Resumé

The purpose of this article is to answer the following question: When did Swedish bureaucracy arise? That is, to determine a point in time when the organizational technology that Weber called bureaucracy became dominant in the process of Swedish state formation, as well as when bureaucracy finally replaced the feudal form of government based on the authority of nobility and the hierarchical ties of individual loyalties. The importance of this question relates to the debate on the sequential logic of economic and political development. The change from an aristocratic/particularistic to a bureaucratic/universal state apparatus can be understood as a change from despotic to infrastructural state power. Most empirical material indicates that, in terms of institutional structure, the transition to a bureaucratic administration started in the 1850s. Contrary to what most Swedish historians have argued, the Swedish state remained feudal and particularistic all they way up to the mid-19th century. If any particular decade is key to this transformation, it would be the 1870s. By then the last of the noble privileges had disappeared, a uniform salary system had been introduced, and the various state apparatuses had begun reorganizing toward a higher level of efficiency and rationality.

Introduction

When did Swedish bureaucracy arise? With this question I seek to determine a point in time when the organizational technology that Max Weber called bureaucracy became dominant in the process of Swedish state formation, as well as when bureaucracy finally replaced feudalism as a form of government based on a combination of the authority of nobility and the hierarchical ties of individual loyalties.

The more specific characteristics of bureaucracy as the means for organizing state activities will be examined later in the article. For now, it suffices to define bureaucracy as an organizational technology for (primarily) state activities, which bases itself on a distinct form of specialized knowledge (in German: Fachwissen). This type of knowledge is used to make decisions in

---

* Bo Rothstein, Department of Political Science, Göteborg University, Box 711, SE-405 30 Göteborg, Sweden, E-mail: Bo.Rothstein@pol.gu.se

accordance with previously established and calculable universal rules without regard for the individual; and it is an ingredient in a distinctive blend of specialization and hierarchy based on an impersonal and formal rationality (cf. Weber 1922/78; Tilly 1990; Therborn 1989; Frohnert 1993).

It should be clear that my purpose is not to determine a point in time when the country established an administrative center, nor when a caste of royal servants arose that performed administrative rather than purely military functions. The point of regarding the concept of bureaucracy as a distinct form of administrative activity resides in its relation to capitalism as an economic system. In other words, bureaucracy can be viewed as the unique administrative institution that (together with parliamentary politics) arises in association with the market economy and, consequently, as the unique class-based character of the capitalist state. To quote Weber: "predictability, continuity and objectivity of legal order are all essential for ... capitalism on a large scale" (Weber 1978, 1095). The importance of this institution for the development of a capitalist economy and economic prosperity has been argued by a number of contemporary social theorists, e.g., Therborn (1980), North (1990) and Rosenberg & Birdzell (1986).

The importance of this question relates to the debate on the sequential logic of economic and political development. The change from an aristocratic/ particularistic to a bureaucratic/universal state apparatus can be understood as a change from despotic to infrastructural state power (Mann 1988). The latter type of power is a qualitatively different type of power which is vital for establishing as well as maintaining a modern market economy (Blomkvist 1992). Is the political development of, in this case, a universal nonparticularistic state apparatus driven by economic necessities in some sort of functional logic as traditional Marxism would argue? Or is it the other way around, i.e., is it changes in the state machinery that spur economic development in a capitalist direction? (Weiss & Hobson 1995, ch. 1). Or was, perhaps, the rise of the modern bureaucracy an unintended consequence of a specific technological development in the military forces that took place in the West, as argued by Tilly (1990, 25-36)?

Be that as it may, if the hypotheses developed primarily by Therborn (1980) and North (1990) mentioned above have any validity, a Swedish state organized in one way (feudal/particularistic) can be identified before a specific point in time, and a state organized in another way (bureaucratic/ universal) after that point. Furthermore, the difference and the transition between these two types of states would be qualitative in the Marxist sense of the term. The timing of this development is important for an understanding of the relation between state formation, administrative institutions and economic change.

The trouble with empirical research on Sweden is that we do not really
know when the economy became primarily capitalistically organized. The

country started industrializing, quickly and belatedly, from around 1870; but long before then there was a substantial amount of business capital for example in the cities of Gothenburg and Stockholm (the so-called Skeppsbroadeln or the harbor nobility). The industries of the time presumably occupied a middle ground between feudalism and capitalism, since they were for the most part oriented toward the production of goods for sale in the marketplace, while the organization of production itself had strong feudal characteristics (Jb'rberg 1961; Schon 1981). There are, however, arguments for saying that a modern capitalist market economy started around 1860-1870. The feudal guild system was abandoned by the four-estate Riksdag (the Swedish Parliament)¹ in 1864 when the principle of free tråde and commerce was established. It was also in this period that the Stockholm stock exchange, banks and the legal system for joint-stock companies were established. Another important change is the decisions between 1855 and 1869 to eliminate land-based property taxes, which meant that the barter economy was finally replaced by the money economy in the finances of the state.

On Problems in Writing History

In his introduction to The English Administrative System 1780-1870, Sir Norman Chester writes that one of his motivations to write the book was limitations of existing literature on the subject. He is especially critical of the tendency in studies of the history of administrative institutions to explain the past in terms of the present. Sir Chester warns us that terms such as "civil service," "central or local administration" and "ministerial," for example, either were not used or had an entirely different meaning during the period being studied, and that using them carelessly in their current sense would paint the wrong picture of the historical reality of the time (Chester 1981). Yet, this is often what happens when the history of the Swedish state apparatuses is presented. The military efforts during the 17th century, when Sweden under Gustavus Adolphus became a major European power, necessitated the construction of a strong, and in many ways impressive, administrative center under the famous Chancellor Axel Oxenstierna (ef. Carlsson 1913; Eden 1902; Nilsson 1988). Many Swedish historians and political scientists have uncritically taken the establishment of the collegial system of administration during the 17th century as the starting point for the administrative formation of the modern Swedish state (see, e.g., Asker 1993; Nilsson 1993; Molin et al. 1979, 27-30). As I will show below, this is misleading for exactly the reasons given by Sir Chester, i.e., even if some administrative concepts and terms have remained until this day, they had a totally different meaning and function during the 17th and 18th centuries.

