University of Stockholm

1. Introduction

The share of the gross national product devoted to public consumption and investment has grown in all industrialized economies. It is also larger in rich nations than in poor ones. Thus, there appears to be some relationship between the level of economic development and the size of the public sector, but the nature of this relationship is not quite clear. Since many social scientists have tried to come to grips with this question in recent years, this article seeks to review some seminal efforts, to propose a typology of explanations, and to suggest a few directions for further research.

The fundamental issue on the macro-level is to account for (1) the differentials in spending between various political units, and (2) the observable change over time within various units. The empirical base for hypotheses in this field consists of cross-section comparisons and time-series analyses, either of all government expenditures or of particular sectors such as education or defence. While most studies have dealt with nations, some important work has also been done on American states and on municipalities in several countries.

At the micro-level, the main problem is the causation of expenditure decisions in general. As such decisions make up a considerable part of all political decisions, the question is very close to the core concerns of the political scientist. Inquiries into the distribution of power, the moulding of public opinion, the processing of demands, the political role of bureaucracy, the capabilities of political systems, and the procedures of decision-making are all relevant to this topic. Hence, a large number of studies on politics and administration could afford insights into the determination of public expenditures. Some of my information is drawn from this research, but much more should be available for further analysis. The political science literature is richly strewn with comments on the causes or conditions of particular decisions and with middle-range inductions of great potential value.

2. The Discussion on Wagner's Law

The theory of growing government activities is of old vintage, but before it was
launched some ninety years ago an opposite hypothesis held sway for several

hundred years. The faith in progress was often combined with a belief in the decline of political power. In the age of Enlightenment it was a common view that the scope of government actions would diminish through the moral and economic evolution of mankind. Authority was required in so far as people were unable to control themselves and manage their own affairs. As they matured, however, they would no longer need political guidance. This outlook was shared by several leading French and German social theorists, including Karl Marx. Yet the early socialists were also convinced that production would become increasingly 'social' - or collective - and in a certain sense they held that the public sector would ultimately encompass the whole economy. In their analysis of the present society, however, the nineteenth-century socialists saw no secular trend towards growth in public expenditures. Some classical economists, among others Adam Smith, paid attention to tendencies in this direction, but there was no attempt to translate such observations into general theory.¹

The first scholar to recognize a positive correlation between the level of economic development and the size of the public sector was Adolph Wagner, German economist and Kathedersozialist.² In Finanzwissenschaft (1883) and Grundlegung der politischen Wissenschaft (1893) he contended that an increase in public spending was a natural consequence of economic growth. Several processes accounted for this. First of all, the costs of defence would go up as all nations invested in more sophisticated weapons. Secondly, domestic protection would also become more expensive through industrialization. The density of modern living would lead to more social frictions, and the complexity of an advanced economy would require new forms of public control.

A third reason for more public expenditures was the greater ability of the government to meet certain income-elastic demands. Education and culture were fields where collective producers were by and large more efficient than private ones. Hence, the public sector would grow as basic needs were satisfied and the consumption pattern shifted in this direction. In some other areas - e.g., railroads - the capital required for investment would become so large that it could hardly be provided through private accumulation.

The modern discussion has dealt less with Wagner's original propositions than with the general trend he predicted. An Italian economist, F. S. Nitti, made an early survey of public expenditure trends in various nations. Among the chief causes of expansion in the last part of the nineteenth century, he listed: (a) the continued increase in military expenditure, (b) the beginning of great public works connected with the use of steam and electricity, (c) the growth of public debts, (d) the development of all forms of social prevention, and (e) the increasing participation of all the people in public affairs. When the control of the government rested with the people, wrote Nitti, they did not look upon taxation as a loss. Hence democracy was one important determinant of government growth. But it could not be the principal cause, since despotic and oligarchical regimes such as that of Tsarist Russia appeared to be spending as much as the democratic governments.³

In a similar study, Arnold Brecht compared public spending in various German

Lander during the Weimar republic. As state activities in Hamburg, Liibeck, and Bremen by far exceeded those of Prussia, Baden, and Hesse, he inferred that population density was a crucial variable, explaining at least the differentials in expenditures on housing, welfare, and internal security. With Brecht, the most powerful determinant of growth in collective consumption appeared to be urbanization ⁴

Yet the effects of the social processes discussed by Wagner, Nitti, Brecht, and others were not automatic and immediate. Another German theorist, Herbert Timm, sought to account for the variations in growth among different nations by introducing the notion of 'lags'. First of all, there was a 'natural lag' involved in the emergence of new demands: such public goods as education or health were required in large amounts only when certain basic needs had been satisfied. Second, there was a 'system lag' in a capitalist economy that retarded the distribution of economic and political resources. Third, some time passed before new demands could be processed through the organs of political pressure and decisionmaking ('institutional lag'), and a final countervailing force was the lingering influence of laissez-faire ideas of 'financial classicism' ('ideological lag').⁵

Along a similar line, but from the reverse viewpoint, Karl Deutsch dealt with the accelerating impact of modern communications. The process of social mobilization had rendered the popular pressure for public services more efficient, and the diffusion of knowledge had made people more aware of what there was to get. Thus, several processes conspired to expand and intensify the demand for government action.⁶

Numerous analyses of the development of government expenditures in different countries have corroborated the assumption of tendential growth and added new suggestions as to its causes. In tracing the expansion of Swedish civil government expenditures from 1913 until 1958, Erik Hook put particular weight on various shifts in consumer demands and on the complementarity between private and public goods. He also emphasized that many activities had been moved from the private to the public sector and that cost increases in such areas as health were partly due to new scientific discoveries and technical inventions.⁷

In his work on American government activity in the first half of the twentieth century, Solomon Fabricant mapped interstate differences in public expenditures. He concluded that the government's expansion was a concomitant of economic growth and that most variance between states could be explained by variations in income, urbanization, and population density. Income was by far the most important variable, but urbanization was a better explanation of expenditures in such areas as fire protection, sanitation, and social welfare.⁸ Harvey Brazer made a similar comparison of U.S. city expenditures and found density, median family income, and intergovernmental revenue to be the most powerful explanatory variables.⁹

Several later American studies have included political variables, such as indicators of party competition, the strength of the Democratic and Republican parties, the turnout in the elections, and equity of apportionment with respect to urban and rural districts.¹⁰ According to the findings of Thomas Dye, interstate variance

could be better explained by socio-economic conditions than by differences in the constitutional framework, electoral system, party system, power structures, or political style. But Dye's conclusions have been criticized, and other works ascribe more influence to the political context.¹¹

