The rise of the modern state coincided with tremendous scientific and technological advances and vast improvements in the standard of living for much of what is now thought of as the Western world. Until recently the most accepted premise has been that such developments led to improvements for mankind, that the world became more “civilized”. Instead of locking up and hiding away the mentally ill, society began to treat them. Prisons were reformed with the hope that the prisoners could be taught to be responsible citizens. The “feeble-minded” were to be provided with an elementary education so that they too could function in society. Science began to offer solutions to problems, even in the social sphere. The individual could be treated and an ordered society was possible. During the nineteenth century those working with social problems were optimistic.
As the nineteenth century faded, so too, did some of this optimism. Not all of the mentally deficient
There are numerous terms which may be used in contemporary... could be taught, and not all criminals could be reformed. Despite these setbacks, the architects of the modern welfare state sought new solutions. This was also the case in Sweden, which, although in a sense on the European periphery, was by no means isolated. Scientific developments in such fields as sociology, psychiatry, and biology offered new alternatives for dealing with those who were deemed untreatable. In Sweden, racial biology coupled with social darwinistic ideas led, for example, to laws forbidding the mentally deficient to marry (1915) and somewhat later to laws (1934 and 1941) that allowed the sterilization of such individuals. By stopping individuals who did not fit into the new emerging modern society from reproducing, Swedish social engineers thought they could improve society for future generations. At the time, these measures were little debated and the basic premise, that one could in fact improve society, was generally accepted.
After the Second World War and the horrors of the Holocaust the entire social service sector came under debate and by the 1970s most earlier forms of care and treatment were being reformed. To accept that not all individuals can or should be molded into model members of a modern society appears now to be the guiding principle in the field of social welfare. Society must now be arranged so that even those who have problems fitting in, for whatever reason—physical, mental or social—can also reap the benefits of progress. Despite a number of political, economic, and environmental setbacks, most of the advanced nations of the world are committed to the idea that scientific and technological advances will continue to improve conditions for mankind.
It is against this background that the development and use of statistical categories must be seen. In order to help the mentally ill, the mentally deficient, the criminal and the handicapped, and/or protect society from these groups, one needs to know who they are. As the Swedish historian Anders Berge has pointed out, in Sweden as elsewhere, both the motivations for and the ramifications of the creation of these categories are many, both positive and negative (Berge, 1998, 616-627). In this paper we shall trace the development of a number of statistical categories that were of relevance for emerging social welfare policies. At their inception, Swedish population statistics were concerned primarily with age, sex, civil and occupational status. There was also a column for those adhering to “foreign” religions, as all Swedes, by definition, were regarded as members of the Lutheran state church (Sköld, 2001, 60-63). Rather than treating statistical categories dealing with demographic behavior or occupations, we will focus on those categories that might be termed “the others”. The reasons for analyzing these types of statistical categories are twofold. First, the arguments underlying the choice of categories provide us with insights into how statistical data became part of developments in the field of social welfare. Second, by tracing the statistical development of “the others”, we are better able to understand the ways in which “modernity” has helped shape statistical data collection.
During the latter part of the twentieth century, the idea that progress and modernity, in this case expressed as the modern welfare state, resulted in an entirely positive development came under attack. Michel Foucault and others began to question the idea that the development of modern society led to a civilizing process and the continued betterment of all members of society. Instead, they suggested that incarcerating criminals, the insane and others was a disciplinary act implemented in order to protect society from such individuals (see, e.g., Foucault, 1983, 1987). Recent research questions the premise that the two theoretical approaches are mutually exclusive. Empirical evidence suggests that most of the policy decisions included both an element of disciplinary action as well as a civilizing motive. The development of social institutions such as prisons, asylums, and reformatories were intended to provide treatment for the individual and to protect society (Larsson, 2001; Nilsson, 1999). Social policy was and still is for the benefit of the individual and for the benefit of society, not one or the other.
James C. Scott in Seeing Like a State. How Certain Schemes to Improve the Human Condition Have Failed and Zygmunt Bauman in Modernity and the Holocaust further discuss the idea that modernity and the process of civilization go hand in hand and lead to a better society (Scott, 1988; Bauman, 1989). Utopian goals were an integral part of the modernization process. They suggest, however, that such developments as the Holocaust in Germany, collectivization in Russia or The Great Leap Forward in China were a feature of modernity rather than a temporary interruption in the quest for a better society. Their arguments offer a useable theoretical framework for the study of the development of statistical categories related to “the others”. Scott argues that an essential part of the process of modernity is the ability to comprehend complex social and natural phenomena. In order to achieve this, the leadership of developing modern states needed to know more about their subjects and more about what natural resources were available. To this end they tried to organize society and make it “legible” by creating maps, gathering statistics, standardizing names and measurements and a whole host of other actions (Scott, 1988, 4-5). Both Scott and Bauman argue that with modernity (Bauman’s term) and high modernism (Scott’s term), engineers, technicians, architects, scientists, and other related groups developed an uncritical optimism as to their ability to control and plan both the environment and the population. These administrators and planners believed in technical and scientific progress.