The prominent Swedish historian Per Nyström has described attempts by historiographers to create continuities in history and cites as an example efforts to trace the rise of Sweden's parliament further and further back in time (Nyström 1974). As Sunesson puts it, this is done to "make tradition more beautiful by extending it." Sunesson argues that,

The administrative apparatus changes at decisive points during the transition to a capitalist mode of production and societal structures of capitalism. To describe this transformation as if the current administration had developed from an embryonic state that existed four hundred years ago is ambiguous and misleading. Yet, this is exactly what occurs (Sunesson 1981, 59f).

With the publication in 1985 of the path-breaking volume Bringing the State Back In (Evans et al. 1985), political scientists became more aware of the shortcomings of the ahistorical view of the state and of the dominant society-centered and functional view of the state-society relation. In The State as a Problem, Swedish political scientist Mats Dahlkvist (1982) demonstrated very convincingly how Easton's systems theory, which dominated Anglo - and thereby Swedish - political science in the 1960s and 19705, led to this ahistorical, precipitous and unproblematic view of the state. Dahlkvist's main criticisms of Easton concerned his construction of political concepts which aimed at reaching beyond historical boundaries, but instead was founded on 20th century Western capitalism. Under the heading "The State As Uncharted Territory in Anglo-Saxon Political Science," Dahlkvist stated that:

The historical reality of the modern state is left unexplored, likewise its social significance, that is, the societal role it has played during the course of its existence. Consequently, whether this role has changed and what aspects this role has encompassed from (in the case of Sweden) Gustavus I (Vasa) to Carolus Gustavus XVI is a question that never arises (Dahlkvist 1982, 27).

Dahlkvist also focused precisely on the types of difficulties that the crude application of Easton's systems theory created in studies of various feudal state structures. The scope of the current study does not allow for the ambitions that Dahlkvist recommends in studies of the state. It is nevertheless my intention to demonstrate if not why, then at least when the role of the Swedish state underwent the transformation to a bureaucratic and universal mode of operation.

Swedish sociologist Sune Sunesson (1981) has greatly contributed to our understanding of the Swedish state during the 17th and 18th centuries. He argues that the organization of the 17th century Swedish state was in fact genuinely feudal in nature. To illustrate, he presents, among other things, the example of the enfeoffment of the Swedish Postal Service to Magnus Gabriel de la Gardie in 1677. Another contributor is Peter Englund (1993), who has shown that particularistic patron-client relations were abundant in the

Swedish state during this period. In his major work Lineages of the Absolutist State, Perry Anderson (1974) effectively pursues his thesis on the feudal nature of absolutist states, and in the chapter on Sweden he cites the regime of Gustavus 111 during the late 18th century as an example.

The Feudal and Capitalist State

Therborn has introduced a state model which suggests that states, just like other organizations, be studied as a function of ongoing social processes and not as goal-oriented subjects. His model focuses on the nature of the factors that constitute an organized activity. These factors are personnel, material resources and tasks, and they are studied according to how they flow into, transform and flow out of the organization/the state. Regarding the personnel factor, for example, it is their recruitment and promotion, position within the organization, relationship to colleagues in other positions, and their relationship to other members of their social group that are specified in the model. The concept of organizational technology is introduced to explain how these factors are ideally regulated under different modes of production. The dominant organizational technology of the feudal state is described as aristocratic, the capitalist state as bureaucratic (Therborn 1980, ch. 2).

Since my objective is not to analyze the concrete historical development of the Swedish state, but rather to determine a point in time, I shall preserve the fundamental approach of Therborn's model but at the same time disaggregate his model into a somewhat simpler form. Five different factors will be analyzed. The theoretical attributes of each factor will be presented at the beginning of the relevant empirical section. First, however, I shall present the following general attributes.

At the end of the 19th century when Weber recorded the features of modern bureaucracy, he clearly did so having perceived it as something new. Administrative centers, often large and complex, had existed for a long time, but the strict separation of individual and post, the professionalization of state activities and the firmly rooted hierarchy of bureaucracy were all novel phenomena. Modern bureaucracy was, above all, founded on a money-based economy. As Hviid-Nielsen (1977) has demonstrated, Weber saw modern bureaucracy as the unique political order of capitalist society, and later, after 1917, together with democratic parliamentarism, as its guarantor.

At a general level, Therborn describes the changes as follows: The struggle of the rising bourgeoisie centered on issues concerning both the state and society. It demanded that the state be separated from the private realm of aristocratic lineage and be firmly based on "public opinion"; it should be concerned only with matters subsumable under general principles, and not with the material or legal interests of particular individuals or categories of

individuals. The distinction was intrinsically related to the concept of bureaucracy - of the bureau as a public office separated from the household of the king and the aristocrats. This was accompanied by the strict differentiation between private matters and public activities, which was an historical innovation. Especially the de-personalization of the public administration (the "public") was an important change (Therborn 1980, 66).