Time-series studies of government expenditures have been carried out for Canada, Britain, Germany, Italy, Japan, and many other countries.¹² Ronald W. Crowley has boldly tried to survey the fluctuations in government activity in Western Europe since the eleventh century, finding some support for Henri Pirenne's theory of swings between periods of laissez-faire (11-12th centuries, 15-16th centuries, 19th century) and periods of public control (13-14th centuries, 17-18th centuries, 20th century).¹³ Other works have focused on national variations in a region, e.g., Irving J. Goffman and Dennis J. Mahar's study of six Carribean countries.¹⁴

In an important contribution, Frederic L. Pryor has compared the development of public expenditures in a sample of market economies and centralized economies. The economic system variable was found to have played a statistically significant role in accounting for variations in expenditures on education, research, non-military external security, and possibly internal security, but not for defence, health, and welfare. Spending on internal security, foreign aid, and research and development appeared to be related to per capita income, but not spending on military purposes, education, and health.¹⁵

All long time-series analyses of industrialized nations have confirmed the assumption that the public expenditure share of the GNP tends to increase with the growth of the GNP per capita. For the developing countries, however, the data are less convincing. Between 1955-1957 and 1963-1965, the level of government tax receipts as a percentage of the gross domestic, as reported in United Nations World Economic Survey of 1967, rose in 18 developing countries, sank in 7, and remained constant in 4.¹⁶ Several cross-nation comparisons have failed to provide evidence of any clear correlation between per capita income and government expenditures. The most comprehensive one, by S. Lall, covered 46 developing nations and found no statistically significant relationship between GNP per capita and total government expenditures or several functional subcategories of government expenditures.¹⁷

As Wagner's prediction concerned only the 'progressive' nations, i.e., the industrializing countries, these results hardly vitiate his thesis. Nor do they undermine those explanations that link the growth of the public sector to the transformation and increasing complexity of the modern economy. One author, Bernard Herber, has proposed a three-stage variant of Wagner's law, according to which there is first a less than proportionate rise in government spending, then a more than proportionate rise, and then again a less than proportionate rise as the industrial economy comes of age. In the pre-industrial phase, subsistence goods would be more vital to the consumers than public goods, and in the post-industrial stage government would already have provided those economic goods which it could provide with an efficiency advantage.¹⁸

Some data support the hypothesis that the growth curve of government

expenditures bends downwards as a society passes from industrialization to postindustrial development, but the evidence is far from conclusive.¹⁹ In Sweden, to quote one counter-instance, the public expenditure share of the GNP rose more swiftly in the last fifteen years than in any previous fifteen-year period in peacetime. And while it appears evident that pre-industrial economies exhibit no unequivocal tendency to steady expansion of the public sector, there is no proof of regression in the ratio of government expenditures either. Wagner's law, if conceived as a simple prediction that the share of the GNP devoted to public expenditures tends to grow in 'progressive' nations, has been corroborated by most studies in the field, but Herber's law of a decline-growth-decline cycle stands as yet unconfirmed.

3. Three Levels of Explanation

It should be plain from this sample of studies that many propositions have been offered as to the causes of the growth of government activity. Some explanations appear to be merely specificative, i.e., they specify the particular subsectors that account for the net increase in public expenditures. Thus, when we are told that growth of defence costs or the transfer of certain activities from the private to the public sector is responsible for a certain change in government spending, we learn something about the structure of the expansion but nothing about its determinants. Other explanations, however, purport to give us reasons for the rise hi government expenditures and not solely a profile of growth. These substantive explanations will be more relevant to our inquiry.

Among the substantive explanations there are references to several kinds of determinants. Some scholars focus on economic variables while others emphasize ideological or institutional conditions. Often, one type of explanation does not seem to preclude another. A growth process may be explained by changes both in the economic system, the social value system, and the political system. Many authors apparently assume a 'causal flow' of a more or less historical-materialist character: innovation in the technological sphere leads to a transformation of the relations of production and the pattern of living, which in turn leads to the emergence of new demands that slowly work their way through the institutions of political articulation, aggregation, and decision-making, and finally result in government action. Yet even in this picture, events or processes on one or several levels in the flow are usually singled out as triggers of government expansion. Schematically, we may distinguish between three different planes in the explication of public expenditure growth: one socio-economic, one ideological-cognitive, and one political-institutional.

(1) Explanations on the socio-economic level refer to changes in the technological, economic, social, or demographic structure. Several scholars have dealt with the various concomitants of industrialization. For instance, new production methods and larger markets require a better provision of communications, manpower training, and internal security. The density of urban living gives rise to

social frictions and control problems. Shifts in the population pyramid may entail greater costs for pensions, old-age care, health, education, and child care. The need for public services may also increase as a larger share of the female population joins the labour force. Urbanization carries with it a number of costly investments in the towns, and may also make it more expensive to provide collective services to a diminishing population in the countryside. New patterns of production lead to one kind of growth in public expenditures; new patterns of consumption to another.

(2) Explanations on the ideological-cognitive level focus on changes in knowledge, beliefs and desires. These may but need not be closely related to the development in the socio-economic sphere. Shifts in consumer demand is a case in point. Interest in superior goods - among which a proportionately large number are provided by the government - is likely to increase as income per capita goes up and certain basic needs have been satisfied, but the emergence of new demands may also be due to influences less directly tied to economic growth. Intellectual trends, moral ideas, scientific discoveries, and religious beliefs all play a part in the determination of public expenditures.

Views may affect the scope of government action even if they are shared by only a small number of people. In the extreme case, one man's opinion may be sufficient, but any system is liable to be influenced by elite values and norms. A new theory communicated to a small circle of policy-makers may thus have an impact on public spending. In other contexts, the crucial effects are reached only when certain views are spread to large strata. Ideological-cognitive explanations can concentrate either on innovation or on diffusion. In the first instance it is a question of new tastes or beliefs; in the second a question of old ones gaming foothold in new groups.

(3) The mere existence of a belief or a demand does not make it politically efficient. To make an impact on the decision-makers it must be articulated, aggregated, and channelled so as to affect the political calculus. Shifts on the political-institutional level may therefore influence the effective policy output. Constitutional reforms or the emergence of interest organizations or parties can change the balance of power and add weight to the demands of hitherto neglected groups. An extension of the franchise, for instance, enables new strata to influence public policy. Changes in parliamentary procedure, the 'code' of political behaviour, or the political culture may have similar effects. Explanations on this level may also refer to more transient constellations: an alliance between political forces, a successful party strategy, or a charismatic personality.

4. The Three Sides of the Public Sector

Most explanations covered thus far have departed from a 'consumer perspective' on the public sector. New public outlays are held to be the result of socioeconomic changes, changes in values or changes in the power of values. Expenditures grow because there is a growing demand for public goods.