Bauman has likened this modern bureaucracy and its attempts to improve society to a gardener who plans and cultivates his garden, caring for those plants that are useful and weeding out those that are not (Bauman, 1989, 40). Scott suggests that such a bureaucracy is not necessarily dangerous as long as a strong civil society exists that is capable of controlling its leaders. When an authoritarian state emerges at the same time civil society is weakened, then planning and control can lead to tragic consequences (Scott, 1988, 5). In the light of Scott’s and Bauman’s work, we will consider when and, if possible, why certain statistical categories were adopted in nineteenth century Sweden. We will trace their development and consider the problems associated with trying to determine who belonged to these categories. Furthermore, we shall consider how Swedish social engineers and planners used these categories. The selected categories include the “Lapps” (Saami), Finns, Jews, gypsies, and “idiots” (mentally deficient) because they were major categories singled out over time and are indicative of the kinds of categories selected by Swedish officials.
Swedish population statistics
The collection of population statistics has a long history in Sweden. The world’s oldest continuous population registration system on a national level began in 1749 with the establishment of Tabellverket. Swedish intellectuals, businessmen and government officials believed that the country suffered from under-population because of the tremendous losses in manpower that occurred during the xviith century. The xviiith century did not begin well. Not only was the country involved in devastating wars eventually resulting in the loss of most of its overseas possessions; it also was visited by the plague, resulting in significant population losses wherever it struck. King Charles XII died in 1718 and Sweden’s role as a major European military power came to an end, but the leaders of the country hoped that Sweden could develop economically. Influenced by mercantilistic ideas, they accepted the premise that a populous nation was a prerequisite for progress. However, the mercantilist writers were not interested in breeding more soldiers. Rather they were concerned with exploiting the country’s natural resources and furthering its industrial growth, a task which demanded human resources (Johannisson, 1988, 96-102). In order to promote this growth, they needed to know the size and make-up of the Swedish population, as well as what could be done demographically to improve the situation (Rogers, 1999; Sköld, 2001). Well aware of developments in political arithmetic, Swedish academicians and parliamentarians set about to create a system for the collection of population data. They opted for a system based on the country’s ecclesiastical registers rather than collecting data through census taking. Three forms or tables were to be completed each year containing basic demographic information as well as information on occupation and social status. The first form referred to baptisms and burials with baptisms arranged by sex, legitimacy and month and burials by sex, month, and three categories that reflected social dependencies
The categories reflecting social dependency were (1).... The second table was a cross-tabulation of deaths by cause, age, and sex. The third form supplied census-type data on the population by sex and age including information on marital status and occupation or social standing (Rogers, 1999, 189-190). The parish ministers compiled the required data from their respective parishes on pre-printed forms that were all sent directly to Tabellverket in Stockholm, the Swedish capital, and then converted into provincial and national statistics (Rogers, 1999, 194-198).
The use of these printed forms meant that the various categories in the three tables were set and could not be changed. Printed forms were supplied for a 25-year period. This insured uniformity in the collection of data and comparability over time. However, since no changes were allowed, the system was not flexible and did not permit the introduction of, for example, new occupational categories or changes in the cause of death terminology. From 1774 until 1860 when the National Bureau of Statistics replaced the Tabellverket, the forms were revised and updated several times at irregular intervals. With the establishment of the National Bureau of Statistics, statistical information was published and widely distributed for the first time. Population data had often been considered a state secret, and earlier statistics had been either handwritten or printed in a limited number of copies for the use of different government agencies. The gathering of statistics still relied on information sent in by parish ministers but the information was more thoroughly controlled and the statistics more accurate. New statistical series were quickly added and by 1900, 23 new series covering such areas as transport, foreign trade, agriculture, and health care were published. In 1862 the National Bureau of Statistics began publishing a statistical journal, Statistik tidskrift, devoted to scientific articles related to the statistics compiled by the Bureau (Sköld, 2001, 284-286).
Interest in statistics varied throughout the twentieth century, but the production of statistics steadily increased (Höjer, 2001). Only four to five people worked at the National Bureau of Statistics in its early years, as the initial work of gathering statistics remained in the hands of the state church. The number employed increased steadily, reaching approximately 1900 in the 1970s. In 2001, the bureau had 1369 employees (http:// www. scb. se, Feb. 2003, Statistics Sweden Annual Report 2001, 25). The collection and production of statistics remained fairly decentralized until the 1960s. With the more frequent use of computers, statistical collection and production became more centralized. In 1994, responsibility for a number of statistical series was transferred to other governmental agencies
(http:// www. scb. se/ omscb/ organisation/ historia. asp. February 2003).