Birth, Recruitment and Promotion

Under feudalism, birth and later even purchase of positions were normal avenues to acquiring public office in Sweden as elsewhere. Being of noble birth was primarily a requirement that applied to high-level civil service posts and officer positions. There were two related key factors in the organizational technology of the feudal state: individual loyalty and subordination between the sovereign and his vassal. The prevalent mode of integration of the feudal nobility into the Absolutist State in the West took the form of acquisition of "offices." He who privately purchased a position in the public apparatus of the State could then recoup his expenses by licensed privileges and what would be called corruption today, in a kind of monetarized caricature of investiture in a fief (Anderson 1974, 34-36). In contrast, the recruitment system of bureaucracy is strictly based on individual intellectual merits, where formal qualifications in the form of university degrees and prior experience are the deciding factors in individual cases.

The number of commoners in the lower levels of public administration in Sweden was not insignificant during the eighteenth century. As early as 1720, the Constitution of the same year stipulated that experience and merit were the only factors of importance and that nobody was to be displaced due to inferior status or birth. As so many other laws in feudal society, this came to have little, if any, significance. Throughout the entire "Age of Freedom" (a period of semi-liberal rule 1720-1789), all offices were, in practice, filled by the nobility, from the highest ranks and down, and within both the military and the civil service. However, a substantial number of commoners were ennobled during the 18th century. This, among other things, in 1762, led to the refusal of the House of the Nobility to admit any new noble families until the existing number declined. The widespread ennoblement of commoners was probably due in part to the fact that on an international scale the Swedish nobility was relatively small in number, and in part to the fact that the nobility was not able to recover, numerically or statuswise, after Sweden's defeat in the Great Northern War. The ennoblement of commoners can also be viewed as a sign of the dissolution of the political order based on the four estates and that an incipient merit-based recruitment system had taken root within public administration. This interpretation is, however, doubtful, partly because, as

Sunesson (1981) has demonstrated, a significant economic, cultural and social fusion was taking place between the nobility of birth and the civil service nobility; and partly because the sons of nobility had significant advantages in their civil service careers: they could start out at a high level at a very early age and thereby had a greater chance of reaching high offices (Carlsson 1949, 42).

Constitutionally, 1789 became an important year in that the Act of Union and Security issued by Gustavus 111, which replaced the semi-liberal Constitution of 1720, stipulated that the "highest and most distinguished" offices would be reserved for the nobility, but that otherwise estate would not be taken into consideration. The break with nobility initiated by Gustavus 111 had a significant influence on the recruitment of commoners to public administration, albeit only to low offices. When the king died in 1792, there was only one commoner in the high-ranking body of civil servants. Between 1792 and 1820, the growth in the number of commoners in public service came to a standstill (ibid., 57).

The new Constitution of 1809 removed almost all preferential treatment of nobles within public administration. The last statute to uphold noble priorities, stipulating that half of the members of the Supreme Court must belong to the House of the Nobility, did not disappear until 1845 (Carlsson & Rosén 1961, 304). However, the Constitution of 1809 was not, as often claimed, the decisive turning point in the development of modern bureaucracy. Already in 1800, Anders Håkansson was the first commoner to be appointed county governor, which was, of course, against regulations. He was, however, ennobled in 1801. The first county governor to remain a commoner was P. E. Georgii, appointed in 1802. The provisions set forth in the Act of Union and Security had thus already become obsolete before the Constitution of 1809. The first commoner to become cabinet minister was appointed in 1828, the first to become prime minister in 1885, and the first to become minister of foreign affairs in 1914 (ibid., 304).

Carlsson identifies 1840-65 as the period when the last formal privileges of the nobility disappeared within public administration. This mainly concerned the nobleman's possibility, as compared to the commoner's, to start his career both higher up in the hierarchy and earlier in life (Carlsson 1949,59). Another historian estimates that equality was formally established around 1865 (Elmroth 1962). This, of course, did not stop the nobility from, even later, remaining disproportionately represented within the administrative apparatus through wealth, family traditions, as well as by other means. Even around the turn of the century, the nobility held approximately 25 percent of high offices at colleges, and 11 of the 25 county governorships (ibid.; cf. Frohnert 1993, 287).²

Position Purchasing

Direct sale of offices to reinforce the treasury, which was normal practice in France, does not seem to have been common in the central Swedish administration during the 18th and 19th centuries (Kettering 1986). On the other hand, the state avoided paying pensions to civil servants thanks to the so-called ackordssystemet (accord system) which was übiquitous and completely dominating in the armed forces (Myrberg 1922; Frohnert 1993, 288).

The accord system allowed civil servants who wanted to advance in their careers to persuade higher-ranking civil servants to resign their offices by paying them an accord in the form of a certain sum of money (Frohnert 1993, 287). The higher-ranking civil servants could then use the money to purchase new positions, or they could use it as pension. The system worked, partly because an effective pension system for civil servants was not established until after the 1809-10 parliamentary assembly (Riksdagen). But it also worked because there was no upper age limit for when a civil servant had to leave his post. Neither illness nor any other gross inability to fulfill one's duties were valid grounds for taking a public office away from someone. This was a consequence of public office being regarded as the officeholder's property comparable to land and goods (Frohnert 1993; cf. Westerhult 1965, 149-52).

During the 18th and 19th centuries, a number of royal prohibitions were issued that addressed this position-purchasing system (Myrberg 1922). The fact that they hardly made an impact provides insight into the legislative system of feudal society (Frohnert 1993, 287). Proposals to introduce criminal laws against purchasing positions were rejected by Riksdagen of 1828-30 (Rabenius 1896, second part, § 108). In order to protect those civil servants who had at one time paid large sums of money for their positions, prohibitions were not accompanied by penalties. An effective prohibition would have entailed the loss of large investments that had been made in offices.

Ultimately, the system was abolished. First, resignation regulations were gradually introduced. The right, but not the obligation, to resign from office at 70 years of age had first been instituted in the armed forces in 1774; one hundred years later, in 1879, similar regulations were introduced in the civil service requiring civil servants who had reached 65 years of age and who had been in service for 35 years to retire, but with the possibility of deferment if the party concerned was still considered to "serve the public good" (ibid.; Westerhult 1965, 149-52).