The assumptions of the 'consumer perspective' are widely shared among social

scientists. When economists discuss public expenditures they frequently posit that political decisions in one way or another reflect consumer preferences and, inversely, that shifts in consumer demand will make an impact on the policymakers. Political scientists have taken great interest in the transmission of such preferences and developed a variety of analytical frameworks to describe the process. The paradigms of David Easton and Gabriel Almond are two examples of schemes that, though theoretically 'empty' and capable of accommodating many different theories of policy initiation and distribution of power within a political system, nevertheless lend themselves particularly well to an analysis of politics from the 'consumer perspective'. The Eastonian analyst is predisposed to look for demands originating in the social environment, i.e., chiefly among the citizens, and then combining and pushing their way to the institutions of rulemaking. In a similar vein, the Downsian analyst wishes to know how parties or candidates vie for power by adapting to each other and by making an appeal to various groups in the electoral market. A common presupposition in both cases is that, given a multi-party system and universal suffrage, it is ultimately the citizens as consumers of public goods who decide on the scope and substance of government action. This comes very close to the political doctrine taught in all democratic nations. The notion of 'the citizens in power' is übiquitous in our civics textbooks.

According to a somewhat tougher appraisal of the democratic system, the political process is both a competition between consumer demands for different public goods and a competition for power and influence between incumbents and their challengers. But even this analysis tends to blur the likelihood that there may also be competition between producers. In politics, as on the market, we have supply and demand, and there is no reason to believe that the suppliers of public goods are idly awaiting the decisions of the consumers. Their interests are also at stake in the determination of government expenditures, and it seems likely that they have their part in the decision-making process.

Who, then, are the suppliers of public goods? In one sense, it is the political actors who offer the electorate a choice between several 'packages' of policy outputs. In the Downsian vision of the political market-place, the politicians are providers of values sought by the citizens. Yet a more solid supplier interest is represented by the actual producers of the public goods: the civil servants, the teachers, the nurses, the police, etc. A second category consists of those in private enterprises who sell most of their production to the government. In this group, it should be noted that both owners and employees may be counted as public producers. As used here, then, the term 'producer' does not refer to industrialists or entrepreneurs in general but to those groups in society that produce public goods.

When the growth of public expenditures is viewed from the 'producer perspective', emphasis is placed on the efforts by various supplier collectives to influence the scope and substance of government activities. Such efforts may be made either through regular bureaucratic channels or through producer politics in the electoral or parliamentary arena. As the public sector grows, it seems likely that the producer interest will carry greater weight in the political process.

A third angle from which the expansion of the public sector may also be viewed could be called the 'financial perspective'. On the private market, the phenomenon 'demand' might be said to be a product of two separate dispositions: the willingness to acquire a good and the unwillingness to part with another, usually money. A purchase takes place if the former of these dispositions outweighs the latter, and the preferences of the consumers are thus revealed in manifest action. In the consumption of collective goods, however, the two elements of consumer demand are split and the citizen cannot strike a balance. As purchase decisions can seldom be individualized to each particular good and to each separate consumer, the two dispositions appear in the guise of, on the one hand, financially unrestrained demands for certain goods and, on the other, unwillingness to contribute to the financing of public goods in general. The confrontation between these two dispositions must take place at a super-individual level through certain institutionalized processes.

Or, to put it differently: whereas transactions in the private market are a function of two variables, i.e., demand and supply, transactions on the collective market are a function of three variables, i.e., demand, supply and finance. Graphically the relationship could be illustrated by an equilateral triangle (Fig. 1). A given volume of government activity presupposes a certain quantum of each variable. An increase of the public sector would then require an increase in both supply (production), demand (consumption), and finance; all three sides of the triangle must grow. A new undertaking would be detectable as growth along three different dimensions. The crucial question is then what causal relationship exists between the three sides.

The advocates of the first perspective obviously presume consumer demand to be the independent variable. When more or new public goods are required by the citizens, the system will respond by arranging for more finance and more production. Yet one might also argue that the causal pattern is a different one. Some authors presume a reverse process: an increase in public consumption could be a consequence either of new resources (on the finance side) or of expanded production. A few examples of such hypotheses may give a better picture of the 'producer perspective' and the 'financial perspective'.

Forces on the 'finance side'

The best known representatives of the 'financial perspective' are probably Alan T. Peacock and Jack Wiseman, whose study of government expenditures in Great Britain broke new ground in the search for explanations of Wagner's law. Instead of emphasizing new elements hi consumer demands, this work focused on the citizens' opinion of a tolerable tax burden as the decisive restriction hampering growth in public spending. In normal times, this opinion is reasonably stable and does not allow any substantial increase in government expenditures. War and crisis, however, may drastically upset established values and make citizens more prepared to abstain from private consumption. As public expenditures are easier to shift upwards than downwards, however, they will never return to their previous level. Over time, therefore, the curve of public expenditures as a ratio of the gross national product will be shaped like a stair-case. This pattern of development has been called 'the displacement effect' or 'the ratchet effect.'²⁰

In this theory, then, the immediate cause of the growth process is a change hi mass attitudes. The explanation clearly belongs to the ideological-cognitive level. The same is true of G. S. Gupta's hypothesis, which deduces both the expansion of the public sector and its presumed slowing-down at a later stage from the individual voters' cost-benefit analysis of government spending. With a proportional tax system, the large majority will have a positive interest in the provision of more services as long as their benefits exceed their costs. Hence, the growth of the public sector will be greatly stimulated by the extension of the franchise but then retarded when a point is reached where most citizens are affected by fairly severe taxation.²¹

But the 'financial perspective' can be employed on other levels too. Harry Oshima has pointed out that the taxable base is more substantial in an advanced economy where most basic needs have been satisfied.²² If a 'subsistence minimum' is deducted from the national resources that the government can tap, then an increase in the GNP will lead to a more than proportional increase in its taxable part. The mode of production and the scope of economic exchange are also crucial to what Gabriel Almond and J. Bingham Powell have called 'the extractive capability' of a political system.²³ Institutional factors may be of similar importance. For instance, several American studies have shown that a high proportion of state revenues received as federal aid is positively related to state expenditures ²⁴ The structure of the tax system, the distribution of incomes from different sources, and the collection methods employed are likely to influence the size of the public sector.

