper, that left-wing parties and Social Democrats have particularly strong links with those organizations (labor, white collar, non-economic) which are most in favor of public sector growth, while Liberals and Conservatives are aligned with private producer interests, the organizations

of which are not nearly as happy with public spending (in areas other than their own). The real world is of course much more complex, but a partial test is possible by means of two questions in the MP-survey probing into the importance of affected interest organizations and groups of employees in relation to appropriation decisions. If, as shown by Table 9, left wing and Social Democratic MPs are most happy with increases of public expenditures and if, as Tables 6 and 8 indicate, the same MPs have special links with interest organizations which are most supportive of government spending, one should expect MPs of these parties to pay more attention to the demands of affected organizations find groups than MPs of other parties. The reason, of course, is that such groups can provide votes at election day.

Table 10 demonstrates that left wing and Social Democratic MPs indeed are more attentive to the demands of affected organizations and groups, or at least more willing to admit such attention ijust as they are more ready to perceive themselves as representatives of special groups or classes in society, cf. Damgaard 1982; ch. 2). In general, there probably is a difference between parties of the left and right with respect to public spending, but it is important to remember that all parties have special interests to cater for, and that spending coalitions are much easier to organize than cutting coalitions (Kristensen 1980). Furthermore, parties essentially represent inconsistent voters, that is, voters who, on the one hand, want tax reductions and, on the other, increased public spending for a variety of purposes (Kristensen 1982b). The problems of the advanced, and organized, welfare state are, therefore, related to the dilemma of political representation and governing in a democratic order.

Concluding Comments

The data presented in this paper do not allow for firm conclusions, as more refined analyses are obviously needed. But they do provide some evidence for the points stressed by way of introduction. Danish society is highly organized and politically mobilized. The macro-political organization of society, reflected in the patterns of interaction among bureaucrats, interest organizations, and MPs, is tied up with special interests and entitlements. An attempt to reorganize or rationalize the public sector is, in essence, an attempt to reorganize or rationalize some part of society and, therefore, not likely to succeed given the resistance potential of interested actors.

Although interests associated with the public sector to some extent differ from those of the private sector, they are no less 'political. Some differences are revealed in the patterns of interaction between ministries and interest organizations. The organizational environment of private sector ministries is dominated (at least numerically) by employers' and business organizations and that of public sector ministries by white-collar and non-economic organizations. However, in other respects, the similarities are as pronounced as the differences. Bureaucrats in all types of ministries tend to think that their resources ought to be increased, and the organizations to which they have contact are more in favor of increased spending than of budget cuts, although the tendency in both respects is most pronounced for public sector ministries.

The level of specialized contact of MPs with bureaucrats does not vary with type of ministry, but there are important party differences concerning contacts with organizations (Table 6) and attitudes toward public expenditures (Tables 9 and 10). Political parties, and not least those of the highly fractionalized Danish party system, may have come under pressure as representational and governing instruments (Damgaard & Kristensen 1981), but parties should not be regarded or analyzed in isolation from other politically relevant forces in society. On the contrary, and as the data indicate, MPs of various parties participate in a very complex set of alliances formed around private and public, organized, and institutional, interests in society. It seems clear that 'public' types of interests have become more important in the last decade, but further studies are needed to show the full consequences of this development.

However, the 'problems' of the public sector alluded to in the introduction are not just problems of the public sector. They are related to the fundamental dilemma of reconciling interest representation and government in a democratic policy, and, as a massive literature demonstrates,

this dilemma has no genuine solution (Dahl 1982). Viewed in that
perspective, dreary pessimism is just as unfounded as naive optimism,
while some realism perhaps might help.

