Navigation – Plan du site
Articles

Illuminating the Labeling Impact of Incarceration: Life-Course Perspectives of Young Offenders’ Pathways in Comparison to Non-Offenders in Nineteenth-Century Northern Sweden

Lotta Vikström
p. 81-117

Résumés

Historiquement, on sait peu de choses du sort des condamnés après leur libération. La théorie de l’étiquetage suggère que ceux qui ont été définis comme criminels subissent les conséquences négatives de cette étiquette. Pour le vérifier, le présent article se penche sur l’existence de jeunes délinquants après leur emprisonnement, dans une zone de la Suède du nord connaissant une très forte expansion de l’industrie forestière au XIXe siècle. On a dépouillé les registres des prisons et on les a croisés avec les registres paroissiaux numérisés dans la Demographic Data Base (DDB) de l’université d’Umeå. Ces sources historiques permettent d’examiner la question dans une perspective longitudinale d’histoire de vie, telle que l’ont développée la criminologie et d’autres disciplines. En procédant à une comparaison avec un groupe de contrôle composé de non-délinquants et en examinant quatre circonstances de leurs vies (réinstallation, mariage, carrière et mort), on peut examiner si les délinquants ont réellement souffert de ce stigmate. Ce dernier a bien eu une influence négative sur leur parcours, mais il n’a pas provoqué de crise et n’a pas modifié le cours de leur vie de manière dramatique. Toutefois, les condamnés n’ont pas aussi bien réussi sur le marché du travail et les quelques femmes dans ce cas apparaissent comme les plus vulnérables, parce qu’elles étaient perçues comme ayant particulièrement transgressé les attentes à l’égard de leur sexe. Néanmoins, le constat selon lequel leur emprisonnement n’a pas ruiné l’avenir des jeunes concernés remet en cause l’idée de stigmatisation qu’implique le concept d’étiquetage. Il semble qu’ils ont été peu victimes d’exclusion sociale et qu’ils ont bénéficié d’attitudes tolérantes et d’un marché du travail prospère.

Haut de page

Texte intégral

1. Introduction

  • 2 Like scholars focusing on today’s criminology, those studying past times discover unfortunate facto (...)
  • 3 Cf. O’Brien (1982, pp. 226-257); Taussi Sjöberg (1986, pp. 145-160); Vejbrink (1982), all of whom s (...)

1Little is known about what happened to past offenders upon release. The limited opportunities to illuminate their life courses have narrowed the perception of them. Although individual offenders do appear in historical studies, they are seldom chosen to show typical pathways, as these are difficult to detect without data that reach beyond criminal sources where glimpses of offenders in retrospect are only available2. Laws and court cases document how control mechanisms confronted people in the past,but do not represent how offenders generally experienced life or how other people viewed them. There are more data on the lives and crimes of recidivists or those charged for felony. Given what is known and the anxious debate, control and societal transformation that encountered nineteenth-century lawbreakers, there is reason to believe that they were faced with social exclusion that made itdifficult for them to chart a pathway similar to that of non-criminals. A few pioneering studies in history indicate that crime and incarceration invited harsh living conditions or added to prevailing difficulties3.

  • 4 The study concerns the town of Sundsvall and five neighboring parishes within the region that witne (...)
  • 5 In the digitalized parish registers of the Demographic Data Base (DDB), every individual is given a (...)
  • 6 Brita Cajsa Wedin (DDB-ID: 848001367). Prison registers of Västernorrlands läns kronohäkte 1870-11- (...)
  • 7 Frans Walfrid Wedin (DDB-ID: 861001991). Prison registers of Västernorrlands läns kronohäkte 1877-0 (...)

2Whether individuals suffered from being labeled as offenders is explored below through highlighting young offenders in one of Europe’s fastest expanding areas in the nineteenth century – the Sundsvall region, situated about 400 kilometers north of Stockholm, Sweden4. Based on the growth of the sawmill industry, this area experienced some structural factors associated with high levels of criminality such as industrialization, migration and urbanization, but these factors are not central in this study. To test the stigma emphasized by some labeling concepts, it goes beyond criminal statistics and large-scale structures by combining different data that allow individual lawbreakers to come to the fore. Jonas Petter Wedin, born in 1857, exemplifies one of these lawbreakers5. His father was a farmer who took up employment as an unskilled laborer in the town of Sundsvall, where the family moved in 1861. When his father died of fever in 1868, his mother was left alone with five children. The next year she received poor relief, and one child was thrown upon the goodwill of others. Jonas Petter Wedin thus had a difficult start in life, and at the age of only 15 he entered the criminal scene by stealing something that did not belong to him. By then his older sister, Brita Cajsa, had been arrested for drunkenness6. In the following years Wedin was repeatedly caught for disturbing the order by his drinking, fighting and stealing, as was his younger brother, Frans Walfrid7. This probably explains why Wedin failed to find steady employment in the 1870s and ended up as a sailor traveling around the world. It seems that he took his troubles with him, however, as he left life on Earth after a fight during the summer of 1884. He was shot by an Icelandic man in Ohio, in North America.

  • 8 Carl Jacob Andersson (DDB-ID: 852001372). Prison registers of Västernorrlands läns kronohäkte 1868- (...)

3The Wedin case illustrates the fragility of the family background in which Jonas himself and his siblings were raised and rooted. Even though this was probably a partial reason for their incarceration and experiences later in life, more fortunate pathways are distinguished among young offenders. Carl Jacob Andersson, born in Sundsvall in 1852, epitomizes one of these8. When his stepfather, a day laborer, died in 1865, his stepmother was left alone with three children. Three years later and aged 16, Andersson was incarcerated for vagrancy, but this did not stop him from taking up employment as a machinist and marrying the daughter of a farmer. He even became a sea captain and the father of three children. In the end, it looks as if things worked out for the parentless boy who broke the law and was labeled a criminal in his teens.

4This article explores whether young lawbreakers generally shared Wedin’s or Andersson’s pathway. Neither plausible reasons for their misconduct nor the legislation that affected the definition of crime are focused on. Their types of crimes and demographic backgrounds are only looked at to frame the major analysis, which is to study the level of social exclusion among young people who were arrested and thus labeled offenders. To expand our knowledge of past misconduct, their experiences of life upon release are investigated with regard to four demographic events: relocation, marriage, upward occupational mobility and death. The trajectories of average young people are analyzed to help explore whether offenders’ pathways reflect the negative consequences associated with the label of criminal.

  • 9 For an overview and some critique of both traditional concepts and the application of those associa (...)
  • 10 See the introductions of Farrington (2005b), Thornberry (1997b), and the contributions in the two b (...)

5Labeling perspectives emerged almost fifty years ago as a response to the positivism that had then invaded more traditional concepts in criminology9. These perspectives comprise a number of ideas, some of which emphasize the stigma process. Scholars have frequently employed the labeling concepts and disputed their application, but empirical tests of past criminality and offenders are rarely conducted more systematically. Applying life-course perspectives shows that this article is also inspired by the new theoretical territory in criminology represented by David P. Farrington, Terence P. Thornberry and colleagues.10 Since the 1980s, the criminologists have challenged developmental and life-course (DLC) theories. Similar to the labeling concepts, they unfold multi-dimensional perspectives that nevertheless share some common features of interest for this study. Without neglecting structural impacts, DLC concepts emphasize the individual level of analysis to understand criminal behavior and why it changes over a lifetime. To distinguish behavioral differences, most DLC theorists take on comparative and longitudinal views of individuals’ life courses. However, instead of considering the causes of criminality as most DLC theories do, this labeling approach raises developmental perspectives on those who bore the consequences of their criminal action and were defined as offenders.

1.1. Theoretical considerations incorporating labelling and developmental perspectives of offenders

  • 11 Since the nineteenth century a comprehensive body of literature on criminal behavior and its reason (...)
  • 12 Howard S. Becker (1963), Erving Goffman (1963) and Edwin M. Lemert (1951) were among the first to d (...)
  • 13 Such labeling theories are discussed in Le Blanc (1997, pp. 263-278); Farrington (1977); Paternoste (...)
  • 14 Muncie (1999, p. 310).

6With structural factors linked to economic modernization and undermined relationships in mind, scholars have long developed theories of urban anomie, social disorganization and gradual lack of social control to explain nineteenth-century crime11. These structural concepts proved insufficient for understanding criminal behavior, and in the 1960s the labeling theorists proposed alternative ways to approach crime and the complexity of human relationships12. Broadly conceived, labeling perspectives consider how and why some individuals come to be defined as criminals by the law and others as well as how this definitional process changes over time, and discuss what happens to people who are so ascribed. Some labeling theorists argue that a degraded status characterizes those who are labeled and suggest that this stigma is difficult to overcome because it generates negative consequences concerning social networks and occupational options13.This latter perspective is emphasized in this survey, as it explores what happened to those afflicted upon release. Being arrested manifests the primary deviance, which might cause a secondary deviance implying a stigma that negatively affects how those subjected to it view themselves and how others treat them. Social exclusion is the result of this process, but is difficult to pinpoint. John Muncie assesses the social side of this and argues that it is “a broad concept which refers not simply to the way in which the poor are marginalized from the economic mainstream, but which emphasizes isolation from relationships”14. In this context, those Muncie calls poor are offenders under arrest. Analyzing their socio-spatial mobility, marital and life chances upon release indicates their level of social exclusion and how other people viewed them, particularly if their demographic experiences are compared to those of non-offenders.