Forces on the 'Producer Side'

Some explanations of public expansion point out the growing supply of labour as a critical factor. In the neo-Marxist analysis of Paul A. Baran and Paul M. Sweezy, it is suggested that the secular stagnation in the advanced capitalist economy pushes the government to absorb idle resources.²⁵ Another common assertion is that the costs of public goods tend to rise faster than the costs of private

goods because productivity cannot be increased as much in the services as in industry ²⁶ On the ideological-cognitive level, such changes become manifest as demands for more public employment or demands for higher salaries and better work conditions by the public employees. Other interpretations on the same level look to the efforts of civil servants to expand their territory.²⁷ If, as several authors suggest, all administrative units have an inbuilt tendency to widen their 'markets', then the crucial question is whether the routines of the administrative and political system promote or bar such attempts.²⁸ Thus, institutional arrangements appear significant on the 'producer side' also; changes in bureaucratic structure or decision procedures could in some circumstances explain growth in public expenditures.

5. Explaining Post-Industrial Growth in Public Expenditures

As shown in this brief survey, numerous theories link the expansion of government activity to the process of industrialization and its social concomitants, and some go on to suggest retarded growth or even a declining ratio of public spending to the GNP as a society enters into the post-industrial phase. Thus, Gupta maintains that the majority will become less tolerant towards increases in the tax burden when they must carry the brunt of it, and Herber asserts that an abundant consumer society will gradually turn to goods that can be produced with an advantage by the private sector. Though some evidence may substantiate these claims, they do not appear to be universally tenable. The continuous growth of Swedish public expenditures, for instance, has not slowed down but rather accelerated in recent years. The 1960s in particular were a time of rapid expansion, especially in regional government (landsting) and local government (communes). In the last two decades, Sweden has experienced a greater increase in government spending than most other OECD countries. Its total public civilian expenditures as a proportion of the GNP are appreciably higher than those of similar countries, e.g., Canada, the United States or New Zealand.

Two questions are raised by this case. First, why is the Swedish public sector so large? And second, why does it continue to grow? In using these two questions as focal points for our discussion, I shall proceed to illustrate the nine types of explanations distinguished in this article (Fig. 2) by listing some pertinent factors that may have contributed to sustaining the continued growth of public expenditures in a post-industrial society. The purpose is not to present a detailed or exhaustive analysis but merely to point at some variables that may go into the boxes and may be relevant for a more thorough examination of causal relationships hi this domain.

It should be borne in mind that the 'causal flow' discussed above renders it difficult to make a firm classification; many hypotheses could rightly be placed in several fields. To avoid repetition, however, I deal with most explanations on one level only.

A. The Consumer Perspective: Socio-Economic Determinants

Industrialization is featured in most theories as one of the key factors behind the rise in public expenditure. If this rise continues in post-industrial society, two major reasons for this may be suggested: (1) that post-industrial society is not as post-industrial as it seems, and (2) that many of the expansive effects are not tied to industrialization in a narrow sense but rather to the transformation of the economy.

On the first point, it should be noted that the pace of structural development in industry has hardly slowed down. It may be true that the secondary sector, as it is often called, no longer increases its share of the labour force and that most new hands are absorbed by the tertiary branches of the economy, but in terms of output and productivity gains industry is still pre-eminently dynamic. Secondly, the consequences of a growing service sector are in many respects similar to those of a growing industrial sector. The transfer of manpower from one predominantly urban sphere of the economy to another has no strong impact on the need for public consumption. The decisive break with the past was performed by the mass exodus from agriculture.

Some processes of post-industrial change with immediate relevance to the scope
of government activity are the following:

(1) Geographical Mobility. The pace of urbanization in Sweden has been relatively even over the last hundred years. Some of the 'urban' costs of this migration are probably smaller in the present late phase because of economies of scale, but others have increased because larger investments in social and educational

infra-structure are now required for each migrant. New towns are severely affected by these demands. At the same time, the 'rural' costs of urbanization appear to increase over time. As the population base shrinks in the countryside, public services become more expensive to maintain, and this tendency is aggravated by the simultaneous development towards larger units of production.

Other costs are incurred by the trans-national migration typical of post-war Western Europe. Since a very high proportion of the immigrants are part of the labour force, their contribution to government revenue at various levels probably exceeds the benefits they can obtain, but some new expenditures for resocialization, immigration control, interpreters, etc., should nonetheless be taken into account.

(2) Demographic Change. Aging is a considerable strain on the public purse. Expenditures for pensions and old-age care increase with the life-expectancy of the average citizen. Health costs are also affected in a rather complex fashion, since more spending in this sphere is likely to raise the life-expectancy, which will in turn increase the demand for medical care.

At the other end of the spectrum, an increasing ratio of children to adults may lead to new expenditures for day-care and education. In a modern society this is more likely to be due to migration than to mounting rates of nativity. Imbalances in the population pyramid in particular parts of the country, especially in new towns, entail two kinds of costs: an increase in current expenditures and higher depreciation costs for capital investments that can only be exploited for a limited period of time.

(3) Technological Change. New techniques of production affect the public sector in several ways hi that they give rise to demands for several new services by the government and for increased public control. The innovative enterprises themselves are in need of manpower, infrastructural investments, legal protection, communication, information, research and development, and many other forms of assistance. For the trade unions and their members, technological change may create new problems that are best met by government intervention. And with the consumers, finally, a multitude of control and inspection needs arise out of the rapid proliferation of new industrial goods, polluting methods of production, and growing complexity of everyday life. As Adolph Wagner suggested, changes in technology may also pave the way for a higher degree of government intervention in the economy because investments tend to become so large that they cannot be financed by private interests.

(4) Market Extension. In this context, it is noteworthy that many costs are related to the extension of markets. This process increases the effective integration of various geographical areas and carries with it a demand for more public control. When a country becomes much dependent on foreign trade, the government is likely to become involved both in the marketing effort and in a number of related activities carried out in cooperation with other states. Another dimension of the extension of markets is the transition from home production to industrial provision of goods and services. As households reduce then* autarchy and come to rely on economic exchange, part of their needs can most efficiently be

catered for through collective production. The increasing division of labour may
also require new safety measures to protect the uninterrupted supply of necessary
goods to the families.

(5) Changes in the Labour Force. One aspect of this development is that an increasing share of women join the labour force, i.e., that part of the population that works in the market sector or in the collective sector but not in the households. Many services no longer supplied in the households are now provided by communal or regional institutions, especially care in various forms.

B. The Consumer Perspective: Ideological-Cognitive Determinants

On this level, our main concern is to examine those beliefs, values or aspirations that make consumers keen on a more extensive or intensive provision of public goods. We are interested both in views that are new in the sense that nobody ever held them and in views that have spread to new groups and thus become politically efficient. Moreover, we are interested in the very process of transmission, for in an historical perspective it is obvious that one of the most remarkable features of the modem society is the greater scope and velocity of the diffusion of information.