Lapps, Finns and Gypsies
The Saami, who earlier were called Lapps, are a nomadic people living in northern Norway, Sweden, Finland and northwestern Russia. Originally hunters and fishermen, the Saami began herding reindeer on a regular basis sometime during the sixteenth century. Interest from Swedish authorities in the wealth the North had to offer was evident from the fourteenth century, and they began increasing control over present-day northern Sweden and Finland, in part by taxing the Saami. During the reign of Gustav Vasa in the sixteenth century, development of a conscious policy can be discerned. Christian missionaries sought out the Saami in the seventeenth century. At the same time, Swedish and Finnish peasant-farmers began encroaching on traditional Saami lands. A system of taxation based on territorial units was introduced that would last until the end of the xixth century, the period when the Swedish State once again directed its attention to the Saami (Beach, 1988, 346-347).
Statistical information on the Saami became available for the first time in 1805. It has not been possible to determine from the existing sources why the Saami were included as a category in the 1805 population forms, but these new forms were a substantial revision of earlier ones (Underdånigt betänkade…, 1856, 33; Sköld, 2001, 194-203). Two important changes may explain why the Saami were included. First, more information on agricultural production was sought. Since the Saami were involved in reindeer herding, fishing and hunting, it is reasonable to assume that they should be included in the statistics. Three categories were included: those herding reindeer, those who did not herd reindeer, and those who were vagabonds or hired hands (Sköld, 2001, 201). Second, the occupational categories were redone and instructions were sent out to parish ministers that they should avoid double registrations. Tabellverket was trying to get the occupational statistics and the population statistics to match (Sköld, 2001, 210). If this was to work, then the Saami population would have to be included, since at least some of the Saami (those who were not nomads) were included in the population statistics. One can suspect underlying economic motives such as the potential control of land and other resources.
During the 1820s, debates concerning the Saami became increasingly complex. On the economic side, there were the conflicts between the reindeer herding Saami and new settlers, as the settlers pressed the cultivated area further west, encroaching on the traditional Saami territory, Lappmark (Beach 1988, 347-348; Lundmark, 2002, 47-62). These conflicts, with clear roots in the struggle to develop new policies in the early xixth century, culminated in the final loss of control by the Saami over the traditional Saami “tax lands” in 1886 (Lundmark, 2002, 244-262).
In 1825 the statistical forms were once again revised. After protests from the parish ministers the information on agricultural production was dropped. Apparently, there was some confusion regarding the categories for the Saami population in the earlier forms. In the first volume of Bidrag till Sveriges Officiella Statistik (BiSOS) the information on the Saami between 1805 and 1825 was deemed unreliable. In 1860 the statistics on the Saami between 1825 and 1860 were recounted using the primary data originally received by the Tabellverket (BiSOS A. I, 3, 1851-1855, 69
In 1900 the series changed names to Sveriges officella...). In the nominative forms sent in by the parish ministers for the 1860 statistics, the ministers were to designate who belonged to the Saami “race” and who belonged to the Finnish “race”. But, interestingly, the column heading for this information was “nationality” (BiSOS A. I, 1856-1860, XVII). When the material concerning the Saami was summed up in 1870, different terms were used interchangeably to describe the categorization: nationality (nationalitet) and stock (stam) (BiSOS A. XII,3, 1870, XXVIII). The frequency with which the latter term was used increased in the census reports, and by the 1880 report was more common; the term nationality had definitely taken a back seat (BiSOS, XXII, 3, 1880, XXI). It was later that the term ras (race) became more common in the reports and then was often used synonymously with stam
Stam, the oldest of the three terms, appeared in the.... From this point on until after the Second World War, the category “Lapp” was assumed to be a racial category. This early confusion of terminology and use of these terms as synonyms, at least initially, is an apparent reflection of a certain ambiguity that was erased in the early xxth century.
Determining who was to be categorized as a “Lapp” presented problems from the very outset
The circumpolar area has long been inhabited by distinct.... A marginal comment on one of the forms submitted to the Central Bureau of Statistics in the 1860s remarked that it was difficult to differentiate Saami from Finns and Swedes in parts of the far North. The writer added that language and life styles were used as guidelines when doubtful (SCB: 1870-års folkräkning, Hib: 1). Even the doctors made similar notes in their official reports (RA: Medicinalstyrelsens arkiv, Provinsialläkares rapporter, 1867, Övertorneå district). Bi-and tri-lingualism was not uncommon in the early xxth century (Collinder, 1926, 121-124). Between the 1860 and 1900 censuses (censuses were compiled every tenth year
The censuses were not carried out by census takers...), the language used at home determined who was a Lapp, and this was still recommended by the 1894 regulations governing registration by the church authorities. In 1910 the first name of the individual was also used to determine who belonged to the Saami. An article in Statistik Tidskrift in 1913 pointed out that this was still the practice at the time and that the use of language to determine “race” was unreliable (Arosenius 1913, 264).