  • 15 Farrington (2005b) and Thornberry (1997b). See also the contributions in the two books on this topi (...)
  • 16 Elder (1985, quote p. 17). For an overview of the life-course concept, see Elder (1985); Giele, Eld (...)

7A turning away from a structural to an individual level of analysis has paved the way for incorporating the DLC perspectives into criminology15. The life-course concept has emerged in other fields as well; Glen H. Elder was the first to define it in terms of “pathways through the age differentiated life span”16. These pathways are also labeled trajectories. From cradle to grave, they uncover a line of development that involves different phases such as childhood, adolescence, education, working life, adulthood and parenthood, all of which affect human behavior. Life events such as getting a new job, committing a crime, marrying or moving away intersect this line of development until death breaks it. Although these events are embedded in the trajectories because individuals’ propensity to be involved with them varies with age, life events also depend on human agency and the available opportunities and strains located in the structural setting. As a result and marked by these life events, most individuals pass through a sequence of transitions. Under the impact of long-term trajectories and the social surroundings, these transitions evolve over shorter time spans. Altogether, these complex circumstances shape individuals’ life courses. This is true for offenders too, but their pathways might take different directions if incarceration implies a stigma.

  • 17 Graff (1977, pp. 477ff). Other scholars also emphasize the advantages of focusing on the characteri (...)
  • 18 Loeber, Le Blanc (1990, quote p. 421).

8Aware of the limits of structural approaches, and inspired by the potential of computerized population data that were making headway in the 1970s, Harvey J. Graff discussed the need to form a ‘collective portrait’ of offenders based on their socio-economic and demographic characteristics17. He argues that knowing this and the offenders’ history helps to explain criminal behavior. Graff’s belief that individuals’ features reflect their life experiences and indicate causes of criminal behavior is fruitful, and is further developed in this article. It suggests that offenders’ behavior upon release shows how they experienced life. Whether imprisonment had any negative impact on their pathways is explored with regard to the four demographic events: marriage, migration, mortality and career. These events indicate offenders’ social bonds to society and institutions such as work and family. If individuals, because they led lives that included imprisonment, were markedly marginalized by society and people around them, they would be less fortunate on the labor and partner markets. A possible stigmatization would also be indicated by untimely deaths and limited careers among offenders, or manifested in a desire to relocate after incarceration to settle where their criminal past was unknown. If these findings appear, they confirm the stigmatizing effects linked to the labeling theory referred to. The labeling theory fits well into this approach, as it is “truly developmental in nature because of its explicit emphasis on processes over time”18. To find whether such a process operated among offenders, some of the dynamic DLC concepts addressed in recent criminology are incorporated. The methodological issue of how this is done is discussed below.

1.2. A matched data set, definitions and methodology

  • 19 The court books of the area under study were destroyed by the fire that hit the town of Sundsvall i (...)
  • 20 Concerning the gap between actual criminal behavior and registered criminality that is particularly (...)
  • 21 Although Swedish parish registers are unique in delivering continuous data, gaps exist because they (...)
  • 22 Concerning the identification process of the offenders in the two sources, see footnote 39.
  • 23 Cf. footnote 41. The following circumstances have helped to delineate the age interval of this stud (...)

9Prison registers are utilized to uncover offenders in the period 1840-188019. These registers recorded individuals who were arrested and informed the authorities about the nationwide situation concerning delinquency. Beyond this, some misconduct escaped the law or was corrected without consulting the authorities, but to what extent this happened remains unknown20. Prison registers identify by name people who were arrested and indicate their type of crime.They also contain information on date of birth, parish of birth and residence. Longitudinal data such as the Swedish parish registers comprise the other source. As this source gives a general picture of people’s pathways, through combining it with the prison registers it is possible to suggest how offenders experienced incarceration21. The time-consuming task of combining the two types of data was facilitated by the Demographic Data Base (DDB) at Umeå University, which has digitized and linked the parish registers for the area under study. As a result, 320 offenders with sufficient data available in the two sources are considered below and gathered in the ‘criminal cohort’22. All of these offenders were unmarried and aged 15-25 when arrested in 1840-188023.

  • 24 The extra-informative church examination lists have been utilized to exclude criminals among the co (...)

10The DDB’s digitalization made it possible to design a control cohort that matches the criminal cohort at the individual level. It is constructed to show the ­general pathways of those who were not arrested and includes 3,864 individuals24. Following the ‘control cohort’, ‘average people’ and ‘non-criminals’ refer to these individuals. For comparative reasons, each offender has a number of paired ‘controls’ (from two to 60) with whom he/she shared the following demographic features on the occasion of the offenders’ incarceration: gender, year of birth, marital status (unmarried), residence (either the town of Sundsvall or same neighboring parish). Having these characteristics under control helps to discern whether offenders adjusted to life and society upon release, or whether they were somehow ­marginalized.

  • 25 In principle this is so, but as the offenders must also be identified in the parish registers to be (...)
  • 26 Cf. Cavender (1995, pp. 352ff); Williams (2004a, pp. 374-379); Lindgren (1998, pp. 73-76).
  • 27 Breslow (1992); Breslow, Day (1980); Lynn, McCulloch (1992). The construction of this control cohor (...)
  • 28 The types of individual characteristics most suitable for constructing a matched control cohort and (...)
  • 29 For studies using matched historical and modern demographic data, see Broström (1987); Brändström ( (...)
  • 30 Glueck, Glueck (1950); Laub, Sampson 1998; Sampson, Laub (1997a; 1997b).
  • 31 Agnew (1997, pp. 101-103); Farrington (2005c, pp. 74ff); Sampson, Laub (1997b, pp. 133ff).

11What constitutes an offender has been debated. In this study, being an ‘offender’ or a ‘criminal’ is equivalent to being acknowledged as such in the prison registers25. Incarceration is an official act sanctioned by legislation and is thus made evident to other people. This definition marks the ‘primary deviance’ and connects to labeling theorists who consider individuals to be criminals if they are defined as such26. The stigma and level of social exclusion linked to the labeling idea of ‘secondary deviance’ is easier to distinguish if the findings regarding offenders are related to more conventional pathways, represented here by the control cohort. Matched data analyses first came into practical use in health surveys to evaluate the effects of medical treatment and have aroused interest among statisticians27. To avoid random effects and possibly examine fairly unusual features, such as being a criminal here being indicated by incarceration, there is a need for access to a large matched sample28. In historical research these approaches are rare because they demand sufficient longitudinal data and some statistical skills29. One sociological study of offenders and their pathways has inspired this survey. Robert J. Sampson and John H. Laub elaborate on the ways to use matched data and incorporate life-course perspectives on offenders and non-criminals by reconsidering the classical study of Sheldon and Eleanor Glueck, Unraveling juvenile delinquents (1950)30. Sampson and Laub show how to methodologically operationalize developmental theories and empirically test them. In accordance with these theories, within-individual changes over the life course are explored, as are between-individual differences, by comparing the pathways of offenders with those of non-offenders31.

  • 32 For a presentation of advanced statistical survival analyses, see Cox (1972); Cox, Oakes (1984); Ka (...)
  • 33 Thanks to the growing access to databases that store past quantitative material, historical demogra (...)

12Focusing on death, career, migration and marriage and linking these four events to individuals’ demographic and criminal characteristics are keys in this study, but are complicated to grasp methodologically. Statistical survival analyses exemplified by Cox regression models are employed and further discussed in section 3, as these help to explore the individual consequences of imprisonment and allow life-course perspectives32. These perspectives emphasize biographical continuity, which links criminals’ past with their present situation and future pathway. Societal changes such as industrialization are studied but not centered on, as these are all mediated to the criminals as to others through networks such as families and institutions related to work, gender and location that altogether shaped their existence. Life-course perspectives and the longitudinal event-history analyses they demand thus include both the individual and societal levels of change and events related to these two levels33. Regardless of whether being labeled an offender changed one’s life for the worse or originated from large-scale changes beyond individual choice, the four demographic events under study are viewed with respect to this label but also other attributes and time-space features that characterized those at risk of experiencing these events, hence all individuals in both cohorts. If a stigma was markedly associated with incarceration, it would change the direction of offenders’ life courses and invite further crises in accordance with the labeling theory referred to above.