(6) Income-Elastic Demands. A number of new consumer demands are immediately connected to the higher standard of living. As subsistence needs have been satisfied, various 'superior' goods come to the fore. Among these are many that can best be provided by the public sector and others requiring complementary action by the government. New patterns of consumption may also raise the demand for free or subsidized public services such as justice, theatres, or parks. Some of the effects of the rising standard of living are quite subtle. As Erik Hook observed, for instance, the large-scale investments in hydroenergy in Sweden immediately increased the strains on the Water Courts, which deal with the legal problems of water resource management.²⁹ Similar expansive processes, of far greater import, are set in motion when the aspirations for extensive education and health care spread to new strata of the populace.

(7) The Demonstration Effect. This diffusion is significantly accelerated by the news media. The revolution in social communications, particularly the coming of television, has led to increased awareness of the standard of living enjoyed by other segments in society and even in other parts of the world. As a consequence, expectations and pretensions mount, and people get increasingly sensitive to injustices in the distribution of public goods. Through information about services provided in other political units - communes, landsting, or other countries - the citizens and their representatives also know more about what there is to get and can thus better articulate and communicate their views. In this way, the dissemination of inventions will become more rapid.³⁰ The growing concern about 'regional' problems provides an example of this process. Regional redistribution has developed in response to pressures from areas that have become increasingly conscious of their relative deprivation.

C. Consumer Perspective: Political-Institutional Determinants

As the Swedish political structure has remained rather stable over a long time, it is less than probable that changes in this sphere can be brought to bear upon the expansion of government activities. Nor have there been any significant changes in the balance of political forces. But the make-up of the institutions could go some way to explain the differential between Sweden and other countries. Several components of the Swedish political system appear to be favourable to large public consumption. In the terms suggested by Almond and Powell, one could argue that the system has a high 'responsive capability'; decision-makers are likely to lend an ear to claims from many quarters.

(8) Proportional Representation. Under the PR system, a multitude of small groups have their own representatives in the political bodies of local, regional and national government. In the electoral arena, candidates need not necessarily vie for the centre, since they may just as well be returned by a bloc of voters on the right or on the left. This will contribute to a many-faceted representation of interests in the legislatures. In the last instance, of course, only a majority is needed to make decisions, but the very fact that the forming of a majority coalition is postponed to a later stage in the political process may add somewhat to the influence of minority groups.

(9) 'Corporativism'. The prevalence of strong and well-endowed interest organizations and the close cooperation of these organizations with the political parties and the public agencies is a prominent feature in the Swedish system. Because of the frequency of communication and consultation and the presence of organizational spokesmen on many public investigatory committees and agency boards, this pattern has often been said to be 'corporativist'. The tight integration and the plurality of channels between the government and the various interests is likely to promote the transmission of demands and reduce the volume of the demands that is sifted away in the phase of 'political aggregation'.

(10) Consensus Politics. Political strife may intensify the pressure for government action. Torstein Hjellum, in his study of Norwegian local government, found that communes with elections along partisan lines spent more than those not yet politicized.³¹ In certain circumstances, however, consensus might also contribute to the expansion of the public sector. In virtually all Swedish landsting and communes, the leaders of all major parties are simultaneously involved in the executive functions and carry joint responsibility for the main lines of policy. At all levels of the polity, it is common practice for decision-makers to check out the reactions of potential critics before a commitment is made. This procedure of forankring (lit. 'anchoring' a proposal in relevant parts of the power-structure) ensures political actors against post-decision criticism and may make them more responsive to suggestions from various groups of citizens.

(11) Decision Practices. New policies play a lesser role in the growth of public expenditures than old ones that become gradually more expensive. Part of this cost increase is automatic in the sense that the government has already committed itself to provide certain services. Primary education, for instance, is compulsory

for certain age-groups; a 'baby boom' will therefore entail further obligations for the government, and costs will go up without any new consideration of the question. In the judicial institutions, growing demands for justice will lead to more waiting, and waiting will generate demands for more resources. Once a certain right has been granted to all citizens, the level of expenditures will be affected by forces outside the control of the appropriating authorities.

This calls attention to the style of decision-making. In a system where commitments are made for short periods only or where a thorough review process is built into all policies, the political institutions are still in good command of later developments. This is not the case where decisions are 'open-ended' and laden with long-range implications for the public purse. The high proportion of rule-fixed rather than cost-fixed decisions may be one significant factor behind the growth in public expenditures.

D. Financial Perspective: Socio-Economic Determinants

The 'extractive capability' of a political system is closely linked to the structure of production. Large-scale taxation is possible only where there is large-scale economic exchange, and the amount of extractable resources are related both to the mode and to the volume of this exchange. As external trade is easier to tax than domestic trade, a substantial part of government revenue in less developed countries is derived from this source. Urban trade is also easier to tax than rural, which was a reason for certain restrictions on commerce in the past. And finally, a far-reaching division of labour which follows from the expansion of the market economy and the public sector at the expense of home production is an obvious boon to the tax collector. When more goods and services are traded, there is a better opportunity for the government to intervene and intercept a share of the benefits that accrue to the trading partners.

(12) Increase in Economic Exchange. A notable process in the late stage of industrialization is the transfer of many household functions to the market sector and to the public sector. Some services have been taken over by entrepreneurs or government organs that accomplish them with greater efficiency (but perhaps less warmth and human feeling). Others have been turned into commodities. Simultaneously, a large number of women have joined the labour force. The net result of this transformation is a considerable extension of the volume of economic exchange. The ratio of work included in GNP figures to all work performed in society has grown, which is one of the reasons why we have had 'economic growth' - not a misnomer, but a term that conceals the transfer from the invisible' section of the economy to the visible and taxable section. Through this transfer, the government can extract taxes from a larger part of the populace and thus finance an expansion of its services.

E. Financial Perspective: Ideological-Cognitive Determinants

Taxation, however, is not solely a question of objective preconditions. It is also a
matter of consent. Unless people are prepared to give up an increasing part of

their wealth and income, or can be forced by rather subtle means to do so, there is no room for more public spending. Compulsion alone is not a viable technique for financing the government of a post-industrial society. Thus, the public attitudes towards taxation are highly relevant to the issue.

(13) Income-Elastic Attitudes to Taxation. To some extent, the preparedness to pay taxes appears related to the level of income. When subsistence needs have been satisfied, citizens are more likely to part with their surplus resources. The government of an advanced society would then be in a better position than that of a poor nation to raise money for public programs.

(14) The Legitimacy and Efficiency of the Government. A second variable is the prestige of the regime. A state whose sovereignty and identity is well established and whose rulers are generally considered to be representative of the people is likely to have a better 'extractive capability' than one that is not. As Nitti observed, the introduction of democracy might therefore be of some importance for the size of the public sector, and there might well be long-term effects of this transformation that accrue successively as a result of education and social communication. A peaceful political development and a homogeneous national culture could sustain this development towards a greater acceptance of high taxes.