Racial biology gained prominence in Sweden during the first half of the twentieth century. Sweden was the first country in the world to open a state-run Institute for Racial Biology, established in 1921 in Uppsala, but its roots stretch back into the xixth century and beyond. The xviiith century interest in classification developed into the xixth century physical anthropology, a field in which a number of combined “archeologist-anatomist-physician biologists” became prominent (Broberg, 1988, 314-317). During this period the National Bureau of Statistics increased their efforts to determine who belonged to what they called “the race of Lapps”. They still had difficulties determining who should be designated Saami. Mixed marriages were considered a problem from the classification point of view, and in 1920 the Bureau decided that children in mixed marriages should be classified according to the father’s racial origin (SOS. Folkräkningen, 1920, II, 23). This led to an absurd situation. The children of a Saami woman married to Swedish man would be classified as Swedish; if her son married a Saami woman their children would be classified as Swedish. If this continued, the number of registered Saami would decrease drastically at the same time as their Saami heritage increased each generation (Sköld, 2001, 326). The background to this characterization of the Saami can also be seen in the debates concerning the fate of the Saami as a group. Lennart Lundmark’s discussion of the origins, motives and character of the Swedish policy in regard to the Saami points out how the increase or decrease of the Saami population was essential for various arguments or schools of thought. To the advocates of assimilation or to those intent on isolating the picturesque reindeer herders (the “Lapps should be Lapps” concept), the numerical status and even the desirability of their existence as a separate group had vastly different implications in terms of economics, science, and culture (Lundmark, 2002).
In the 1930s, the Bureau once again tried to improve their statistics on the Saami and find a workable distinction for the definition of a “Lapp”. They conducted a special study on language and also introduced the categories “full-Lapp” and “half-Lapp”. A full-Lapp was one who was born in a Saami village or whose parents were born there. If only the father was born in a Saami village, then the person was categorized as a half-Lapp. If only the mother was born in a Saami village, then the person was classified as a half-Lapp of Swedish or Finnish extraction (Lundmark, 2002, 149). In 1928, attempts were made to restrict the definition to those herding reindeer, something which may be viewed as part of the “Lapps should be Lapps” concept, an idea which, Lundmark has argued, had an economic base. Other possibilities apparently did not occur to the framers of the statistical forms.
After the atrocities of World War II, race biological aspects in data collection were toned down. In 1945, a “Lapp census” was taken and those who spoke Saami or had the right to reindeer herding according to a 1928 law were designated Saami (Lundmark, 2002, 150). Since 1993, the Saami have their own parliament which elects members every fourth year. The law remains ambiguous in the question of designating who is Saami. A person who considers himself-/herself to be Saami, and who has spoken Saami in the home or whose parents or grandparents have done so, may vote. (SFS, 1992, 1433. See also http:// www. sameting. se). The vagueness of the definition obviously beds for discussions of interpretation.
The Finns emerge for the first time as a separate category in 1860 and were included in the same tables as the Saami. When the parish ministers were asked to provide information on “nationality”, Finnish was one of those required. In part this was due to the fact that the majority of the Finns lived in the same areas as the Saami and, in order to sort out the Saami, information on the Finns was also needed. Using language as a determinant of the supposed racial origin of Finns was just as problematic as for the Saami. The problem of determining who was a Finn was complicated by the fact that parish ministers also used place of birth and traditional manners and customs as criteria. The Finnish population was concentrated to the northern part of Sweden along the Torneå River, which was also the boundary between Sweden and Finland. According to the 1870 census, over 90 percent of the Finns lived in 13 parishes in Norrbotten, Sweden’s northernmost province. It is clear that only a small minority of those designated Finnish were born in Finland. According to the 1870 census, of the 14,015 designated Finnish, 2,018 were born in Finland (BiSOS A. XII, 3, 1870, XXVI-XXXII).
The 1890 census looked more closely at the distribution and demographic development of the Finnish population. According to the provincial government, the language problem, i.e. that only Finnish was spoken along the Torneå River, was due to the fact that these settlements were isolated from Swedish communities by large roadless tracts of marshes and forest. On the other hand, just across the river resided a large Finnish population. The provincial government also claimed that the Finnish-speaking Swedish population wanted to learn Swedish (BiSOS A XXXII, 3, 1890, XXXIV-XL). The Tornedal Finnish-speaking population was subjected to various “Swedification” policies in language matters during the xixth and xxth centuries. (Lainio, 1989, 32). In 1910, the problem of distinction still remained, and Finns were considered to be even more difficult to distinguish than the Saami. The 1894 regulations on how population statistics were to be recorded in the parish registers recommended using the language spoken at home when all else failed. When the central authorities tried to reconstruct these categories from the local material that had been submitted to the national bureau, two criteria were used: name, for example, a Finnish given name combined with a Swedish patronymic, and/or place of birth, a Finnish place known to be dominated by Finnish speakers. Again, in mixed marriages, the father’s origin determined the category to which the children were relegated.