13For students of crime within any scientific field, the theoretical aspects encompassed by the DLC concept are most relevant for designing research. Issues of measurements and the tools for empirically testing its multi-dimensional aspects are much discussed, however, and therefore this theoretical territory is still developing. It is largely up to the researcher him/herself to find sufficient means of methodology and select the DLC aspects most interesting to address with regard to the purpose and available sources. Applying knowledge developed in disciplines beyond history and using a matched data set of people who are long gone is one way to deal with these theoretical and methodological issues. Of course, historical sources are limited compared to more recent material that can describe the physical and mental status of modern criminals. However, past data is sufficient if it uncovers substantial portions of people’s life courses. The parish registers of Sweden do this.

2. Crime among young offenders in the sundsvall region and their demographic backgrounds

  • 34 The results presented in section 2 are more thoroughly discussed in Vikström (2006; 2008a).

14A brief look at the criminal development among young individuals in the area and period under study is necessary for the continuing approach, as is distinguishing their offences and backgrounds34. Knowing this helps to build on the demographic portrait of young offenders and is fundamental for the approach of their pathways upon release, as it also includes their backgrounds. The results of these longitudinal analyses are discussed in the next section.

  • 35 Alm Stenflo (1994, pp. 37-47; Tedebrand (1996; 1997); Vikström (2003, pp. 69-79).

15In the mid-1840s, the town of Sundsvall was a semi-agrarian town characterized by fishing, handicraft and shopkeeping. Whereas only about 2,500 inhabitants lived there at the time, this figure had grown to more than 8,000 by the mid-1870s. Free trade and an abundance of natural resources stimulated the industrial breakthrough, as did the growing demand for timber in Europe. No sawmills were established in the town but frequently emerged in the industrial hinterland, which witnessed a dramatic population increase as well, from about 7,000 in the 1840s to 20,000 in the 1870s. Alongside the pauperization process in the countryside, migrants arrived and stimulated population growth35.

2.1. Increasing number of incarcerations but continuity in crime over time

  • 36 The legislation is discussed in Inger (1982, pp. 234-48); Wallén (1973, pp. 5-16). Concerning the ‘ (...)
  • 37 Emsley (1987, pp. 18-47); King (1998, pp. 133-137); McDonald (1982); Taussi Sjöberg (1981, pp. 19ff (...)
  • 38 Young individuals were largely involved in the general rise of crimes during the second part of nin (...)

16Figure 1 indicates what this development implied in criminal terms among young people in the Sundsvall region. Curve A shows the total number of incarcerations, not the number of persons arrested. Similar to industrializing countries abroad, young people were increasingly being imprisoned, particularly after the industrial breakthrough this region witnessed in the mid-1860s. This numerical increase is not necessarily equal to higher crime rates. To deal with the current ‘social issue’ and discipline lawbreakers, the revised legislation of 1864 stressed punishment to prevent crime36. Scholars surmise that a more rigorous registration of misconduct was linked to growing intolerant attitudes toward criminal behavior, although it largely remained the same over time37. Figure 1 supports such continuity when the increase of reported crimes is compared to the total population aged 15-25 living in the Sundsvall region (curve D)38.

Figure 1. Number of incarcerations of young people aged 15-25 according to the prison registers for the period 1840-1880, compared to the number of young offenders identified in the parish register, and in relation to the young population (aged 15-25) living in the region and period under study.

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.

  • 39 There are several reasons why not all young offenders are described in detail. First, some did not (...)

17Figure 1 also shows the first incarceration of the 406 individuals identified in the parish registers (curve B), and that of those qualified for this study (curve C). Recidivism explains why the number of identified offenders (406) is not equal to the total number of offences (714). Another reason consists of under-registration in the parish registers and therefore problems with identifying some prisoners39. As this longitudinal study needs good data and only explores unmarried offenders, the discussion below considers the 320 young individuals included in the criminal cohort.

Figure 2. Type of misconduct among young offenders arrested during the period 1840-1880, their geographical background and parental access. Percentages within each category (N = 320).

Figure 2. Type of misconduct among young offenders arrested during the period 1840-1880, their geographical background and parental access. Percentages within each category (N = 320).

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Database, Umeå University.
Explanations: Concerning the crime categorization, see footnote 40.
1. Violence – primarily physical assault and breach of peace, including a few homicides
2. Theft – primarily petty larceny and property offences, including a few advanced economic crimes (forgery, fraud, swindle)
3. Disorder – disturbing behavior such as drunkenness, vagrancy, street disorder, verbal assault
Comments: Native offenders (2a) had never registered their residence elsewhere prior to their incarceration. The category of intra-regional migrants (2b) includes individuals who had left one regional parish for another, including the town of Sundsvall. Immigrants and those who departed from places outside the Sundsvall region are gathered in the category of inter-regional migrants (2c). The parents of offenders (categories 3a-b) usually lived in the same parish (about 90%).

2.2. Crime characteristics of the offenders and their demographic features when arrested

  • 40 In general, scholars separate violent crimes from those related to property or disorder, cf. Taussi (...)
  • 41 According to Swedish statistics in 1857, only one delinquent in ten among those charged and convict (...)
  • 42 Gender and criminality have aroused increasing interest among scholars in the past few decades. For (...)
  • 43 Gendered reasons for which men and women were arrested and with regard to their demographic past ar (...)

18Based on the three modified categories consisting of violence, theft and disorder, Figure 2 shows the types of crimes the 320 offenders had committed on their first occasion of incarceration40. Although a few were charged with homicide or advanced economic crime, the offending majority were arrested for minor misconduct such as physical assault, petty property offence and disorderly behavior. Only 10% were women, and one explanatory factor for this is the gendered expectations41. The sharper control associated with the female gender reduced women’s options to enter the public sphere where most crimes took place, or at least were detected42. As a result, there were fewer women than men to arrest. However, whereas only one male offender in five was arrested for property offences during the period under study, this held true for every second female offender. This indicates that women confronted difficulties with finding means of subsistence43.

  • 44 Ellis (1890); Emsley (1987, pp. 78-102); Monkkonen (1981); Tobias (1967); Zehr (1976, pp. 20-29).
  • 45 At the turn of the nineteenth century, native-born Canadians dominated among recidivists in Ontario (...)
  • 46 In the Sundsvall region, about 40 kilometers separated the most remote parishes from one another. I (...)

19Figure 2 contradicts the notion that criminality flourished among newcomers in unfamiliar settings44. Not even one offender in three had traveled a lengthy distance to reach the Sundsvall region (bar 2c)45. The majority was rooted in the Sundsvall region, which comprised a tiny area in terms of Swedish geography, and thus lacked any ‘disruptive’ experience of relocating (bars 2a-b)46. This characteristic did not make the offenders differ from average young people, as about 60% of those in the control cohort also originated from this region.

  • 47 Today most scholars emphasize such ties, especially in times of urban-industrial change, and these (...)
  • 48 Spouses, parents and children are the only type of network the parish registers acknowledge, except (...)
  • 49 Drugge, Lindgren (2001, pp. 135ff); Muncie (1999, pp. 207-219); Sampson, Laub (1997a, pp. 18, 64-98 (...)

20The criminals’ pronounced regional background suggests that they had social networks at close distance, and Figure 2 confirms this (bars 3a-b)47. The parish registers illuminate the offenders’ family ties represented by fathers and/or mothers living in the region. All were alive and formed a potential resource when their children were arrested48. This held true for about seven offenders in ten, and a similar extent of parental ties also characterized average youths in the region. Students of crime assess the informal control imposed by such ties49. For the young offenders, these ties were of little help in keeping them away from imprisonment; thus the actual functioning of this ample access can be disputed. At least this result rejects the notion that fragile parental ties, or none, provided less control whereas considerable ones offered sufficient emotional and socio-economic support to avoid incarceration. However, among the offenders, some had only access to one parent and usually the mother, just as Wedin and Andersson did when they were arrested. Average youths had slightly larger access to both parents (50 versus 44% among the offenders).

Figure 3. Social status of individuals and their fathers when the young offenders were arrested during the period 1840-1880: a comparison between the criminal and control cohorts (percentages).

Figure 3. Social status of individuals and their fathers when the young offenders were arrested during the period 1840-1880: a comparison between the criminal and control cohorts (percentages).

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Explanations: Upper social strata
1. Large-scale business entrepreneurs
2. Higher civil officials
Middle social strata
3a. Small-scale entrepreneurs in trade and industry, master artisans
3b. Farmers, tenant farmers
4. Lower civil officials
Lower social strata
5. Skilled laborers, craftsmen and artisans below the rank of master
6a. Unskilled laborers in industry and urban commerce
6b. Agricultural workers, farmhands and maidservants

  • 50 Cf. Boritch (2005); Emsley, (1987, pp. 48-77); Lindgren (1998, pp. 48-54).
  • 51 Most scholars agree that occupation can indicate a person’s status and they have also developed dif (...)
  • 52 Similar results appear among delinquent boys in nineteenth-century France, see Berlanstein (1979, p (...)