The prestige of a regime is obviously affected by its record. Where citizens believe, rightly or wrongly, that there is a great deal of corruption and inefficiency in government operations, they will hardly be enthusiastic taxpayers. While no systematic comparison has been made, most knowledgeable observers would probably agree that Swedes are less suspicious on this score than, for instance, Americans. Relative mistrust of politicians and bureaucrats appears to be a pertinent variable when it comes to explaining the level of public spending in different political systems.

(15) Doctrines of Government. None of this, however, could account for the significant differential in public spending between such countries as Sweden and Switzerland. A partial explanation for this gap might be found on the level of ideology. Herbert Timm has theorized that the lingering influence of the financial doctrines of laissez-faire was one of the 'lags' that retarded the impact of the economic transformation on government expenditures.³² Surely, wide-spread views of the legitimate borders of collective action must be accorded some weight in the political calculus. Where the government is generally regarded as a due caretaker of most public concerns, taxation may be easier to accomplish than where the functions of regime are circumscribed by deep-seated convictions about civic rights and free enterprise.

F. Financial Perspective: Political-Institutional Determinants

Upgrading the system of tax collection is one way of increasing revenue. If growth in available resources is regarded as an independent variable influencing the scope of government activities, such measures may have some significance. But one may also look to the structure of the fiscal system and to the political techniques used for making decisions on taxation. It seems probable that an

incongruence between the distribution of political influence among the citizens and
the distribution of the tax burden can accelerate or retard the growth of government
expenditures.

(16) Fiscal Decision-Making. Under the Swedish system, three bodies beneath the national level (landsting, communal councils, and church councils) are entitled to levy proportional income taxes upon the citizens. In addition, all kinds of taxes can be imposed by the national parliament. Though the party system provides certain links between these four bodies, it is sometimes argued that this practice of independent decision-making by separate organs, none of which is responsible for the total tax load and some of which - the landsting and the church council - are rarely in the public eye, confers a heavier burden on the tax-payers than would be the case if all fiscal policies were made by one legislature.

Another expansive force is the automatic increases built into progressive taxes. As nominal incomes go up because of inflation, a larger share of a person's real income will be absorbed by the state without any formal decision to this effect. With steep progression and rapid inflation, the government can sit back and collect substantially more revenue each year without ever touching the tax rates.

(17) Revenue Sharing. When expenditure decisions and fiscal decisions are made by separate bodies, the overall volume of government activities tends to increase. Several American studies have confirmed the positive relationship between federal grants-in-aid and state and local spending.³³ While no corresponding studies have been made of the effects of the ample support of the Swedish national government to the landsting and the communes, it can safely be claimed that this separation of fiscal and policy responsibility contributes to the willingness to spend at the regional and local levels. Another interesting feature in the Swedish system which might have had a similar impact was the deductibility of taxes paid to the regional and local authorities on the assessment of taxable income for the national income tax. This provision, which was abolished a few years ago, meant that in raising its own tax rate, a commune or landsting would also enable its citizens to pay a lower national income tax, and less than a hundred per cent of the revenue raised would therefore impact upon the local or regional community. The period in which this roundabout subsidy was at its largest was also a period of rapid growth in commune and landsting expenditures.

G. Producer Perspective: Socio-Economic Determinants

Technological evolution has many repercussions on the supply of public goods. With the acquisition of modem techniques and equipment, the government's provision of services becomes more efficient, sophisticated, and capital-intensive. Emerging economies of scale propel the creation of larger units of production. New professions join the civil service and new skills are employed in carrying out the functions of public administration. Several of the changes noted under section A appear in this sphere also and may affect the production of collective goods.

(18) Idle Manpower. The government is often under pressure to employ more people or to provide idle hands with the means to earn their livelihood. Industrialism, with its periodic recessions, has reinforced this pressure, and novel economic doctrines - on the cognitive level - have lent it political respectability. A greater concern for economic stability and full employment is the outcome of tendencies on both planes - in the objective economic conditions and in the values of the electorate.

Assar Lindbeck has suggested that this demand for government employment or government support for maintenance of particular activities in the private sector tends to increase and differentiate over time.³³ In the welfare state people get more fastidious about their work. They not only require employment but want to hold on to the same kind of employment that they have held before or even to the same job in the same enterprise. Thus, the goal of the labour market policy evolves from stabilization of the whole economy towards stabilization of branches and regional subsectors.

According to another theory there is a growing need for employment and production support because of the tendencies towards secular stagnation inherent in the advanced capitalist economy.³⁴ A Swedish version of this hypothesis advanced by Rudolf Meidner asserts that: (1) employment in Swedish industry is unlikely to increase, (2) more jobs can only be created in the service sector, and that (3) a large part of these must by necessity be financed by the government.''⁵

(19) Rising Costs in Public Production. A suggestion already made by Adolph Wagner, and repeated by many writers since, was that the services provided by the public sector had an inbuilt propensity to become relatively more expensive over time. Wagner's chief explanation of this was the growing capital intensity in defence imposed upon each country by the international arms race. Later theorists have focused on differentials in productivity increase between the private and public sectors, largely due to the greater proportion of services in the latter.³⁶ Part of the problem, of course, may be that efficiency gams in such areas as education or hospital care are not as easy to gauge as in manufacture, but even if some criteria are found, a large number of government activities appear to be difficult to rationalize. To secure stable output, one would therefore need to increase the transfer of resources to the public sector, provided that the wage structure is to be maintained.

H. Producer Perspective: Ideological-Cognitive Determinants

Through communication, citizens get new ideas about what they can demand from their government, but civil servants and government purveyors also get new ideas about what to produce. The extension of contacts and the growth of knowledge on the producer side stands out as an important condition of government expansion. In the private sector, it is scarcely the consumer who initiates new production. The intelligence functions connected with innovation largely reside with the producers, and it is only through their search for profitable combina

tions that the market economy can retain its dynamism. The situation in the
public sector cannot be entirely different.

(20) Professional Norms. Professional excellence is a common human goal. What doctors do is in many respects determined less by patients or administrative superiors than by the views of other doctors. Professionals develop a code of behaviour and a set of values that strongly influence the implementation of many government services and often have expansive effects on the volume of spending. Innovations in the political sphere - 'social inventions' - are frequently disseminated through professional contacts between producers.³⁷ New doctrines and skills disseminated through the educational system may make the younger generation of public producers more enterprising and expensive than the old one.

(21) Demands for Equality. A related source of demands for more appropriations is the civil servants' acquaintance with principles and practices in other parts of the public sector. The 'demonstration effect' is not merely a consumer phenomenon. When civil servants learn about the remuneration and work conditions enjoyed by their colleagues, they demand equal treatment. The standardization of wages and other benefits is likely to shift the total wage level upwards.