Second, in addition to Riksdagen's establishment of a pension system,
special state funds were introduced during the 1830s and '40s to compensate
retiring civil servants for the accord they had paid to attain their position.

Furthermore, civil servants who filled these positions in the future would have to sign a sworn personal pledge that they would not demand accord from their successors (Rabenius 1896; Myrberg 1922). The last order against the accord system was issued for county governorships in 1852, where accords had been particularly high, and it was not until 1886 that a sworn assurance on the part of the civil servant that he would not accept accord within the civil service was no longer required (Myrberg 1922, 6).

In the army, the accord system had been especially common. During the second half of the eighteenth century and the early nineteenth century it was almost universally employed. The armed forces were also more open about dealing in positions (ibid., 44). Efforts to abolish this system did not begin until the 1820s. This in turn led to a new order in 1833, resembling the one that would later develop within the civil service with the aforementioned sworn personal pledges and the government amortizing the accord that had been paid (ibid.).

The abolishment of the accord system reflects some interesting organizational characteristics of the feudal state. The public prohibitions had very little impact on the behavior of civil servants. Not until civil servants provided sworn personal declarations against taking accord could the system be broken, and then only when the state acted as intermediary by compensating those who had at one time paid accord for a position. This underscores how, under feudalism, public office was regarded as personal property that could not be confiscated without compensation. Not even if a civil servant was too ill to fulfill his duties could he be dismissed. Being personal property, public office was also something the officeholder had the prerogative to neglect. Under such conditions, it was not unusual for the formal officeholder to delegate the actual duties of the position to another person, who would then receive a portion of the salary, but also, and more importantly, the income associated with and generated by the position itself (see below).

Education and Duties

In contrast to the organizational technology of the feudal state, which was built on a system of personal service and loyalty between superiors and subordinates, the distinguishing feature of bureaucracy is its call for intellectual capability and knowledge of the rules on the part of government personnel. The feudal civil servant required no special training, since his job was to interpret customary law and natural law. For this, the universal authority of nobility and a noble upbringing were sufficient. The growing commerce and sale of goods in the market place revived the Roman legal system, characterized not only by its strong emphasis on the right to property, but also by its traditions of equity, its rational canons of evidence, and its

emphasis on a professional judiciary (Anderson 1974,27). Therefore, in order to create a modern bureaucracy, a connection had to be established between higher education and public employment. The most important entrance qualification to the civil service became a university degree in law.

During the 19th century, Sweden had only two universities, one in Lund and one in Uppsala. In his history of Uppsala University, Lindroth describes the education at the School of Law as being stuck in a veritable intellectual as well as organizational morass from the mid-18th century and up to the first decades of the 19th century. The same was true at the University of Lund (Lindroth 1971, 163-66; Lunds universitets historia 1971, 224). In 1797, the chancellery college complained to the governing board of the University of Lund:

we must recognize that the young men who seek entrance to the chancellery offices do not
possess the knowledge in science and language necessary for a chancery subject,
notwithstanding they were furnished with academic qualifications (ibid.).

This slump pertained not least to the so-called ämbetsexamina (degrees qualifying for high civil service posts) which served as recruitment instruments for employment in the central administration. In 1859, Olivercrona wrote in his historical account of the legal education at the Uppsala University School of Law from 1785-1823:

during the long period when Hemberg, Lundström and Drissel occupied the prominent juris patrii profession, the purely legal studies sank to their lowest point of ruin. The so-called Hoffrättsexamen became insignificant, the Bachelor of Law degree was implemented with the highest degree of ease, and in the study of Roman jurisprudence, even the most cursory knowledge was not required (emphasis added; Olivercrona 1859, 14).

In their 1828 report, the famous Genius Committee, (a committee of learned scholars formed by the government to investigate the educational system), described one of the civil service degrees, the kameralexamen (finance degree), with which the very highest positions in the state apparatus could be reached, as pure parody. In Uppsala Universitetet 1477-1977, Lindroth writes that the civil service degrees were considered mediocre and that those who studied for them were regarded as second-class students (Lindroth 1971,165). Frohnert's (1993) detailed study of the local tax administration also shows a lack of formal education requirements for local civil servants during the 18th and early 19th centuries. The Crown did not reward academic degrees when such local positions were filled, apparently because tax collection rules and practice differed among counties (Frohnert 1993, 165). This is of course an indication that "general rules" in a Weberian sense did not play a significant role, even in such a crucial matter as taxes.

In response to the criticism against the quality of the graduates' education,
the Chancellor of the University of Lund explained in 1797 that the most
important reason for this unfortunate situation was that adolescents in their

early teens, who for understandable reasons could not benefit from the
education, were being enrolled at the universities. This, in turn, was a
consequence of the faet that

parents and relatives, whose main goal is only to hasten the early entrance of their children
to a civil service career, either out of conceit or ignorance, . . . build their hopes of future
advancement more on wealth and privileged connections than on duly founded ability
through hard-earned studies (Lunds universitets historia 1971, 238).

Since the principle of seniority was the most important for career advancement, it was of vital importance to gain entrance for one's offspring to a department where he could be employed already as a child; naturally without being required to work or receiving compensation.

Eventually rules were introduced that established guidelines for minimum time of study, as well as age requirements for employment in civil service departments. In 1812, the minimum age for entering the civil service was 18, and in 1828 it was 21, based on the Training Committee's recommendations; an age limit of 18 was also required for university enrollment (ibid., 244). Nevertheless, even in 1831, in response to a letter regarding the education of future civil servants, Professor Holmbergsson at Lund University stated that education was not going to be improved by establishing regulations for minimum time of study and age limit requirements. Holmbergsson proposed instead to change the promotion system of governmental departments so that merits and years of service, rather than personal relationships, were the deciding factors for career advancement (ibid.).