Since Finland was part of the Swedish realm until 1809, numerous Finns had migrated to the Swedish side of the Gulf of Bothnia, particularly the inland forested regions, for example, in Dalarna and Värmland. Indeed, such areas were often referred to as Finnmark, the Finnish territory. As assimilation took place, these descendents of Finnish settlers were no longer considered as part of the Finnish population (Arosenius, 1913, 266-269). The Finnish population in the northernmost province, Norrbotten, was at least in part a result of the population left behind in 1809, as the traditional border between the Finnish and Swedish population was the Sangis River, located west of the 1809 border (Erixon, 1945, 3-4). Thus the question as to why it was necessary to keep track of the Finns is far from clear. The economic and even later, racial, motives apparent in the case of the Saami are not as evident here. The discussion of classification in the 1890 and 1900 census provide some clues. Discussions began to emerge concerning the homogeneity of the Swedish people. The notable exception to this supposed “purity” was the northernmost province of Norrbotten as a result of the presence of the Finns and the Saami (BiSOS A, XXXII, 3, 1890, XXXIV; BiSOS A, XLII, 3, 1900, XXXI). It was even maintained that Swedes in the area had absorbed Finnish culture and language (BiSOS A, XXXII, 3, 1890, XXXVII). Henrik Höjer suggests that this may have been part of the mid-xixth century concern with “Swedishness” in the development of Swedish nationalism. It was seen as important to know the extent of homogeneity in the Swedish population, perhaps in comparison with other nations. This was part of the shift, as he sees it, from a quantitative to a qualitative emphasis in population statistics that embracing moral statistics implied (Höjer, 2001, 177-189).
Gypsies were also considered a foreign element and from 1860 attempts were made to provide statistical information on this group. The National Bureau of Statistics, however, found that the information from the parish ministers did not provide an accurate account of how many gypsies were in the country. Since the gypsies were on the move so often, it was assumed that the parish ministers did not register many. Gypsies were designated a separate category in 1870, although that year no figures were reported (BiSOS A, XXII, 3, 1870, 3, XXXII). Later policies were aimed at assimilation, and in 1914 the country’s borders were closed for the immigration of gypsies (Arnstberg, 1988, 485-486). Although the Bureau was aware that the gypsies were clearly underestimated in the statistics, they continued to publish the reported figures and saw some improvement in registration.
In 1910 another group was included with the gypsies, the tattare (tinkers), who were defined as being part gypsy, but were thought to be more likely to have a permanent residence (SOS. Folkräkningen, 1910, IV, 32). The inclusion of this group probably explains the steady increase in numbers of gypsies. One characteristic of this itinerant group has been their involvement in occupations otherwise eschewed by the majority population: horse trading, rag-collecting, metalwork of various kinds, etc. (Lindholm and Svanberg, 1988, 418-419). Being designated a tattare was a judgement call by the parish minister and the term was often used in a derogatory manner to single out those who lived on the fringes of society, often those with dark hair, eyes and complexion. The definition of what constituted a tattare shifted over time. By 1920 the tattare were often considered “travelers”, vagabonds who were thought to be cheats, drunks, who often fought with knives. In a sense they were considered Swedish-born gypsies. In the 1940s, a new government investigation could still not decide definitely on the tattare issue. A study by the State Institute for Racial Biology concluded that the tattare were of Swedish origin but had a different life style (Broberg and Tydén, 1991, Chapter 6). The ethnologist Birgitta Svensson has interpreted the social categorization of the tattare as part of the attempt to mould desirable lifestyles in the process of building the welfare state (Svensson, 1998, 532-546; see also, Svensson, 1993). Although the tinkers were often alluded to (but seldom specifically mentioned) in the sterilization discussions in the early xxth century, evidence exists of only a few operations having been carried out. The term was used in the official context until the 1950s (Tydén, 2002, 62-64).
Why were the Saami, the Finns and the gypsies treated as separate statistical categories? Although the exact numbers may readily be disputed, it is evident that these groups represented only small minorities in Sweden. The Saami numbered approximately 7,000 during the period 1860-1920 or about 0.015 percent of the entire population. The Finnish speaking population was larger, increasing from about 15,000 to 30,000 (approximately 0.05 % of the entire population) during the same period. The recorded number of gypsies rose to 773 in 1920 (SOS. Folkräkning, 1920, 24, 25, 40).