21Contemporary social observers believed that criminals were recruited from the lower social strata, and a long time scholars shared this belief50. Focusing only on the occupational status of young offenders would be misleading, because their youth places most of them beyond social classification or at the bottom of it, since they were not yet established on the labor market. To illuminate their origins, the occupational status of their fathers is also analyzedinFigure 351. Unskilled laborers, including sailors and apprentices and to some degree skilled laborers, were over-represented among the young offenders, as were especially boatswains and journeymen. Compared to average youths,those incarcerated were allocated to higher levels of the social hierarchy to a lesser extent. This difference is most evident among the fathers. Nevertheless, and although the status is not available for about 50% of all the individuals, there is little evidence of that those under arrest were recruited from only the bottom of the social strata52.

3. Life upon release

  • 53 This happened to almost 75% of them. Only 6% were sentenced to spending more than one year in penal (...)

22To approach the individual price of crime, this section incorporates the offenders’ pathways. Whether incarceration had any impact on their life courses is explored with regard to marriage, migration, mortality and career. Table 1 shows briefly what happened to the offenders during the time span under study, consisting of a maximum of 15 years (cf. below). About 40% of all persons in the two cohorts remained in their parish, whereas almost 50% migrated. The criminal label did not change this. As only few of the offenders were punished for more than one month, they seldom left to serve extended penalties. Most were told to pay fines they could not afford and were therefore imprisoned for some time53. However, this possibly triggered the frequent departures of the few female offenders, who often met an untimely death. These findings indicate that women were confronted with difficulties they either wanted to move away from or died from, as a result of hardship originating from their criminal label and subsequent stigma.

  • 54 That the socio-economic and cultural backgrounds of individuals affect their pathways in terms of r (...)
  • 55 The number and type of independent variables (covariates) are restricted to the quality of the data (...)

23To further illuminate differences between the cohorts and genders, all individuals must be more explored in multiple statistical analyses such as Cox regression models.These models help to evaluate the impact of incarceration by considering demographic characteristics that also can shape the life courses such as age, gender and type of residence (the town of Sundsvall or the surrounding region). Such features are termed ‘covariates’ (or independent/explanatory variables). Other covariates considered are socio-geographical background and access to parents, as these can also influence individuals’ chances in life in terms of socio-cultural capital54. In subsequent models, all these features are usually defined for all individuals in the two cohorts at the moment the offenders were arrested. Whether this occurred prior to the industrial breakthrough or after, when the revised legislation was introduced in 1865, is accounted for, as societal structures implied by the time period can affect individuals’ trajectories. Controlling for all these covariates makes it possible to examine the effect of the characteristic that those in the criminal cohort did not share with their non-offending peers in the control cohort, i.e. being incarcerated and labeled an offender55.

Table 1. Percentage of duration, departure and death within the time interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the genders and the criminal and control cohorts.

Percentage of duration, departure & death (of all within each category)

The criminal cohort

The control cohort

  

All

Men

Women

All

Men

Women

1. Duration (15 years after incarceration)

40.3

41.6

28.2

41.9

41.9

40.5

2. Departure

49.7

49.0

56.2

49.0

48.4

52.8

intra-regional destination

(14.4)

(13.9)

(18.6)

(14.2)

(13.4)

(17.8)

inter-regional destination

(28.1)

(27.1)

(37.6)

(28.6)

(28.4)

(30.1)

destination abroad

(7.2)

(8.0)

(6.2)

(6.6)

(4.9)

3. Death

10.0

9.4

15.6

9.1

9.4

6.7

Total (%)

100

100

100

100

100

100

Total (N)

320

288

32

3864

3116

748

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Comments: The category of duration includes individuals still residing in the same parish 15 years after the longitudinal analysis begins, which is equal to the incarceration date of the criminals. For comparative reasons this date also defines the start of the analysis of those in the control cohort to which each offender is linked. The three subcategories within the category of departure, presented in italics plus brackets, show how the departure percentage is distributed according to type of destination.

  • 56 To specifically analyze the pathways of those arrested twice or more, there is a need to have acces (...)
  • 57 The time interval of fifteen years was constructed for comparable reasons, as the parish registers (...)

24The date of the offenders’ first imprisonment forms the entry of the longitudinal analysis of the two cohorts. Among the offenders in the criminal cohort, 25% were to be arrested once again or more, but in this case their first imprisonment marks the criminal label, which initiates the longitudinal observation of all individuals in the two cohorts56. It continues for as long as they stay on in the parish, but not more than fifteen years57. Meanwhile the four events of marriage, migration, mortality and upward social mobility are examined. With event-history analyses these transitions are explored by including the length of time between entry (incarceration) and exit from a specific state, i.e. from unmarried to married, from residing to relocating, etc. For each event a threefold regression model is applied, starting with one considering all individuals in the two cohorts. The two other models examine the genders separately. The results of Cox regression models are presented as risk ratios, which show the probability of experiencing the event under observation and relate this to the individuals’ different characteristics.

  • 58 Cf. Glueck, Glueck (1950); Sampson, Laub (1997a).
  • 59 Separating the criminal cohort into three crime categories of violence, theft and disorder jeopardi (...)

25With the help of these tools that corroborate those of Sampson and Laub, it is possible to analyze whether incarceration manifested a ‘turning point’ in the trajectories of those labeled offenders because they were socially excluded. If so, untimely deaths or relocation would characterize them upon release, as would limited careers and marital chances. When these findings appear, they confirm the stigma process of being termed offender that label theorists assess. Of course, other key variables that describe persons, such as educational level, psychological character and peer networks, influence human pathways, but such information is impossible to uncover systematically in historical material58. Whether the type of misconduct affected offenders’ life upon release is discussed in another article59.

3.1.Marital chances among the young offenders and average young people

  • 60 Hajnal (1965). Cf. Nilsson, Tedebrand (2005, pp. 15-44); Vikström (2003, pp. 185ff); Laslett (1977) (...)
  • 61 Sampson, Laub (1997a, pp. 178-245; 1997b, pp. 145-152). Adams (1997, pp. 328-332) and Matsueda, Hei (...)

26Marriage was the life plan of most young people in nineteenth-century society, although John Hajnal’s ‘European Marriage Pattern’ illuminates late age at marriage and a high proportion of celibacy60. This predicted pathway is perceived as indicating the offenders’ adjustment to ‘normal’ life, and it was also encouraged by parents and institutions such as the church and the government. Sampson and Laub find a correlation between marriage and desistance from crime, because according to their age-graded theory of informal social control, the incentives to commit crime decrease with increasing age as solid bonds to social networks are being established along the life course61. This socialization process is marked by marriage. Given the criminal label, however, there is reason to assume that lower chances to unite with a spouse separated offenders from others.

  • 62 Alm Stenflo (1994, pp. 64-68); Lundh (1997, p. 10); Vikström (2003, pp. 188-191).

27As not everyone found a spouse and duration of residence influenced this event, Figure 4considers the time between incarceration and wedding with regard to the two cohorts. The larger and more rapid decline of the curves, the more marriages are found within the time interval considered. No sharp difference between the cohorts appears, but about ten years after the incarceration the gap between the curves grows. By then the delinquents and juveniles in the control cohort had entered the phase in life when marriage usually occurred in Sweden and the area under study, on average at the age of 27-2862. It seems as if offenders approaching this age were not as likely to marry as were their non-criminal peers.

Figure 4. Life table of the time between incarceration and marriage within the interval ­covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.

Figure 4. Life table of the time between incarceration and marriage within the interval ­covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Comments and explanations: Life tables illuminate the survival chances by considering the event of death with regard to the period under study. In this case, marriage replaces the event of death, whereas remaining unmarried is equal to survival. The larger and more rapid the decline of the curve, the more marriages are found. Within ten years of the time span under study about 55% of the individuals in both cohorts had experienced this transition. The life table shows that a difference between the criminal and control cohort appears, as the young individuals had reached the phase in life when most marriages occur. The Cox regression model of Table 3 indicates that this difference is significant at the ten-percent level.

28Table 2 develops these findings by emphasizing that the impact of misconduct on marriage must be explored with regard to the demographic features discussed above, not least gender. Model 1 shows that the offenders were almost 17% less likely to marry than were average young people. Models 2 and 3 suggest that misconduct hindered women in particular from becoming wives. Those charged were 36% less inclined to experience this event compared to women in the control cohort. Male offenders were only 13% less likely to marry than were average men. However, characteristics other than having a criminal label were more significant. Because men dominated the partner pool in this sawmill region that mainly attracted male laborers, women were generally more inclined to marry. Compared to younger individuals in the two cohorts, those in their mid-twenties were more likely to do so, as they were closer to the average age when marriage occurred. The type of residence was also of greater significance for this event than was imprisonment or parental access. Those living in the town of Sundsvall were half as likely to experience a wedding as were those residing elsewhere in the region.