(22) Organizational Expansion. For a variety of motives covered in the literature on bureaucracy, people want the organization they are involved in to expand. Organizational growth may create opportunities for personal self-fulfilment, job enlargement, careers, organizational slack, increased security, prestige, jobs for friends and relatives, and a number of other values sought by civil servants. 'Parkinson's Law' is a facetious representation of this desire for more elbowroom in administration, but the desire is real enough, and there is no reason to believe that it declines in post-industrial society.³⁸

I. Producer Perspective: Political-Institutional Factors

The degree to which pressures for more public production have an impact on the level of government expenditures is determined by the routines of politics. The procedures of decision-making may be weighted in favour of certain demands, whereas others are screened away in the process of political aggregation. This has reverberations both on the power balance between producers and consumers in toto and on the competition between various groups on each side. To assess the influence of certain producers, it is vital to examine a variety of attributes such as participation in policy elaboration, access to decision-makers, communication with consumers, authority in public opinion, and indispensability of the public good provided. Shifts in these conditions, or in the activities of the producer groups themselves, may contribute to increasing public expenditures.

(23) Pressure Along Administrative Channels. Plans and decisions taken at the top are often founded and largely dependent on data furnished from below. The annual budgetary process, in which agencies submit their requests for appropriations, is a major channel for bureaucratic demands for more government activities. The expansive effects of this process are to a great extent due to the selective character of the transmitted information. Subordinate organs tend to be

very frank about the new needs they perceive but rather tacit about opportunities for reductions. This is well known, however, and superior authorities have some experience in knocking down the excess demands received from lower levels. What incidence this 'dialectic' will have on the level of government expenditures is influenced by a number of cultural variables, such as the degree of trust, conservatism, and rationalism in the administrative system.

(24) Pressure Along Electoral-Parliamentary Channels. Producers are citizens, too, and in a society with a large public sector they are numerous. With the particular expertise they possess, they often become dedicated opinion-moulders and advocates of more forceful policies in their own fields. When the rank and file of all parties contain substantial portions of public producers, pressure groups for various kinds of government services are built into the political process, and civil servants become an important constituency for the individual representative. In many instances, alliances between producers and consumers may be formed within the framework of the political parties: concerned civil servants and prospective consumers of public services unite to promote a rise in government expenditures.

(25) Insensitivity to Change. Government decisions are made seldom. Even if the budgetary process provides the opportunity for an annual review of all public expenditures, there is rarely an effective examination of ongoing programs, and 'disjointed incrementalism' is the chief principle of action.³⁹ This means that feedback in various forms is normally taken into account only when a large amount of experience has been accumulated. Adopted policies are likely to remain valid for a long time and will not be much affected by less than substantial pieces of new information and less than substantial changes in the environment.

The relative insensitivity to change inherent in political institutions has an immediate bearing on the size of public expenditures. Theoretically, a decisionmaking body may establish the scope of a policy commitment in three different ways:

The decision may stipulate that a certain value of public goods shall be
produced.
The decision may stipulate that a certain volume of public goods shall be
produced.
The decision may stipulate that the value and volume of public goods to
be produced shall be determined by certain rules, which in effect hands
over the decision to somebody else.

A budgetary decision will often contain elements of all these principles, but the crucial question is which of them is likely to prevail if conditions change. What happens, for instance, if the price of the goods goes up? If the first principle has priority, a smaller volume will be purchased by the government. Yet this is not often the case. More frequently, decision-makers will feel more strongly attached to the second or third principle and be prepared to increase their appropriations to meet the new situation. Public policies are more often volume-determined or rule-determined than value-determined.

This has wide-ranging implications for the level of public expenditures. First, it signifies that a drop in relative productivity (treated under 19) will lead to more spending rather than less production. Second, it makes government a rather weak bargaining or trading partner for owners of other factors of production, both labour, commodities, and raw materials. If a higher price is unlikely to lead to decreased purchases on the part of the public buyer, there are few restraints on sellers to keep prices down.

6. Some Concluding Remarks

The above catalogue of potential determinants of rises in government spending is by no means an exhaustive list but merely a set of examples to illustrate the diverse approaches that may be chosen in an effort to explain public growth. If the multitude appears staggering, it should be borne in mind that a truly thorough inventory of related factors would be far more extensive. Subdividing the nine fields in our diagram, we might easily arrive at fifty or seventy-five different processes or events that advance increase in collective expenditures. A general theory of government growth does not seem to be around the corner.

It is doubtful, however, whether one could ever reach comprehension of the phenomenon by taking so many circumstances into consideration. Unless crucial variables are identified, the complexity of the process might even deter from further attempts to explain and predict trends in public policy and public spending. By way of conclusion, I shall therefore suggest some areas that appear to be fertile and worthy of further research. In doing so, I take into account that most studies thus far have applied the consumer perspective. To complement this picture we need now turn our attention in other directions.

First of all, to the problems of financial structure and financial psychology. The fine art of taxation is evidently a fundamental precondition for collective activities, and it is only governments that have mastered it and practise it with skill that can ever maintain a large-scale public sector. The technology of resource transfer between different sectors of the economy is a largely unexplored field, and one that appears quite promising for comparative research as well as for constructive thinking. It is closely related to such key issues of development theory as the achievement of political legitimacy and the attainment of greater costefficiency in the carrying out of public functions.

Secondly, we should pay more attention to the problem of bureaucratic politics in its widest sense. Although the political role of administrators is now widely acknowledged and much good work has been done on the relationship between politicans and bureaucrats, the essential part of the producers in policy determination has not yet been sufficiently explored. A good deal remains to be learned about the practices and principles of the budgetary process and about the extracurricular' activities of bureaucrats in the political arena.

An important domain is finally the routines of decision-making, the procedures
of evaluation and review, and the very structure of the political decisions, which

determine, among other things, the extent to which government activities are affected by changes in the environment. These questions belong to the heartland of political analysis, but it is nonetheless true that little has been done to generate abstract theory and tenable generalizations. A mass of scattered observations are available for sifting and synthesis. This article has suggested one way in which they might be organized.

---

NOTES

1. Cf., my Beyond the State: The Future Polity in Classical and Soviet Marxism (Uddevalla: Laromedelsforlagen, 1972), esp. chapter 1.