---

NOTES

1. The data have been collected within a Danish research project, 'The Political Process. Participation and Decision-Making.' The first part of that project is a mass survey on the political activities and attitudes of Danish citizens, cf. Damgaard, 1980, which focuses on party, interest group and single-issue activities as well as on contacts with public authorities and power perceptions, in chapters by Jørgen Goal Andersen, Jacob A. Buksti, J. Grønnegård Christensen, Erik Damgaard, Kjell A. Eliassen, Lars Nerby Johansen, and Ole P. Kristensen. The second part consists of a series of parallel elite questionnaire surveys comprising MPs, government bureaucrats, interest organizations (both organizations as such and individual staff of interest organizations). Only the MP data have been ready for analyses for some time, cf. Damgaard, 1982. This paper therefore only includes preliminary analyses of a few variables in the bureaucrat and organization (per se) surveys. The bureaucrat questionnaires were addressed to the total population of civil servants in the central administration, but only top and middle level bureaucrats are included in this paper (response rates: 60 and 53 per cent, respectively).

2. Similarly, roughly one-fourth of the bureaucrats in all types of ministries report that they are often or occasionally contacted by individual MPs.

REFERENCES

Andersen, J. G. 1981. 'Vælgernes holdninger til socialpolitikken' paper, 6th Nordic Political
Science Congress, Åbo, Finland, August 17-19.

Andersen, J. G. & Glans, I. 1981.'Socialklasser og partivalg i 70'erne\ Politica, no. 2, pp.
5-45.

Andersen, J. G. & Glans, I. 1980. "Sociale klasser og partivalg ved folketingsvalgene 1977
og 1979' (Aarhus: Institute of Political Science).

Barry, 1979. 'Is Democracy Special?' In Peter Laslett & James Fishkin, eds., Philosophy,
Politics and Society. Fifth Series Oxford: Basil Blackwell.

Beer, S. H. 1976. The Adoption of General Revenue Sharing: A Case Study in Public
Sector Polities', Public Policy, vol. 24, No. 2, pp. 127 - 195.

Buksti, J. A. 1982. 'Organisationer og offentlig politik', paper, seminar om korporatismen,
Stockholm, April 28.

Buksti, J. A. & Johansen, L. N. 1979, 'Variations in Organizational Participation in
Government: The Case of Denmark', Scandinavian Political Studies, vol. 2 (New Series),
No. 3, 1979, pp. 197-220.

Buksti, J. A. & Johansen, L.N. 1977. Danske organisationers hvem-hvad-hvor, København:
Politikens Forlag.

Christensen, J. G. 1982. 'Growth by Exception or the Vain Attempt to Impose Resource
Scarcity on the Danish Public Sector', Journal of Public Policy, Vol. 2, Part 2.

Christensen, J. G. 1981. 'Blurring the International-Domestic Politics Distinction: Danish
Representation at EC Negotiations'. Scandinavian Political Studies, Vol. 4 (New Series),
No. 3, pp. 191 -208.

Christensen, J. G. 1980. Centraladministrationen: organisation og politisk placering, Copenhagen.
Samfundsvidenskabeligt Forlag.

Dahl. R. A. 1982. Dilemmas of Pluralist Democracy, New Haven: Yah University Press.

Davidson, R. H. 1974. 'Representation and Congressional Committees', The Annals of the
American Academy of Political and Social Science, vol. 411, pp. 48 - 62.

Davidson, R. H. 1977. 'Breaking up Those "Cozy Triangles": An Impossible Dream?' Susan
Welch & John G. Peters, Legislative Reform and Public Policy, New York: Praeger
Publishers, pp. 30 - 53.

Davidson, R. H. 1981. 'Subcommittee Government: New Channels for Policy Making',
Thomas E. Mann & Norman J. Ornstein, eds., The New Congress, Washington, D. C:
American Enterprise Institute, pp. 99- 133.

Damgaard, E. 1982. Partigrupper, repræsentation og styring, Copenhagen: Schultz.

Damgaard, E. 1981. 'Politiske sektorer: Jerntrekanter eller løse netværk?', Nordisk Administrativt
Tidsskrift, vol. 4, pp. 396-411.

Damgaard, E. ed., 1980. Folkets veje i dansk politik, Copenhagen: Schultz.

Damgaard, E. 1980a. 'The Dilemma of Rational Legislative Action: Some Danish Evidence',
Leif Lewin & Evert Vedung, eds., Politics as Rational Action, Dordrecht: D.
Reidel Publ. Co., pp. 217 - 236.