The question of whether legal schooling had any relation or relevance to future employment in public administration became a heated issue in Swedish politics from the 1820s to the 1860s. Liberal and radical forces inside and outside of Riksdagen together with the newspaper Aftonbladet, among others, even demanded that the program at Uppsala be discontinued and that it be established at Stockholm University so that it could, both geographically and even by virtue of its contents, be closer to practical administrative activities.

This battle was of course related to the one being fought over the nature of academic education in general. Here the contenders were the old educational ideals of romantic classicism and those of liberalism and progressivism which demanded rationality and utility in higher education. The abolishment in the 1860s of Latin as an academic language symbolized the victory of the latter (Lindroth 1971, 158).

The quality of legal education for civil servants and its relevance to future administrative employment improved dramatically in the first half of the 19th century. The number of professorships at the law schools experienced a sharp increase - from one to four at Lund and from two to five at Uppsala (Lindroth 1976, 159-61; Lunds universitets historia 1971, 230). An 1863 reform of the law degrees also established a firm relation between education and employment.

It is interesting that these changes coincided in the 1860s with the state assuming financial responsibility for university management and salaries. The universities had formerly functioned as financially independent institutions with revenues from fiefs.³

In terms of content, Riksdagen demanded that students who graduated to the civil service must receive instruction in economics, administrative and financial legislation, and in Swedish constitutional law. This was motivated by rationalist principles; accordingly, it was reasonable to expect that students who were going to be in charge of the country's administrative operation would, in the course of their education, be imparted knowledge with a certain degree of relevance to those activities. Revealingly, no coherent systematic proposals regarding the Swedish administrative apparatus came forth until 1866-73 (Rabenius 1866-1873).

Work and Pay

In the feudal state there was no direct relation between the duties of a post and its pay. The feudal civil servant lived primarily on income from land. Every office usually had a number of manors tied to it, and the owners had to pay taxes in kind. The officeholder also received a large portion of his income from the various duties of his post, based on a system that nowadays would qualify as bribery or corruption. On the other hand, he himself often had to assume expenses in connection with the performance of duties related to his office, such as travel, assistants, materials and premises. Also, the same civil servant could hold several posts simultaneously and a strict work distribution among the various posts and/or various state apparatuses did not exist. Again, the best way to understand this is to remember that public office was considered property/fiefs.

Bureaucracy presupposes a direct relation between the position and pay of a civil servant, based on established rules and in monetary terms. A full-time position normally demands all of the civil servant's work time and holding two positions is not tolerated. It further presupposes that individual and post, as well as private income and income derived from the post are strictly separate. In contrast to his feudal colleague, who held his public office around the clock, the bureaucrat has no power or authority whatsoever outside of the job, where he is regarded as a private person. The feudal outlook on public office also entailed that its duties could not be altered without a new personal agreement between the civil servant and his lord (which in turn led to new benefits). The bureaucrat, in contrast, is part of a strict hierarchy where job duties are decided by a superior.

During the first half of the nineteenth century, the public administration
pay system was rather mixed. Even though pay in kind was most common,

monetary remuneration did exist. Perquisites and fees for job-related services (henceforth, service fees), or bribes by today's standards, were abundant (Cavallin 1996). It was also common for civil servants to receive income from land and residences that belonged to the office (Westerhult 1965, 107-23; Rabenius 1866, 324-30; Kammarkollegiets historia 1941, 274-78). Civil servants were often appointed, especially within the higher administration and the universities, in the absence of sufficient funds to pay them, and without the expectation that they were going to perform any work (Lunds universitets historia 1971, 25-27). Income for the very same job often varied. For example, in the local administration, income for the same job was "so varying that many of the best-paid civil servants had twice the pay of their less fortunate colleagues" (Westerhult 1965, 197). Likewise, Frohnert's study of 18th century local bailiffs shows that "through indelningsverket (the allotment system), a large portion of the bailiffs' income was tied to individual peasants who were obliged to deliver grain, money or other goods" (Frohnert 1993, 367).

From 1809 to 1880, one of the most debated topics in Riksdagen was the introduction of a modern, uniform salary system (ibid., 277). From 1809, the value of money dropped drastically, which hit salaried civil servants quite hard. In an 1811 letter to His Majesty the King, the directors of the Exchequer Board pointed out that since the pay had fallen to such a low level, a large number of civil servants had taken jobs in other national boards or agencies or with county governments (all the while retaining their old positions, of course), and that this had considerably complicated and delayed work in the agency (Kammarkollegiets historia 1941, 278). Holding two posts was very common and the system was not prohibited until 1879.⁴ In a 1822 report, the departmental committee appointed in 1819 declared that poor pay forced civil servants who did not have private fortunes to either look for other posts and public office or gain their livelihood through private business affairs.³

Both civil servants employed in the central government and at the county level were paid in full or in part through their direct ties to private manors and farms, which were required to deliver determined amounts of primary products to the civil servant every year. During the mid-19th century, this tax system was very complicated with 295 different persedlar (articles) in the various tax codes of the country (Westerhult 1965, 120-26). The private taxpayer and the civil servant could agree upon a monetary tax instead of a tax paid in kind. However, the civil servants could always demand to receive products instead of cash.

Whenever a citizen contacted a civil servant with a business matter, he had to pay him a fee for his services. Such perquisites and service fees were regulated by government ordinance (Rabenius 1866, 120-23). For some civil servants, such as local sheriffs and bailiffs, these incomes were often higher than their ordinary salary.