Table 2. Cox regression of the time between incarceration and marriage within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the genders and the criminal and control cohorts. (Risk ratios presented in boldfaced italics = P-value below 5%, boldfaced = P-value below 10%)

Covariates
Characteristics of the two cohorts

Three Cox regression models of the two cohorts

Demographic features of all individuals when the offenders were arrested

Model 1
P-value (0.000)
All individuals
(N = 4,184)

Model 2P-value (0.000)Men
(N = 3,404)

Model 3P-value (0.045)Women
(N = 780)

  

P-value

Risk ratio

P-value

Risk ratio

P-value

Risk ratio

1. Criminal label (ref: The control cohort)
The criminal cohort

0.070

0.834

0.173

0.865

0.152

0.641

2. Sex (ref: Women) – Men

0.000

0.614

3. Age (ref: 20-25) – 15-19

0.000

0.829

0.001

0.808

0.820

0.974

4. Time period (ref: 1865-1880)
1840-1864

0.314

0.944

0.406

0.947

0.775

0.964

5. Residence (ref: The surrounding region)
The town of Sundsvall

0.000

0.516

0.000

0.470

0.018

0.719

6. Geographical background (ref: Native)

0.086

  

0.549

  

0.034

  

From abroad

0.578

1.145

0.757

0.909

0.126

1.855

Inter-regional background

0.017

1.183

0.124

1.132

0.024

1.389

Intra-regional background

0.042

1.152

0.387

1.070

0.011

1.467

Unknown

0.124

1.243

0.419

1.136

0.053

1.864

7. Parental access (ref: Both parents in the region)

0.240

  

0.294

  

0.433

  

No parents available in the region

0.178

0.856

0.142

0.825

0.841

1.051

One parent available in the region

0.932

0.992

0.609

0.948

0.266

1.247

8. Socio-economic status
(ref: 6b. Agricultural workers)

0.000

  

0.000

  

0.454

  

1 & 2. Upper social strata

0.663

0.872

0.904

0.962

3a. Small-scale entrepreneurs

0.070

1.410

0.035

1.514

3b. Farmers

0.197

0.746

0.116

0.696

4. Lower civil officials

0.355

1.165

0.126

1.305

0.753

0.727

5. Skilled laborers

0.021

1.307

0.012

1.372

0.856

0.832

6a. Unskilled laborers

0.000

0.644

0.000

0.643

Unknown, unspecified

0.000

0.746

0.001

0.732

0.107

0.786

9. Socio-economic origin (status of the father)
(ref: 6b. Agricultural workers)

0.354

  

0.025

  

0.304

  

1 & 2. Upper social strata

0.041

0.660

0.015

0.560

0.854

0.926

3a. Small-scale entrepreneurs

0.518

0.911

0.072

0.732

0.238

1.410

3b. Farmers

0.840

1.022

0.996

1.001

0.706

1.110

4. Lower civil officials

0.606

1.087

0.882

0.972

0.345

1.359

5. Skilled laborers

0.577

1.071

0.231

1.177

0.446

0.804

6a. Unskilled laborers

0.632

1.052

0.533

1.076

0.932

0.979

Unknown, unspecified

0.731

0.960

0.792

0.965

0.400

0.796

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Comments: Concerning the geographical background and social classification, see Figures 2 and 3.
Explanations: Event-history analyses such as Cox regression models show the transition of a specific event by including the length of time between entry (i.e. date of incarceration) and exit (i.e. wedding date). The probability of experiencing this transition is presented in risk ratios. To examine what influenced the likelihood to marry, these ratios are related to the individuals’ characteristics (covariates), such as crime, gender and residence. For certain covariates these ratios also give information on the relationship between comparative subgroups and the reference group, whereby the latter is always defined as 1.00. Concerning the impact of misconduct on the likelihood of marrying, the offenders included in the criminal cohort were about 17% (1.00-0.834=0.166) as likely to experience this transition in comparison to average young people in the control cohort (the reference group). The reliability (significance) of the risk ratio depends on the P-value (probability). The lower the P-value, the more reliable the results. Results when the P-value is above the five-percent level must be considered with some caution, as is the case concerning the impact of the criminal label in the above table. Because it is boldfaced, it is significant below the ten-percent level, however. Risk ratios below the five-percent level are presented in boldfaced italics. The P-value of each regression model employed on all individuals, men or women respectively, is displayed at the top of the table.

  • 63 Of course, a more thorough study of the choice of partner with regard to age and socio-economic sta (...)

29Even if there are indications that incarceration jeopardized marital chances, especially among women, the findings reject the notion that this markedly reduced the prospects63. These were determined by other characteristics, some of which affected the genders differently. In contrast to native girls, female migrants benefited from the regional partner pool, whereas geographical background was insignificant for men’s marriage. For them, age and status were more determinant. Small-scale business entrepreneurs and skilled laborers were most likely to join a bride in church. Unskilled laborers in industry were less likely to do so in comparison to these men and toagricultural workers. According to the sources, men with no occupation experienced the lowest marital chances. Their difficulties on the labor market were echoed on the partner market but cannot be linked to past crime or incarceration.

3.2. Leaving in the name of shame?

30Relocating could be one strategy for criminals to escape a bad reputation and the place where their transgression was known. To find out whether the offenders’ migratory behavior differed on this plausible ground, the time between incarceration and departure is explored.

Figure 5. Life table of the time between incarceration and departure within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.

Figure 5. Life table of the time between incarceration and departure within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Comments and explanations: Cf. Figure 4

31The parallel curves of the two cohorts in Figure 5 suggest that offenders were not more motivated to migrate. Table 3 confirms this and shows that gender also influenced the incentive to relocate to a limited extent. Instead, young age and geographical background determined this event. Native regional residents and those with access to both parents were least interested in moving. Such factors downplayed the labeling impact, as did the individuals’ socio-economic status as well as that of their fathers, especially for men. Skilled laborers, mostly journeymen, were the most mobile. ­Children of agricultural workers were least likely to leave. After the industrial breakthrough in the mid-1860s, all individuals were more inclined to relocate. Hence, neither incarceration nor gender but instead other characteristics affected migration.

Table 3. Cox regression of the time between incarceration and departure within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the genders and the criminal and control cohorts. (Risk ratios presented in boldfaced italics = P-value below 5%, boldfaced = P-value below 10%).

Covariates
Characteristics of the two cohorts

Three Cox regression models of the two cohorts

Demographic features of all individuals when the offenders were arrested

Model 1
P-value (0.000)
All individuals
(N = 4,184)

Model 2P-value (0.000)Men
(N = 3,404)

Model 3P-value (0.000)Women
(N = 780)

  

P-value

Risk ratio

P-value

Risk ratio

P-value

Risk ratio

1. Criminal label (ref: The control cohort)
The criminal cohort

0.480

1.061

0.489

1.063

0.741

1.085

2. Sex (ref: Women) – Men

0.562

0.964

3. Age (ref: 20-25) – 15-19

0.000

1.234

0.000

1.253

0.093

1.214

4. Time period (ref: 1865-1880)
1840-1864

0.004

0.857

0.045

0.888

0.034

0.749

5. Residence (ref: The surrounding region)
The town of Sundsvall

0.465

1.040

0.533

1.037

0.614

0.929

6. Geographical background (ref: Native)

0.000

1.465

0.000

  

0.000

  

From abroad

0.054

1.901

0.121

1.434

0.223

1.622

Inter-regional background

0.000

1.881

0.000

1.896

0.000

1.944

Intra-regional background

0.000

1.501

0.000

1.802

0.000

2.270

Unknown

0.005

  

0.001

1.556

0.716

1.159

7. Parental access (ref: Both parents in the region)

0.000

  

0.000

  

0.178

  

No parents available in the region

0.000

1.546

0.000

1.569

0.076

1.524

One parent available in the region

0.903

1.012

0.911

0.988

0.438

1.175

8. Socio-economic status
(ref: 6b. Agricultural workers)

0.000

  

0.000

  

0.469

  

1 & 2. Upper social strata

0.411

0.796

0.485

0.821

3a. Small-scale entrepreneurs

0.008

0.569

0.014

0.587

3b. Farmers

0.202

0.611

0.226

0.624

4. Lower civil officials

0.545

1.085

0.502

1.102

0.256

1.955

5. Skilled laborers

0.000

1.532

0.000

1.562

0.534

1.388

6a. Unskilled laborers

0.219

0.902

0.483

0.936

Unknown, unspecified

0.746

0.977

0.854

1.017

0.460

0.903

9. Socio-economic origin (status of the father)
(ref: 6b. Agricultural workers)

0.000

  

0.000

  

0.177

  

1 & 2. Upper social strata

0.001

1.785

0.002

1.771

0.123

1.973

3a. Small-scale entrepreneurs

0.037

1.383

0.087

1.339

0.148

1.743

3b. Farmers

0.325

0.859

0.444

0.882

0.382

0.662

4. Lower civil officials

0.010

1.567

0.041

1.497

0.089

1.983

5. Skilled laborers

0.000

1.685

0.001

1.666

0.057

1.947

6a. Unskilled laborers

0.024

1.339

0.081

1.279

0.104

1.724

Unknown, unspecified

0.020

1.405

0.061

1.353

0.133

1.714

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Comments and explanations: Cf. Table 2. Concerning the geographical background and social classification, see Figures 2 and 3.