From the few sources we have on the early introduction of the Saami as a separate statistical category, it would appear that they were included for at least two reasons: economic and demographic. Early on, race does not appear to have been a dominant factor. Several authors, and particularly Lundmark, emphasize the predominance of the economic motives over time. As the star of racial biology ascended, so too did the presence of such motives in government policy. During the latter half of the nineteenth century, the continued interest in the Saami had to do with governmental campaigns that were in part aimed at assimilation. In particular their nomadic life in conjunction with reindeer herding was thought to be out of step with modern society, and it was upon this aspect that the categorization in the late 1920s was based (Lundmark, 2002, 132-169). As Sweden became more industrialized and urbanized, the Saami were seen as a quaint relic of the past. They had no place in modern society. The importance of keeping accurate statistics could sometimes be viewed by some as an end in itself, but in the long run their role in larger questions seems more important. In both the cases of the Finns and the Saami it is clear that influences from international contacts were also important as interest in racial homogeneity gained primacy and became the subject of comparison among nations. The reasons why the gypsies originally were treated as a separate category are less clear. Their late appearance in the statistics suggests racial factors were present, because, as in the case of the Saami, their lifestyle was not commensurate with a modern progressive society. By the 1920s, it was clear that the gypsies and the tattare were considered racially inferior even though in the case of the tattare their origins were unclear and debated. In a 1923 government investigation of the tattare problem, one of the solutions discussed (although not implemented) was that the tattare should be eliminated directly or indirectly. When Sweden passed its laws allowing sterilization in 1934 and 1941, this was considered as a way of coming to terms with the tattare. Although sterilization was rarely directed at the group per se, the fact that one was considered a tattare certainly reinforced the decision to sterilize many individuals (Broberg and Tydén, 1991, Chapter 6).
Although these groups were small and numerous problems were involved in collecting data on them, the fact that they turn up as statistical categories is a logical consequence of the development of the modern Swedish State. As pointed out by Scott, in order to run an orderly society, information on that society must be available. Statistical information on the inhabitants of a country is vital for the planners of modern societies, whether it is to exact taxes or to try to reshape the categories so that they will more readily fit into modern society. It is clear that this could not be accomplished without prior knowledge as to who belonged to these groups. By the xxth century concepts of race played an increasingly important role. The scientific backing provided by racial biology, with a state-sanctioned institute for research on the subject, supplied the necessary support to administrators and planners. There is another side to the coin. In the end, these statistical categories also led to positive treatment and the official recognition of these groups. The creation of the Saami parliament in 1993 has already been mentioned. On January 17, 2000, Sweden signed two European conventions officially recognizing the Saami, Swedish-Finns, Finns in the Torneå River Valley and the Romany as national minorities. Their languages, Saami, Finnish, Meänkieli (the Finnish dialect spoken in the Torneå River Valley) and Romany, were also recognized, and the Swedish government agreed to provide support and protection for these languages (Messing, 2000).
The history of the Jewish population in Sweden has roots which predate the laws that permitted the immigration of the Jews into the country. In spite of the regulations and decrees to the contrary, individual Jews settled in Sweden in the xviith and early xviiith centuries. An enclave was permitted in Marstrand, a free port, from 1775, and finally in 1783 Jews were allowed to settle in Stockholm, Göteborg and Norrköping, and later, in Karlskrona. They were allowed to engage in trade and become artisans, but with specific limitations. Nor were they permitted to own property except in designated towns. The opportunity of serving in the national parliament or of entering the civil service was not open to them, and marriage was restricted to their own religious group. In 1838 economic limitations were removed, followed by a gradual lifting of other restrictions. Political emancipation occurred in 1870. This Jewish population, prominent in the economic and cultural life of xixth century Sweden, became in many respects assimilated, although the assimilation question was debated among the Jewish population itself. The late xixth century saw the immigration of Jews from Eastern Europe. Conflicts often arose between the poor newcomers and the older, often wealthy, families. Indeed, substantial wealth had been one of the prerequisites for the early Jewish immigrants (Runblom and Tydén, 1988; Zitomersky, 1988; Svanberg and Tydén, 1992, chapters 14 and 18).
There were few early statistics on this group. Regular collection of statistics on the Jewish population in Sweden began in 1831. There had been a report from the Stockholm congregation in 1795, but in general little effort seems to have been made to complement this until the 1820s. This came about when the Tabellverket requested information on the members of congregations not belonging to the Swedish State church, a group including mainly Jews and the Roman Catholics. According to Sköld, the reason for requesting this information was to improve the comprehensiveness of the statistical material produced by Tabellverket (Sköld, 1988, 229-232). As a statistical category the Jewish population was designated as adhering to the Moasic confession or creed and not as an ethnic group, something which was explicitly stated in the statistical reports. There appears to have been no discussion concerning the collection of information for statistical purposes when the National Bureau of Statistics replaced Tabellverket in 1860 (Runblom and Tydén, 1988). In the 1920s, on the initiative of Professor Eil Heckscher, Hugo Valentin studied the Jewish population in Sweden using the primary sources collected in the 1920 census. The study was published in Valentin’s book Judarnas historia i Sverige (The History of the Jews in Sweden) as a separate chapter. The statistical information published on the Jewish population in the 1920 census is to a large degree taken from Valentin’s work. Significantly, in printed reports in 1920 the statistical material on the Jewish population had been moved from the section treating other religious faiths to the section on foreign “races” (SOS, Folkräkning, 1920, II, 36; Valentin, 1924, 541-563).