2. The 'academic socialists' were socialists only in the sense that they took great interest in the 'social question'. For summaries of Wagner's views, see H. Timm, 'Das Gesetz der wachsenden Staatsausgaben', Finanzarchiv XXI (1961), pp. 201-247; or M. Bird, Wagner's "Law" of Expanding State Activity', Public Finance/Finances Publiques XXVI (1971), pp. 1-26; V. P. Gandhi, 'Wagner's Law of Public Expenditure: Do Recent Cross Studies Confirm it?', Ibid., pp. 44-56, lists five different interpretations of Wagner's Law.

3. F. S. Nitti, Principi di scienza delle finanze (1903), reprinted in translation in C. J. Bullock, ed., Selected Readings in Public Finance, 2nd ed. (Cambridge, Mass.: Ginn and Company, 1920), pp. 33-47.

4. A. Brecht, 'Internationaler Vergleich der offentlichen Ausgaben', Grundfragen der Internationalen Politik 11, Leipzig & Berlin, 1932.

5. H. Timm, op. cit., pp. 234-242. Cf., G. Schmolders, Finanzpolitik, 3rd ed. (Heidelberg: Springer Verlag, 1970), pp. 187-196.

6. K. Deutsch, 'Social Mobilization and Political Development', American Political Science Review, Sept. 1961, pp. 493-514.

7. E. Hook, Den offentliga sektorns expansion. En studie av de offentliga civila utgifternas utveckling åren 1913-58 (Uppsala: Almqvist & Wicksell, 1962).

8. S. Fabricant, The Trend of Government Activity in the United States since 1900 (New York: National Bureau of Economic Research, Inc., 1952).

9. H. E. Brazer, City Expenditures in the United States (New York: National Bureau of Economic Research, Inc., 1959); for a list of other American studies, see F. L. Pryor, Public Expenditures in Communist and Capitalist Nations (London: George Allen and Unwin Ltd., 1968), pp. 448-451.

10. For references to this literature, see I. Sharkansky, Spending in the American States (Chicago: Rand McNally & Company), pp. 11, 145 ff.

11. T. R. Dye, Politics, Economics and the Public: Policy Outcomes in American States (New York: Rand McNally, 1967); on criticisms of Dye's views, see pp. 282-301.

12. Bibliographies in Pryor, op. cit., pp. 446-448; Timm, op. cit., pp. 243-247; Cf., also Bird, op. cit., p. 1.

13. R. W. Crowley, 'Long Swings in the Role of Government: An Analysis of Wars and Government Expenditures in Western Europe since the Eleventh Century', Public Finance! Finances Publiques XXVI (1971), pp. 25-43.

14. I. J. Goffman & D. J. Mahar, The Growth of Public Expenditures in Selected Developing Nations: Six Caribbean Countries 1940-1965', Ibid., pp. 57-74.

15. Pryor, op. cit., pp. 280-312.

16. World Economic Survey. 1967 (New York: United Nations. 1968). DD. 86 ff.

17. S. Lall, 'A Note on Government Expenditures in Developing Countries", Economic Journal LXXIX (1969), pp. 413-417.

18. Bernard P. Herber, Modern Public Finance (Homewood, 111.: Richard D. Irwin, Inc., 1967), Cf., Bird, op. cit., pp. 17-18.

19. S. P. Gupta, 'Public Expenditure and Economic Development - A Cross-Section Analysis', Finanzarchiv XXVIII (1969), pp. 26-41; A different perspective may be found in R. Zenker, 'Den offentliga sektorns tillvaxt', Erik Lundberg et al., Svensk finanspolitik i teori och praktik (Lund: Aldus, 1971), pp. 76-103.

---

20. A. T. Peacock & J. Wiseman, The Growth of Public Expenditure in the United Kingdom (Princeton: Princeton University Press, 1961).

21. S. P. Gupta, op. cit.

22. H. T. Oshima, 'Share of Government in Gross National Product for Various Countries', American Economic Review, XLVII (1957), pp. 381-390.

23. G. Almond & J. B. Powell, Comparative Politics: A Developmental Approach (Boston and Toronto: Little, Brown and Company, 1966), pp. 190-212. For an extensive discussion of related problems, cf. A. Musgrave, Fiscal Systems (New Haven: Yale University Press, 1969), and H. H. Hinrichs, A General Theory of Tax Structure Change During Economic Development (Cambridge: International Tax Program, Harvard Law School, 1966).

24. I. Sharkansky, op. cit., p. 146; S. Sachs & R. Harris, 'The Determinants of State and Local Government Expenditures and Intergovernmental Flows of Funds', National Tax Journal, 1964:1, pp. 75-85.

25. P. A. Baran & P. M. Sweezy, Monopoly Capital (Bungay: Penguin, 1966).

26. W. J. Baumol & W. E. Gates, 'Kostnadssjukan inom tjanstesektorn och levnadsstandarden', Skandinaviska Bankens Kvartalsskrift, 1972, pp. 42-52, C. Åberg, 'Den offentliga sektorns kostnadsutveckling', Ekonomisk Debatt 1973, pp. 491-496.

27. C. N. Parkinson, Parkinson's Law and Other Studies in Administration (Boston: Houghton Mifflin, 1957); W. A. Niskanen, Bureaucracy and Representative Government (Chicago & New York: Aldine, Atherton, 1971); Jørgen S. Dich, Den herskende klasse (Copenhagen: Borgen, 1973).

28. W. H. Starbuch, 'Organizational Growth and Development', J. G. March, ed., Hand' book of Organizations (Chicago: Rand McNally, 1965), pp. 451-522; A Downs, Inside Bureaucracy (Boston: Little, Brown and Company, 1967), chapter 2. For a further discussion of producer pressures, see Daniel Tarschys & Maud Edwards, Petita: Hur Svenska myndigheter argumenterar for hogre anslag (Uddevalla: Liber, 1975).

29. Hook, op. cit., chapter 5.

30. Cf., J. Walker, 'The Diffusion of Innovations Among the American States', American Political Science Review, LXII (1969), pp. 880-899.

31. T. Hjellum, The Politicization of Local Government: Rates of Change, Conditioning Factors, Effects on Political Culture', Scandinavian Political Studies. 2/1967, pp. 69-93.

32. Timm, loc. cit.

33. A. Lindbeck, 'Internationella ekonomiska konflikter i en expanderande vårldsekonomi', Ekonomisk Debatt, 1974, pp. 383-394.

34. Baran & Sweezy, op. cit.

35. R. Meidner, 'Utvecklingen av de offentliga tjånsterna och dess betydelse ur facklig synpunkt', Anfdrande vid Public Service Internationals kongress i New York den 10 oktober 1973 (mimeo).

36. Cf. note 26.

37. Cf. note 30.

38. Cf. note 28.

39. D. Braybrooke & C. E. Lindblom, A Stategy of Decision: Policy Evaluation as a Social Process (New York: The Free Press, 1963), chapter 5.