  • 64 Until the poor law of 1847, the parish boards had to approve migrants, and some were not welcome at (...)

32There is little proof that reproachful attitudes confronted the young offenders. Their departure mirrored that of average young people and was due to phenomena other than incarceration. If the offenders were leading a dissolute life, this might have hindered a desired relocation among them, because they either lacked resources to relocate or were not allowed to settle elsewhere, as they might become a burden to the new local authority64. However, the criminal cohort was not more inclined to either relocate or remain. It seems as if the offenders were about as comfortable with life in their parish of residence as were those in the control cohort. Hypothetically, solid social bonds and ample options on the regional labor market contributed to the same length of duration among the individuals, regardless of label and cohort.

3.3. Death differences and similarities between the cohorts and genders

  • 65 Moffitt (1997, pp. 21-23); Sampson, Laub (1997a, pp. 123-138; 1997b, pp. 143ff).

33Perhaps slightly surprising, more similarities than differences between the cohorts have been found thus far. However, Figure 6 shows that a period of arrest at a young age lowered survival chances later in life. Hypothetically, a stigmatization process including hard living conditions limited the life expectancy of those once charged. In combination with the marital chances displayed in Figure 4, this finding indicates that individuals paid some price for being labeled an offender, but years later. They were then less likely to marry and more likely to meet an untimely death. This ‘postponed’ cost of crime suggests that the stigma process accumulated over a lifetime. Life-course scholars term this developmental phenomenon ‘cumulative disadvantage’ or the ‘knifing off’ of future opportunities65. Those affected encounter fewer options for a conventional life, because a bad reputation will weaken their bonds to society and social institutions such as family and work.

  • 66 Brändström, Tedebrand (1993); Edvinsson (1992); Fridlizius (1988); Willner (1999).

34However, Table 4 suggests that incarceration was not the only key to shaping survival odds. Mortality studies show that women in general could expect to live longer than men. Men’s lifestyle, not least their abuse of alcohol and rough working conditions in towns or dirty industries, is used to explain their lower survival rates66. This gendered mortality pattern is reflected among the individuals considered here. The first model of Table 4 shows that men were almost 60% more likely to die. Individuals living in the town of Sundsvall faced a significantly lower life expectancy than did those in the surrounding region. The criminal label was not in a position to affect mortality to a similar degree.

Figure 6. Life table of the time between incarceration and death within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.

Figure 6. Life table of the time between incarceration and death within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Comments and explanations: Cf. Figure 4.

  • 67 Edvinsson (2001); Galley (1995); Kearns (1988).

35Table 1 revealed that female criminals met an untimely death. The two regression models that treat the genders separately provide further clues about how well they and male offenders coped with life upon release. Men with no occupation reported in the sources and those living in Sundsvall experienced the lowest survival chances. Towns were unhealthy places, and people living there exposed themselves to what is sometimes called the ‘urban penalty’67. This especially afflicted the men considered here. Whereas imprisonment was of low significance for their survival, female offenders were more than twice as likely to die as were their sisters in the control cohort, probably because they challenged their gendered expectation. Hence, the knifing-off effect operated primarily among women. For them, achieving the criminal label implied a stigma equal to fewer options to escape cumulative disadvantages. Table 4 suggests that weak bonds were involved in this process, too. Having experienced stable parental backgrounds represented by access to both parents in the region when the observation starts added to women’s survival odds.

Table 4. Cox regression of the time between incarceration and death within the interval covering a maximum of fifteen years after offenders’ incarceration during the period 1840-1880: a comparison between the genders and the criminal and control cohorts. (Risk ratios presented in boldfaced italics = P-value below 5%, boldfaced = P-value below 10%).

Covariates
Characteristics of the two cohorts

Three Cox regression models of the two cohorts

Demographic features of all individuals when the offenders were arrested

Model 1
P-value (0.020)
All individuals
(N = 4,184)

Model 2P-value (0.020)Men
(N = 3,404)

Model 3P-value (0.097)Women
(N = 780)

  

P-value

Risk ratio

P-value

Risk ratio

P-value

Risk ratio

1. Criminal label (ref: The control cohort)
The criminal cohort

0.671

1.083

0.912

0.978

0.060

2.504

2. Sex (ref: Women) – Men

0.003

1.580

3. Age (ref: 20-25) – 15-19

0.754

0.965

0.906

1.015

0.616

0.851

4. Time period (ref: 1865-1880)
1840-1864

0.986

0.998

0.616

1.066

0.454

0.759

5. Residence (ref: The surrounding region)
The town of Sundsvall

0.002

1.459

0.003

1.475

0.996

0.998

6. Geographical background (ref: Native)

0.891

  

0.777

  

0.967

  

From abroad

0.970

1.020

0.944

0.959

0.745

1.415

Inter-regional background

0.547

1.088

0.623

1.078

0.816

1.096

Intra-regional background

0.466

1.115

0.510

1.111

0.481

1.353

Unknown

0.417

1.280

0.222

1.452

0.973

0.000

7. Parental access (ref: Both parents in the region)

0.065

  

0.030

  

0.006

  

No parents available in the region

0.380

0.805

0.039

0.574

0.012

5.022

One parent available in the region

0.309

1.217

0.903

0.975

0.002

3.877

8. Socio-economic status
(ref: 6b. Agricultural workers)

0.193

  

0.015

  

0.998

  

1 & 2. Upper social strata

0.570

0.658

0.720

0.764

3a. Small-scale entrepreneurs

0.132

0.402

0.251

0.461

3b. Farmers

0.349

1.576

0.213

1.874

4. Lower civil officials

0.548

0.788

0.910

0.953

0.991

0.000

5. Skilled laborers

0.552

1.168

0.188

1.477

0.989

0.000

6a. Unskilled laborers

0.654

0.907

0.606

1.141

Unknown, unspecified

0.207

1.265

0.068

1.535

0.852

1.078

9. Socio-economic origin (status of the father)
(ref: 6b. Agricultural workers)

0.699

  

0.625

  

0.153

  

1 & 2. Upper social strata

0.647

0.841

0.604

0.814

0.980

1.032

3a. Small-scale entrepreneurs

0.944

1.020

0.832

1.063

0.641

0.558

3b. Farmers

0.195

0.712

0.182

0.696

0.960

0.950

4. Lower civil officials

0.242

0.622

0.252

0.611

0.938

0.907

5. Skilled laborers

0.977

1.008

0.714

0.904

0.269

2.505

6a. Unskilled laborers

0.912

1.026

0.751

0.925

0.256

2.447

Unknown, unspecified

0.950

0.984

0.648

1.136

0.584

0.633

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Comments and explanations: Cf. Table 2. Concerning the geographical background and social classification, see Figures 2 and 3.

3.4. Career after crime and occupational changes

36Previous results propose that most of the offenders under study stayed on and probably continued their work or took up employment after incarceration. The next issue is to determine whether they faced limited occupational options in comparison to average young people because of cumulating disadvantages based on their criminal label.

37Figure 7 displays the likelihood of moving one or more steps upward on the social ladder, just as the machinist Carl Jacob Andersson did, who was initially introduced. Even though his label did not stop him from becoming a sea captain, incarceration in general impeded the career. Table 5 presents some other explanatory factors of upward mobility and emphasizes that the difference between the two cohorts is both sharp and significant. As this analysis requires occupational information on both occasions of measurement, the number of cases is restricted to those for which such data is available. As a result, this discussion primarily concerns men. The status of the few women is linked to that of their husbands, if they married during the time interval investigated. This and their initially low status explain why they were slightly more likely to achieve upward mobility, but this gender difference is insignificant.

Figure 7. Life table of the time between incarceration and upward mobility within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.

Figure 7. Life table of the time between incarceration and upward mobility within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Comments and explanations: Cf. Figure 4.

  • 68 Cf. Brändström, Ericsson (1995); Erikson, Goldthorpe (1992); Kaelble (1984); Leeuwen, Maas (2002); (...)

38Model 1 of Table 5 shows that besides incarceration, the time period and the individuals’ status including that of their fathers influenced their chances to move upward. Those labeled as offenders were almost 40% less likely to experience this event. Industrialization and the structural ‘openness’ often associated with it did not stimulate the upward mobility of either the offenders or the non-offenders. In research, this openness is increasingly recognized as a modernization myth rather than as being based on empirical facts68. The career of those residing in the town of Sundsvall only benefited slightly from the urban environment. Unskilled laborers and agricultural workers were those most likely to rise in the social hierarchy, as were the children of farmers.When the genders are treated separately in Models 2 and 3, the impeding impact of incarceration remains significant for men, but not for women. Among women, parental access was one key factor; those with access to two parents were more likely to rise in the hierarchy. Living in Sundsvall and being a daughter of a businessman also spurred these chances.

Table 5. Cox regression of the time between incarceration and upward mobility within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the genders and the criminal and control cohorts. (Risk ratios presented in boldfaced italics = P-value below 5%, boldfaced = P-value below 10 %).