As noted with regard to the Lapps, Finns and gypsies, interest in racial questions increased during the 1930s. Discussing these problems the planners of the 1930 census came to the conclusion that “the only way to obtain more accurate information of foreign races appears to be through racial biological research and therefore inquiries should be made to the State Institute for Racial Biology in order to determine if they are engaged in such work. On the question of the Jews at present only those adhering to the Mosaic creed are accounted for; the only possibility to obtain more complete statistical information on the Jews appears to be if one could interest the Mosaic congregations for such an effort.”(Riksarkivet (RA), 1930-års Folkräkning F/, Protokoll, 15 maj 1929)
Early in the century, immigration legislation made the entry of Jews difficult, the first laws being passed prior to the First World War. Sweden’s position on how to treat the Jews during the 1930s and 1940s can be characterized as ambivalent. On the one hand, Swedish policy was, to say the least, restrictive in regard to the admission of Jews in this period. For example, in 1938 the notorious J-stamp in German passports was grounds for non-admission. Sweden also viewed itself as a transit land for Jewish refugees. Anti-Semitism existed openly, at the same time that vocal opposition to such views and actions was also found (Svanberg and Tydén, 1992, chapter 21; Runblom and Tydén, 1988, 195). The efforts of Swedish diplomats to help Jews fleeing from the Nazis has received much press outside the country, and efforts were also made later in the war to help fleeing Jews from neighboring Denmark and Norway. In these matters the picture is exceedingly complex. Paul Levine has argued that Sweden’s earlier policies of nonchalance were converted into activism later in the war (Levine, 1996). In spite of the fact that the Jewish population was known and categorized at an early date, the state did not use this information to prosecute the Jews. But efforts were made to see that the population did not increase in the early xxth century. Sweden’s Jewish population is now officially recognized as a national minority and Yiddish as an official minority language (http:// justitie. regeringen. se/ klarsprak/ , March 20, 2003).
“Idiots or fools” (idioter) were terms used interchangeably with the term “feeble-minded” (sinnesslö) in the late nineteenth century. As a statistical category separate from the insane, this group first appeared in the 1870 census. However, the problem of separating the mentally deficient from the mentally ill was discussed in conjunction with the 1860 census. Since little or no information was provided on the type of illness, the statistics compiled on the metally ill since 1840 were of little use for state administrative purposes or for scientific study. The National Bureau of Statistics concluded in 1860 that the collection of statistics on the mentally ill would have to be done by experts. If this proved to be too expensive, then they suggested that the usefulness of the statistics could be significantly improved if the age at which the mental illness occurred was provided. Two statistical categories, one for the mentally ill and one for the mentally deficient, were thought to be more useful in the planning of institutions for the care of such individuals. Those whose illness occurred during childhood were placed in the latter category (BiSOS A, II, 3, 1856-1860, XXXIV-XXXVI; BiSOS A, XXXII, 3, 1890, XLVII-L; Förhammar, 1991, 23-24).
In line with the move towards modernity, it was felt that these groups could be helped. In particular those involved in this field were optimistic regarding the chances of helping the mentally deficient. In 1870, the first institution to help in their education was established. Others were soon to follow and, while these first efforts were private initiatives, the state quickly began providing support (Hansson, 2000). The Association for the Care of Feeble-minded Children requested that the National Board of Health separate the mentally deficient from the mentally ill in their statistics (RA, Ecklesiastiska departementet, Inkommande brev, 14 dec. 1885), which occurred in 1890. As with the Saami, Finns and gypsies, the collection of reliable statistics was problematic. The statistics on the mentally ill and mentally deficient were based on information received from the clergy. Even though the clergy were to distinguish those who were mentally ill since childhood from those who became ill later in life, the terminology used by the parish ministers indicated that they were confusing the two categories. In 1890, two statistical categories were created. An expert categorized the various terms used by the parish ministers to describe the mentally ill. The figures for 1860 and 1870 were revised (BiSOS A, XXXII, 3, 1890, XLVIII-XLIX). In 1902, the National Board of Health conducted a study of the mentally ill and the mentally deficient in order to determine who was in need of institutional care. Their figures were about twice the number reported by the clergy (SOS, Folkräkning, 1910, IV, 33-36). By 1920 the figures for the two categories were more reliable and compared fairly well with the figures compiled by the National Board of Health. Because the city of Stockholm based its population statistics on poll tax registers instead of ministerial registers, information on the mentally deficient and the mentally ill was less complete there (SOS. Folkräkning, 1920 II, 55-57). The two statistical categories were used in both the statistical series on population and in the series on health.