The pay-in-kind system was gradually abolished in the civil service between 1855 and 1870. In 1855, the state began to liquidate in-kind payment and instead paid the civil servants the corresponding salaries. The same year, salaires replaced the average worth of perquisite and residence incomes (Westerhult 1965, 127-29; Frohnert 1993, 287; Cavallin 1996, 7).

The introduction of a new tax system was a very important prerequisite for the centralization of the state apparatus, which in turn was necessary for the establishment of a modern bureaucratic state. The subtitle of Lagerroth's important work on the change of the Swedish tax system during the second half of the 19th century is telling: A System Shift in the Swedish National Economy. In the introduction, Lagerroth regrets that he was not able to explore the more general significance of the changes since

the system shift in question has had a very significant, until now fairly overlooked impact on
both the government's legal form and on actual power relations between state institutions -
more so than many constitutional amendments. (Lagerroth 1927, emphasis added).

Westerhult describes the job-related duties in the civil service, before
salary regulations were instituted in the middle of the 19th century, as bearing

an unmistakable resemblance to the contract system. The civil servant had an agreement
with the state to perform certain duties for certain compensations; this contract could not be
broken without the consent of both parties (Westerhult 1965, 166).

It was not until 1858-60 that new ordinances forced civil servants to accept any new duties that might be placed on them. The feudal perspective on public office made it possible for the civil servant to manage his office as he saw fit. Westerhult describes the situation before the reorganization around 1870:

It was the civil servant's duty to maintain an office, set office hours, keep records in order, obtain and pay for office supplies and clerks. The state did not keep track of how the civil servant used his government allowances - if the travel allowance sufficed or if as many bailiffs were being retained as were on the payroll. The civil servant was solely responsible for all expenses related to the position. How the civil servant dispensed his work time was not monitored . . . (ibid.).

Even though Westerhult's description applies to local civil servants, these conditions were presumably prevalent throughout the entire public administration (Hildebrand 1896, 653; Myrberg 1922; Cavallin 1996). For local civil servants, the salary regulations issued in 1878-80 meant that salaries would be divided into three parts: salary, service fees and expenses. Thus, the government had actually assumed responsibility for job-related expenses (Westerhult 1965, 166). The salary regulations of 1858-60 also led to the introduction of a modern salary grade system entitling civil servants in similar positions to fairly equal pay (Ekenberg 1952, 26).

Organization, Rationality and Responsibility

The parliamentary debates from 1809 to 1880 provide many examples of attempts to establish a more rational and uniform public administration. Statements calling for "an organized, enterprising and secure treatment of affairs" in the civil service departments were common.⁶ The most obvious evidence of Riksdagen's efforts to bring about these administrative changes is the great number of governmental committees that were set up from 1811 to 1878.

Before the 1870s, however, the results of these efforts were meager. In 1840, the ministries were reorganized, but this did not lead to a reformation of the central administration in a bureaucratic sense (Sörndal 1943, 280; Hildebrand 1896, cf. Cavallin 1996). In the parliamentary debates there were attempts to achieve greater uniformity and rationality in two ways. First by replacing the time-consuming, antiquated principle of collegial decision making with the more effective bureau and delegated decision making system; and second by trying to subordinate independent civil service departments under the ministries. During the 1870s, a number of new instructions were issued to the central civil service departments where the bureau and delegated decision making system was gradually introduced (Hildebrand 1896, 653; Sörndal 1943, 300).

Before the 1870s, there was a great interest in Riksdagen to fundamentally reorganize public administration (Sörndal 1943). Some of the new tasks that the state undertook aimed at the production of goods and jobs, such as railway construction, telegraph services and road/canal construction, presumably played a significant role in founding a rational administration, as defined by Weber. The Railway Board was established in 1862, the Road and Waterway Commission in 1841, and the Telegraph Board in 1856.

In his work on the history of the central administration in Norway, Beinum points to the very arrival of practical-minded personnel in public administration as the most important reason for its transformation. The ability to transform the administration into a bureaucracy also grew dramatically with the new salary regulations in 1858-60 (Beinum 1979). The state could subsequently require civil servants to accept new assignments that were more extensive than those agreed to when taking the position (Westerhult 1965, 162-66).

With regard to the regulatory system, civil servants could, up until around the 1830s, rather safely violate administrative ordinances and regulations (Cavallin 1998). An example is the aforementioned ineffective prohibitions imposed against ackordssystemet. Civil servants could of course cause the disapproval of the king or other superiors, but not for the types of activities that were classified as official misconduct according to the civil service laws. If anything, breaking these rules was met with approval as long as it was

apparent that it was being done in the best interest of the king (rather than of the state). Not until new criminal legislation was enacted in 1864 were there special statutes governing the misconduct of civil servants that applied to the entire public administration (Svea Rikes Lag 1864).

A Short Note on the Comparative Perspective

I have not had the opportunity to examine other Western countries closely, but if we concentrate on only one variable - the transition from the birth and loyalty-based recruitment system of the noble civil service corps to a system based on intellectual merits without regard for social status - it seems that this transition took place during approximately the same period in several comparable nations (Knudsen & Rothstein 1993).

According to Fischer & Lundgren (1975), the French revolution did not represent a definitive breakthrough for a system of merit-based recruitment to the state apparatus. Even though the civil rights declaration in 1791 established that no distinctions other than merit and skill were to apply in the recruitment of public officials, it took about one hundred years before these principles had any practical application. Instead, the republics, empires and kingdoms that, after 1789, quickly followed one after the other opened the way for favoritism and patronage in the politico-administrative sphere.