Covariates
Characteristics of the two cohorts

Three Cox regression models of the two cohorts

Demographic features of all individuals when the offenders were arrested

Model 1
P-value (0.000)
All individuals
(N = 1,901)

Model 2P-value (0.020)Men
(N = 1,611)

Model 3P-value (0.036)Women
(N = 290)

  

P-value

Risk ratio

P-value

Risk ratio

P-value

Risk ratio

1. Criminal label (ref: The control cohort)
The criminal cohort

0.048

0.623

0.038

0.582

0.448

1.609

2. Sex (ref: Women) – Men

0.272

0.847

3. Age (ref: 20-25) – 15-19

0.240

0.855

0.798

0.963

0.335

0.683

4. Time period (ref: 1865-1880)
1840-1864

0.022

1.328

0.010

1.422

0.932

1.030

5. Residence (ref: The surrounding region)
The town of Sundsvall

0.126

1.238

0.465

1.118

0.073

2.749

6. Geographical background (ref: Native)

0.968

  

0.978

  

0.801

  

From abroad

0.881

0.930

0.613

0.735

0.493

1.991

Inter-regional background

0.741

1.065

0.932

0.983

0.990

1.010

Intra-regional background

0.546

1.111

0.772

1.056

0.624

0.701

Unknown

0.884

0.950

0.942

1.026

0.972

0.000

7. Parental access (ref: Both parents in the region)

0.646

  

0.617

  

0.029

  

No parents available in the region

0.106

0.823

0.760

1.088

0.021

0.291

One parent available in the region

0.385

0.818

0.646

0.887

0.881

0.912

8. Modified socio-economic status
(ref: 6. Unskilled labourer and Agricultural workers)

0.000

  

0.001

  

0.999

  

3. Small scale entrepreneurs & farmers

0.000

0.120

0.000

0.127

4. Lower civil officials

0.055

0.584

0.136

0.657

0.984

0.000

5. Skilled laborers

0.703

0.940

0.814

0.962

0.980

0.000

9. Socio-economic origin (status of the father)
(ref: 6b. Agricultural workers)

0.000

  

0.013

  

0.041

  

1 & 2. Upper social strata

0.181

2.024

0.210

1.980

3a. Small-scale entrepreneurs

0.093

2.048

0.346

1.571

0.007

12.385

3b. Farmers

0.005

2.518

0.021

2.341

0.268

2.504

4. Lower civil officials

0.113

2.076

0.144

2.097

0.844

1.255

5. Skilled laborers

0.515

1.266

0.601

1.233

0.632

1.530

6a. Unskilled laborers

0.138

0.602

0.193

0.608

0.383

0.499

Unknown, unspecified

0.072

1.744

0.259

1.506

0.194

2.172

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Comments and explanations: Cf. Table 2. Concerning the geographical background and social classification, see Figures 2 and 3. Upward mobility is equal to occupational changes that cause a change of social group according to the modified classification (cf. Covariate 8 in the above table). Only those below the first two social groups (1 & 2) with an occupation reported when the young offenders were arrested and an occupational notation available within the time span are considered. If women married, they attained the status of their husbands. This and the few females involved in the analysis jeopardize the reliability of the results concerning women.

39Male offenders were thus unfortunate on the regional labor market, and their criminal label definitely contributed to this. Table 6 compares their occupational chances with average youthsand women. Individuals lacking occupational data are also identified, as they could find employment over time. Regardless of gender, the offenders remained socially stable to a larger degree and were less distributed among occupational groups allocated to the higher social strata. This suggests that they encountered employment difficulties because of cumulating disadvantages. However, incarceration generally neither excluded offenders from the labor market nor isolated them at the lowest level of it.

Table 6. Changes by social group in the social strata based on the occupation reported when the offenders were arrested during the period 1840-1880, compared to the last available notation within the time span consisting of a maximum of 15 years: a comparison between the genders and the criminal and control cohorts.

Percentage of

Occupational changes over time between the cohorts and genders

change within

Offenders in the criminal cohort (LO) and non-offenders in the control cohort (NO)

each social group

Men

Women

Social group

3a

4

5

6a

6b

Unknown

6b

Unknown

Cohort type

LO

NO

LO

NO

LO

NO

LO

NO

LO

NO

LO

NO

LO

NO

LO

NO

1 & 2

5.4

1.8

2.9

1.5

1.4

0.3

0.7

2.8

0.4

4.9

3a. Small-scale

50.0

75.0

9.6

5.9

11.9

5.0

3.5

5.9

3.1

1.5

3.3

4.1

7.6

3b. Farmers

25.0

2.6

0.3

1.6

2.9

8.9

6.0

11.2

7.7

2.6

5.1

4. Lower civil

5.4

100

78.1

4.5

1.3

3.2

1.2

1.5

6.0

3.4

5.6

7.8

5. Skilled lab.

8.9

2.6

73.5

68.5

5.0

8.7

8.8

5.5

5.2

7.7

9.3

5.6

7.2

6a. Unskilled

25.0

3.6

3.5

14.7

9.8

85.0

77.9

29.4

33.6

41.8

29.7

23.1

22.4

11.1

15.3

6b. Agricultural

2.9

2.4

1.3

2.7

50.0

47.1

9.7

7.0

61.5

53.4

22.2

10.6

Unknown

1.8

1.2

1.2

2.5

1.1

2.9

0.3

33.6

34.4

7.7

4.5

55.6

41.5

Total (%)

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

Total (N)

4

56

2

114

34

336

80

565

34

327

134

1656

14

268

18

472

  • 69 Among the Gluecks’ male delinquents, Sampson, Laub (1997a, pp. 178-242; 1997b, pp. 145-152) find th (...)

Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
Explanations: Concerning the social classification, see Figure 3.
LO = Labeled offenders, i.e. the young people arrested, the criminal cohort (N=320)
NO = Non-offenders, i.e. the average young people, the control cohort (N=3,794)
Comments: Highlighted cells display the percentage of occupational stability over time for the category under consideration. Figures presented in the same column but above these cells show the percentage of individuals in this category who moved upward. Similarly, figures below the cells showing persistence display the percentage that went downward in the social hierarchy. As so few young individuals were allocated to the upper social strata (social groups 1 & 2), this strata is not considered as regards the horizontal level, which shows the social distribution among all individuals when the offenders were arrested. This has slightly reduced the number of individuals in the control cohort (from 3,864 to 3,794).
Again, there are indications that the offenders had left their criminal past behind, though they were not left entirely unaffected. Their criminal label did not hinder employees from hiring them, just as it did not stop spouses from marrying them.Perhaps people did not know about their incarceration, or cared little about it. DLC theorists provide yet another answer. Ageing involves developing occupational or marital commitment, which also implies socialization and responsibilities. Such circumstances lower one’s incentive to break the law and possibly get caught, and thus favor individuals’ desistance from criminal involvement69. Most likely, these life-course factors made these offenders consider whether to commit crime again or take on a more conventional lifestyle. It seems as if most tried to do the latter, although few managed to match the thriving career of Carl Jacob Andersson.

4. Concluding remarks

40This study examines the pathways of young offenders in a Swedish sawmill region that witnessed rapid socio-economic transformations during the nineteenth century. To apply life-course perspectives on young individuals labeled offenders and explore their lives upon release, prison registers are linked to computerized parish records. The pathways of offenders are thus illuminated to a greater extent than is the case in most other studies concerned with past criminals. In combination with a matched data set of non-criminals, longitudinal event-history analyses of four demographic events (marriage, migration, mortality and upward social mobility) suggest whether offenders adjusted to life and society after incarceration, or were faced with social exclusion as many labeling theorists propose. Lower survival odds and limited chances on the partner and labor markets are assumed to indicate high levels of social exclusion, as is their inclination to relocate due to shame. Below, the most essential findings are discussed and linked to the two concepts of labeling and life course initially introduced.

41What happened to past criminals upon release is especially ignored in historical research due to the limited availability of longitudinal data. According to this survey, most offenders continued life in the Sundsvall region after incarceration and did not set themselves on the road to possibly escape shame. They were not more eager to leave than were average young people, and their pronounced regional background probably contributed to this. Nor did captivity impede their marital chances to any great degree, or cause an untimely death as a result of a stigmatization process. Even though their careers in comparison to those of average youths were less successful, they were not forced to the bottom of the social strata. In all, there is little evidence for utilizing the concept of marginalization when discussing offenders’ adjustment to life and society upon release. To argue that they readjusted is closer at hand, and serves to complete the view of past criminals and their pathways, particularly regarding male offenders.