The adoption of the mentally deficient as a statistical category had far reaching consequences (For further discussion, see Engvall, 2000; Runcis, 1998, 133-141, 232-238). In 1915, marriage between mentally deficient individuals was forbidden. In 1934 and again in 1941, Sweden adopted laws allowing the sterilization of individuals even if they did not agree to the procedure. The laws were in effect until 1975 and during the intervening period nearly 63,000 individuals were sterilized, 93 percent of whom were women. From 1936 until the early 1950s, the majority of those sterilized were classified as mentally deficient and they were sterilized mainly on eugenic grounds. Eighty-six mentally deficient persons were sterilized in 1936, and in 1944 and 1945 nearly 1,400 were sterilized per year. In comparison only about 325 mentally ill were sterilized during the same two years. The grounds for sterilization were vague but a rule of thumb used by the National Board of Health in approving applications for sterilization was that those with a mental age of 12 years or younger were considered incapable of deciding for themselves. IQ testing was frequently used. The stated aim of the modern Swedish state during this period was to improve the racial qualities of the Swedish “race” by eliminating undesirable elements, in this case the mentally retarded (Broberg and Tydén, 1991, Chapter 5). In a study of the application of sterilization laws to those classified as “feeble-minded” in institutions, Tydén has pointed out that certain institutions were more prone to use sterilization than others. Generally speaking, the shift away from the acceptability of this idea came in the 1960s as a result, not only of scientific developments (new theories and “the pill”), but also new views of society, individuals and groups (Tydén, 2002, 411-455). “Improving” society by these kinds of scientific applications was no longer acceptable.
Sweden, modernity and the use of statistics
The creation of various statistical categories was one of the more important means of making a society legible for those attempting to change that society. Population statistics played a central role in that any attempt at change in social, economic or demographic terms was dependent on having correct knowledge about the subjects of a country. Sweden was no exception. In the eighteenth century, mercantilistic ideas in their Swedish clothing influenced policy decisions. If Sweden were to regain a position of power in Europe, the country needed to develop economically. In mercantilistic terms, this meant increasing its population. In order to do this, the government needed to know not only how many subjects were living in the country, but also basic information on their demographic behavior. In particular more information was sought on mortality and fertility. To this end Tabellverket was created to supply the government with relevant statistical information. From the outset this information was processed in various statistical categories, for the most part basic demographic and economic categories such as age, sex, civil status, and occupational position. As the gathering and compilation of the statistics became more accurate, those in charge of producing the statistics sought to improve their completeness and reliability. New categories that had earlier been lacking were introduced. The registering of Saami and Jews in the early part of the nineteenth century are examples of this. At this early stage, it appears that these new categories were not introduced in order to control the members of these two groups, but rather to obtain a more complete picture of Sweden’s demographic development. Economic motives were also more apparent at this stage.
As Sweden became more industrialized, urbanized and modernized in the second half of the nineteenth century, statistical information became more important and emphasis shifted from purely quantitative to qualitative concerns. Government officials needed more detailed information on the population in order to plan and administer government activities in a rapidly changing society. As modern Sweden began to emerge, administrators and planners turned their attention to those groups that were having difficulty adjusting, in their eyes, to modern society. More information was needed in order to help assimilate these groups. Statistics were gathered on Finns and gypsies and later tattare, and statistical information on the Saami was improved. Educational campaigns, in particular with regard to the Swedish language, were launched. In the same spirit government planners and administrators needed better statistics on the mentally ill. By creating separate categories for the mentally ill and the mentally deficient, officials could better plan for their care, which in the spirit of the day meant institutionalization.
With increasing modernity, advances in science and technology offered new solutions to the problems facing the planners of the new Sweden. The statistical categories created in the nineteenth century and the statistical information gathered provided promoters of racial biology and Social Darwinism with the means of implementing their particular solutions. “Idiots”, gypsies and tattare could, in their view, eventually be eliminated through various measures ranging form restrictions on marriage to sterilization. Assimilation was at times considered desirable for Saami, Finns and Jews. Only among the mentally deficient did the measures taken lead indirectly to a loss of life. The sterilization of thousands of mentally deficient during the 1930s, 1940s, and 1950s on eugenic grounds was not an accident, a temporary lapse in the otherwise progressive move toward to a more civilized modern Sweden. Rather it was a logical consequence of the move toward modernity that included the application of the natural sciences to social “improvement” in true xxth century style.
Returning to the thoughts of Scott and Bauman, we find that developments in Sweden both support and refute their ideas on modernity. Scott’s argument that atrocities only occurred when an authoritarian state coincides with a weakened civil society appears to be, at least, partially substantiated. During the first decades of the twentieth century, when Social Darwinist and racial ideas began flourishing, Sweden made the definite transition from an elite authoritarian rule to a democratic parliamentary system. This, in Scott’s terms, should have put the necessary check on development toward the more tragic aspects of modernity. Yet, with the development of the welfare state in the 1930s with its social engineers, Bauman’s gardeners appear on the scene weeding out those who cannot be part of the “new society.” Even in a democratic state with a strong civil society, tragic events could occur. What is clear is that such tragedies could not have occurred without the development of certain statistical categories. However, the same statistical categories have helped to provide governmental guarantees and governmental support for these particular groups. Categorization is a double-edged sword.