The Prussian experience is, as Anderson (1974, 281-95) has demonstrated, far more complex than in other West European states. On the one hand, welldefined rules regarding recruitment to and promotion in the state apparatus had already been established by the mid-1700s. On the other, the civil service corps was completely overrun by Junkers long into the 1800s (cf. Fischer & Lundgren 1975, 524). Anderson describes the Prussian state apparatus as the special territory of the aristocracy in German society even into the late 1800s (Anderson 1974, 295). Furthermore, the ties of personal loyalties to the emperor permeated the Prussian civil service corps. Gillis (1979, 189) points specifically to the mid-1800s as the period when an aristocratic "ethos" was replaced by a bureaucratic one. With regard to qualifications, the connection that had earlier existed between a higher academic education and the prospect of attaining a position in the government seems to have been just as illusive as in Sweden. Gillis argues that the trend toward higher academic norms had begun early in the 1800s, but the effects of this change only became evident around the middle of the century. As a new breed of bureaucrats more attuned to the specialized needs of an industrialized society arose, professionalization ceased to be a cause of division in the bureaucracy and became a source of a new kind of consensus based less on traditional corporate concepts such as honor and pride than on expertness and efficiency. Fischer & Lundgren (1975, 524) also dismisses the idea that the early Prussian merit-based

recruitment system was a break with the aristocratic esprit de corps, and they
identify the 1860s as the dividing line between aristocratic and bureaucratic
recruitment systems.

In the British case, this line was drawn, primarily as a result of the changes in the civil service that took place in the wake of the publication of the Northcote/Trevelyan Report in 1853. Up until then, Britain's government personnel had all the distinguishing attributes of the late feudal state. According to Chapman & Greenaway (1980, 123-27), the implementation of the Northcote/Trevelyan program from 1860-70 entailed the establishment of an efficient, permanent civil service, free from corruption and patronage, recruited by open competition, divided into grades and centrally directed.

If this overview is any indication, it would seem that the timeline of the
Swedish experience corresponds to that of the most influential West
European states.

Conclusion

The preliminary nature of this study notwithstanding, most empirical material indicates that the transition to a bureaucratic administration in Sweden started in the 1850s, at least in terms of institutional structure. Contrary to what most Swedish historians have argued, the Swedish state remained feudal and particularistic until the mid-19th century. If any particular decade is key to this transformation, it would be the 1870s. By then the last of the noble privileges had disappeared, a uniform salary system had been introduced, and the various government agencies had begun reorganizing toward a higher level of efficiency and rationality. The perquisite system had for the most part been abolished and the accord system broken. Although feudal elements remained in the system for some time, the 1870s will have to serve as an approximation as far as institutional structure is concerned. Of particular importance is the fact that it was not until that time that a clear-cut connection was established between acquired intellectual ability (especially acquaintance with the public regulatory system) and the opportunity to enter and make a career in the state machine.

Thus, our preliminary findings indicate that the market economy and the rationalized bureaucratic state developed simultaneously in Sweden. It is also evident that most of the decisive changes toward a bureaucratic administrative apparatus occured within a relatively short time frame. I would identify the thirty-year span from 1840 to 1870 as the actual transition period. The transition period started with the widespread emergence of liberalism around 1840, when the reign of King Karl XIV Johan - the first Bernadotte to take the Swedish throne and who built his reign entirely on the personal loyalties of the noble civil servant - was coming to its end. This was the socalled

Brahe empire, named after the country's most distinguished aristocrat and the king's confidant Count Magnus Brahe. Determining the exact end of the transition period is more difficult. However, with regard to the reorganization of public administration from 1876 to 1878, Hildebrand writes in his widely read political science textbook from 1896:

The new principles gradually came into effect, leading to the dissolution, reorganization or integration of most of the colleges, and the founding of new simpler administrative divisions, normally based on the bureau or delegated decision making system. The old view and treatment of administrative offices as property was beginning to disappear . . .

(Hildebrand 1896, 653, emphasis added).

Whether this penetration of a bureaucratic administrative order - in the 1870s - came before or after or perhaps simultaneously with the establishment of capitalism as a dominant economic system is a question that cannot be answered here. The impression one gets going over the debates of the fourestate Riksdag is that there was a political awareness relatively early on - in the 1820s - that administrative changes toward a modern bureaucratic organization were desirable. Also the many inquiry commissions that were appointed to investigate the matter during the first half of the 1800s are an indication of this. By this time, an awareness of the existing educational inadequacies were also beginning to take shape (see the main reports of the Training Committee in 1828).

The fact that it would take another fifty years before the changes were implemented, at least institutionally, can probably be attributed to two things. For one, the sluggishness that prevails in all large administrative organizations as well as the necessary generational turnover within the civil service. Second, the reign of King Karl Johan from 1818 to 1844 and the King's fear of constitutional change in addition to the feudal authoritativeness of his nature must have had a delaying effect. Finally, a common argument with regard to the modernization/bureaucratization of the administration was the general backwardness of the Swedish economy. The arguments against the changes expressed in parliamentary debates and in rejected committee reports were seldom fundamental in their defense of the old feudal-aristocratic order; they were more often of an economic nature, i.e., that changes would be too costly. In other words, if further empirical research could confirm these observations, it might become apparent that a politico-ideological awareness of the need for bureaucratically inclined administrative changes had formed before the economic situation would allow them to come to fruition.

---

NOTES

1. It should be added that the change from thc four-cstatc Parliament to a two-chamber system took place in 1866.

2. Within certain administrative areas the nobility held on even longer. Sometimes the foreign affairs administration and the officer corps of the Royal Swedish Life Guards are held up as examples of this. See Carlsson (1949, 59-65 and 95).

---

3. In fact, the Johan Skytte Prize in Political Science at Uppsala University, established in 1995, gets its revenue from land donated to the University from Johan Skytte during the 17th century.

4. Records from the Swedish Riksdag 1840 (Allmänna Besvärs- och Ekonomiutskottet Betänkande nr 43) and Royal Ordinance (Kungl. kungörelse) November 14, 1879.