Damgaard, E. 1980 b. 'The Function of Parliament in the Danish Political System: Results
of Recent Research', Legislative Studies Quarterly, vol.V, No. 1 pp. 101 — 121.

Damgaard, E. & Kristensen, O. P., 1981. Party Government under Pressure', paper, 6th
Nordic Political Science Congress, Åbo, Finland, August 17 - 19.

Damgaard, E. 1977. Folketinget under forandring, Copenhagen: Samfundsvidenskabeligt
Forlag.

Damgaard, E. et. al. 1979. Folketingsmedlemmer på arbejde Aarhus: Forlaget Politica.

Freeman, J. L. 1965. The Political Process.- Executive Bureau-Legislative Committee Relations,
New York: Random House.

Heclo, H. 1978. 'Issue Networks and the Executive Establishment' Anthony King, ed.,
The New American Political System, Washington, D. C: American Enterprise Institute,
pp. 87- 124.

Hibbs, D. A. & H. S. Madsen, 1981 Public Reactions to the Growth of Taxation and Government
Expenditure', World Politics, pp. 413 — 435.

Inglehart, R. 1981. 'Post-Materialism in an Environment of Insecurity', The American
Political Science Review, vol. 75, No. 4 pp. 880 - 900.

Jeppesen, S. L. 1982. 'Den offentlige sektors vækst og indretning', Jens Gunst, ed., Offentlig
ledelse og samfunnsøkonomiske nøgleproblemstillinger, Copenhagen: Samfundsvidenskapbeligt
Forlag pp. 127 - 154.

Jordan, A. G. 1981, 'Iron Triangles. Woolly Corporatism, and Elastic Nets: Images of the
Policy Process', Journal of Public Policy, Vol. 1, Part 1, pp. 95 - 123.

Kristensen, O. P. 1982a. 'Oversigt over årsager til væksten i den offentlige sektor', Jens
Gunst, ed., Offentlig ledelse og samfundsøkonomiske nøgleproblemstillinger, Copenhagen:
Samfundsvidenskabeligt Forlag, pp. 155- 188.

Kristensen, O. P. 1982b. 'Voter Attitudes and Public Spending: Is There a Relationship?'
European Journal of Political Research, Vol. 10, pp. 35 - 52.

Kristensen, O. P. 1980. 'The Logic of Political-Bureaucratic Decision-Making as a Cause
of Governmental Growth', European Journal of Political Research, vol. 8, pp.
249-264.

Nordstrand, R. 1981. Personaleforbrug, produktion og arbejdsproduktivitet i den offentlige
sektor i Danmark, Copenhagen: Københavns Universitets Økonomiske Institut, memo
no. 105.

Olsen, J. P. 1978. 'Folkestyre, byråkrati og korporativisme', Johan P. Olsen, ed., Politisk
organisering, Oslo: Universitetsforlaget , pp. 13-114.

Peters, B. G. & Heisler, M. O. 1981, Government: What is Growing and How do We
Know?' Glasgow: Studies in Public Policy, no. 89.

Ripley, R. B. & Franklin, G. A. 1976. Congress, the Bureaucracy, and Public Policy.
Homewood, 111.: The Dorsey Press.

Rose, R. 1981. 'What if Anything is Wrong with Big Government', Journal of Public Policy,
Vol. 1, Part 1 pp. 5-36.

Rose, R. 1976. "On the Priorities of Government: A Developmental Analysis of Public
Policies', European Journal of Political Research, vol. 4 no. 3, pp. .247 - 289.

Scharpf, F. W. 1977. 'Public Organization and the Waning of the Welfare State. A Research
Perspective', European Journal of Political Research, Vol. 5, pp. 339 - 362

Warwick, D. P. 1975. A Theory of Public Bureaucracy, Cambridge, Mass.: Harvard University
Press.

Westergaard-Nielsen, N. 1980, "Trends and Determinants in Public Employment in Denmark,
1921 - 1977', paper prepared for the International Institute of Public Finance,
1980 Congress.