5. Till regleringen av rikets styrelser och verk comiterades underdåniga betänkande. Stockholm 1823, 180.

6. Records from the Swedish Riksdag: Ständernas skrivelse ad. Jan 10, 1809.

REFERENCES

Anderson, P. 1974. Lineages of the Absolutist State. London: NLB.

Asker, B. 1993. "Från godsägarstat till ämbetsmannastat. Byråkraterna, officerarna och
enväldets införande," in Dahlgren, S., Floren, A. & Karlsson, A., eds., Makt och vardag. Hur
man styrde, levde och tänkte under svensk stormaktstid. Stockholm: Atlantis.

Beinum, E. 1979. Sentraladministrasjonens historie 1845-1884. Oslo.

Blomkvist, H. 1992. "Democracy and State-Formation," in Ashford. D. E., ed.. History and
Context in Comparative Public Policy. Pittsburgh: Pittsburgh University Press.

Carlsson, A. B. 1913. Den svenska centralforvaltningen 1521-1809. Stockholm.

Cavallin, M. 1996. "Sportler: Muta eller laglig avloning?" Unpublished paper. Department of
History, Goteborg University.

Cavallin, M. 1998. "Fortroende forpliktigar." Unpublished paper, Department of History.
Goteborg University.

Chapman, R. A. & Greenaway, R. J. 1980. The dynamics of administrative reform. London:
Croom Helm.

Chester, Sir, N. 1981. The English Administrative System 1780-1870. Oxford: Oxford
University Press.

Dahlkvist, M. 1982. Staten som problem. Lund: Arkiv.

Eden, N. 1902. Den svenska centralregeringens un-eckling til! kollegial organisation i bor jan
av sjuttonde århundradet. Uppsala.

Ekenberg, O. 1952. Den svenska statsforvaltningens avloningsvåsende I. Stockholm.

Englund, P. 1993. "Om klienter och deras patroner," in Dahlgren. S.. Florén. A. & Karlsson.
A., eds., Makt och vardag. Hur man styrde, levde och tårikte under svensk stormaktstid.
Stockholm: Atlantis.

Evans, P. 8., Rueschemeyer, D. & Skocpol. T., eds.. 1985. Bringing the State Back In. New
York: Cambridge University Press.

Fischer, W. & Lundgren, P. 1975. "The Recruitment of Administrative Personnel." in Tilly. C.
ed., The Formation of National States in Western Europe. New Jersey: Princeton University
Press.

Frohnert, P. 1993. Kronans skatter och bondens brod: Den lokala forvaltningen och bonderna i
sverige 1719-1775. Stockholm: Rattshistoriskt bibliotek - Ncrenius och Santérus Forlag.

Hildebrand, E. 1896. Svenska statsforfatningens historiska un-eckling. Stockholm.

Jorberg, L. 1961. Growth and fluctuations of the Swedish industnalization 1869-1912.
Stockholm: Almqvist & Wiksell.

Kammarkollegiets historia. 1941. Stockholm: Kammarkollegiet.

Kettering, S. 1986. Patrons, Brokers and Clients in Seventeenth-Century France. New York:
Oxford University Press.

Knudsen, T. & Rothstein, B. 1993. "State-Building in Scandinavia," in Comparative Politics
26.

Lagerroth, F. 1927. Indelnings- och grundskattevåsendets aweckling: ett systemskifte inom
svensk statshushållning. Malmo.

Lindroth, S. 1971. Uppsala Universitet 1477-1977. Stockholm: Almqvist & Wiksell.

Molin, 8., Månsson, L. & Stromberg, L. 1979. Offentlig forvaltning: stats- och
kommunalforvaltningens struktur och funktioner. Stockholm: Bonniers.

Myrberg, I. 1922. "Bidrag till tjånsteackordens historia i vart land," in Från svenska
statsforvaltningen 1922:4, 5 & 6.

Nilsson, S. A. 1988. "Imperial Sweden. Nation-building, war and social change," in Losman,
A., Lundstrom, M. & Revera, M., eds., The Age of New Sweden. Stockholm:
Livrustkammaren.

Nilsson, S. A, 1993. "De stora krigens Sverige," in in Dahlgren, S., Florén, A. & Karlsson, ń
eds., Makt och vardag. Hur man styrde, levde och tånkte under svensk stormaktstid.
Stockholm: Atlantis.

North, D. 1990. Institutions, Institutional Change, and Economic Performance. Cambridge:
Cambridge University Press.

Nystrom, P. 1974. Historieskrivningens dilemma och andra essåer. Stockholm: PAN/
Norstedts.

Olivercrona, S. R. D. K. 1859. Om den juridiska undervisningen vid Universitetet i Uppsala
och om den Juridiska fakultetens flyttande till Stockholm. Stockholm.

Poggi, G. 1978. The Development of the Modem State. London: Hutchinsson.

Rosenberg, N. & Birdzell, Jr., L. E. 1986. How the West grew rich: the economic
transformation of the industrial world. New York: Basic Books.

Schon, L. 1981. Industrialismens forutsattningar. Lund: Studentlitteratur.

Sorndal, O. 1943. "Fattande av beslut i centrala ambetsverk," Statsvetenskaplig tidskrift.

Tilly, C. 1990. Coercion, Capital and European States, AD 990-1990. Oxford: Basil
Blackwell.

Weber, M. 1978. (original published in 1922) Politics as a Vocation. Berkeley: University of
Californa Press.

Weiss, L. & Hobson, J. 1995. States and Economic Development: A Comparative Historical
Analysis. Cambridge: Polity Press.

Westerhult, B. 1965. Kronofogde, håradsskrivare, Idnsman. Den svenska fogderiforvaltningen
1810-1917. Lund: Gleerups.