42This study supports findings suggesting that women were aware of their gendered obligations and seldom practiced criminality, as this would put their honor at stake. If they nevertheless did, female offenders faced low survival chances upon release and were recognized as losers on the partner market. Intolerant attitudes toward them among people and in society at large probably played a part in their difficulties coping with life after imprisonment. In transgressing both the law and rules associated with their gender, they became subjects to double deviance. There are few female offenders to judge from, but they paid a significantly higher price for their criminal label than did their delinquent brothers. Even though women and more unfortunate lawbreakers do not dominate among the offenders, they show that some price was paid in terms of social exclusion but often years later, as was the case concerning the marital and survival chances. When incarceration represented a turning point, the negative effects tend to hit gradually. However, achieving the criminal label did not necessarily initiate the stigma and course of marginalization. This might have accompanied some offenders prior to imprisonment or even formed one reason for it, but being arrested probably spurred the process of cumulative disadvantage. Given this and the fact that about 25% of the offenders were recidivists makes the similarities between their life courses and those of non-criminals seem even more remarkable.

43The life-course perspectives incorporated in this survey show that the criminal label was only one among many determinants that influenced young people’s pathways. Demographic events such as marriage, departure, death and career are affected by multiple factors operating at both the individual and societal levels, which all shape and indicate the quality of human life in history. These factors can downplay the stigmatizing impact of being labeled; the offenders thus did not capitulate if they found themselves arrested. As a result, they were enabled to chart a pathway that echoed that of average people.

44This finding cannot be associated with social exclusion first and foremost, but why? Perhaps those imprisoned learned their lesson and did everything to get back on their feet and avoid future incarceration. If so, the preventive and punitive intention of the control that contemporary social observers praised was achieved. However, about one in four of the offenders fell foul of the law again and young people continued to commit crimes. Regardless of time and space, longitudinal studies show that increasing age generally favors desistance from criminal involvement. This study proposes that this life-course effect held true for most of the young offenders, because their criminal label did not stop them from developing social ties to others and the surrounding society over time. Otherwise, they would have migrated or been markedly disfavored in the partner pool and labor market. In transition from adolescent to adulthood, their adoption of new social roles encouraged their commitment to a conventional lifestyle, including work and family. That the offenders took on trajectories paralleling others further suggests that they were assisted by tolerant attitudes and a prosperous labor market that helped them deal with their situation. It is possible that the substantial regional background and access to family ties governed the offenders in their efforts upon release instead of suppressing them. Additionally, most were young men arrested for minor misconduct. Some drunkenness, disorderly manners and scuffles were part of their lifestyle. This in fact illegal behavior might have escaped the eye of the law, but not that of ordinary people. In contrast to the social observers, they were familiar with this sight and cared little whether it would render in custody. Finally, it must not be forgotten that from all across Sweden and even abroad, newcomers arrived in this expanding region and interfered with local norms. They likely fostered a more tolerant climate. In all, these factors explain why the negative impact associated with the labeling theory remains less obvious than expected.

45These findings suffice to challenge the view of past offenders and the consequences of criminality. Crime has long been linked to social exclusion and future failures among those faced with it. This survey recognizes young people labeled as offenders as individuals with little experience of migration and considerable access to parental ties, thanks to their evident regional background. Nevertheless, they came into conflict with the law and were arrested. This most likely did not help them in coping with their lives upon release, but achieving the criminal label did not significantly alter people’s pathways or ruin their prospects, unless they were women.

Acknowledgments

46I particularly wish to express my gratitude to the Bank of Sweden Tercentenary Foundation (Riksbankens Jubileumsfond) and Professor Ulf Drugge, Kalmar University, Sweden, for a research position in the interdisciplinary project ‘Crisis-biographies: life-course and time-geographical aspects of the causes, cures and consequences of individual crises’. Professor Drugge’s archival efforts and sociological perspectives have contributed to this article, as has funding from Stiftelsen Clas Groshinsky’s Minnesfond. An earlier version was presented at a session on criminality at the 2006 ESSH conference in Amsterdam, a panel organized and discussed by Professor Jean Trépanier. I am grateful for Professor Trépanier’s fruitful comments. This article further owes a great deal to Professor Göran Broström for all his help with statistical methods. I am also grateful for the support from the Swedish Research Council (Vetenskapsrådet), which financed the construction of the matched data set. Finally, I would like to thank my colleagues at the Centre for Population Studies (CPS) and Demographic Data Base (DDB), Umeå University, Sweden, for providing me with constructive critique and for being so helpful in preparing the data files.

Haut de page

Bibliographie

Primary sources

Computerized parish registers, Sundsvall town and the region, 1800-1892, The Demographic Data Base (DDB), Umeå University, Sweden.

Prison registers (fångrullor), 1840-1880, Medelpads kronohäkte, Sundsvalls stadshäkte, Västernorrlands länsfängelse, Västernorrlands läns kronohäkte, Härnösands stadshäkte, The Research Archives, Umeå University, Sweden.

Secondary sources70

Adams, K., Developmental aspects of adult crime, in Thornberry, T. P. (ed.), Developmental theories of crime and delinquency: Advances in criminological theory, Volume 7, New Brunswick, N.J., 1997, pp. 309-342.

Agnew, R., Stability and change in crime over the live course: A strain theory explanation, in Thornberry, T. P. (ed.), Developmental theories of crime and delinquency: Advances in criminological theory, Volume 7, New Brunswick, N.J., 1997, pp. 101-132.

Agnew, R., Strain and subcultural theories of criminality, in Sheley, J. F. (ed.), Criminology: A contemporary handbook, Wadsworth Publishing Company, Belmont, 1995, pp. 305-327.

Allison, P. D., Event history analysis: Regression for longitudinal event data, Beverly Hills, California, Sage Publications, 1984.

Alm Stenflo, G., Demographic description of the Skellefteå and Sundsvall regions during the 19th century, Umeå: The Demographic Data Base, Umeå University, 1994.

Alter, G., Family and the female life course: The women of Verviers, Belgium 1849-1880, Madison, Wisconsin, University of Wisconsin Press, 1988.

Beattie, J. M., The criminality of women in eighteenth-century England, Journal of social history, 1975, 8, pp. 80-116.

Becker, H. S., Outsiders: Studies in the sociology of deviance, London, Macmillan, 1963.

Bengtsson, T., Campbell, C., Lee, J. Z. et al. (Eds), Life under pressure: Mortality and living standards in Europe and Asia, 1700-1900, Cambridge, MIT Press, 2004.

Berlanstein, L., Vagrants, beggars and thieves: Delinquent boys in mid-nineteenth century France, Journal of social history, 1979, 12, pp. 531-552.

Blossfeld, H., Rohwer, G., Techniques of event history modeling: New approaches to causal analysis, Mahwah, N. J., L. Erlbaum, 1995.

Boritch, H., Gender and criminal court outcomes: An historical analysis, Criminology, 1992, 30, pp. 293-325.

Boritch, H., The criminal class revisited: Recidivism and punishment in Ontario, 1871-1920, Social science history, 2005, 29, pp. 137-170.

Bourdieu, P., Cultural reproduction and social reproduction, in Karabel, J., Halsey, A. H. (Eds.), Power and ideology in education, New York, Oxford University Press, 1977, pp. 487-511.

Bras, H., Social change, service and youth in the lives of rural-born Dutch women, Continuity and change, 2004, 19, pp. 241-264.

Breslow, N. E., Covariance adjustment in relative-risk estimates in matched studies, Biometrics, 1982, 38, pp. 661-672.

Breslow, N. E., Day, N. E., Statistical methods in cancer research I: The analysis of case control studies, Lyon, International Agency for Research on Cancer, 1980.

Broström, G., The influence of mother’s death on infant mortality: A case study in matched data survival analysis, Scandinavian journal of statistics, 1987, 14, pp. 113-123.

Broström, G., Event history analysis and historical demography in Brändström, A., Tedebrand, L. (Eds), Health and social change: Disease, health and public care in the Sundsvall district 1750-1950,Umeå, The Demographic Data Base, Umeå University, 1993, pp. 15-22.

Broström, G., Livsförloppsanalys: Kraftfulla analysmetoder vitaliserar demografisk forskning, Tvärsnitt, 2006, 3, pp. 36-39.

Brändström, A., Tedebrand, L. (Eds), Health and social Change: Disease, health and public care in the Sundsvall district 1750-1950, Umeå, The Demographic Data Base, Umeå University, 1993.

Brändström, A., A life after dismissal? Patients’ life histories at a Swedish county hospital, 1845-1890, in Woodward, J., Jütte, R. (Eds), Coping with sickness: Historical aspects of health care in a European perspective, Sheffield, European Association for the History of Medicine and Health Publications, 1995, pp. 93-119.

Brändström, A., Ericsson, T., Social mobility and social networks: The lower middle class in the late nineteenth-century Sundsvall, in Brändström, A., Tedebrand, L. (Eds), Swedish urban demography during industrialization, Umeå, The Demographic Data Base, Umeå University, 1995, pp. 251-283.

Cavender, G., Alternative approaches: Labeling and critical perspectives, in Sheley, J. F. (ed.), Criminology: A contemporary handbook, Wadsworth Publishing Company, Belmont, 1995, pp. 349-367.

Chesney-Lind, M., Women and crime: The female offender, Signs, 1986, 12, pp. 78-96.

Cohen, A. K., Deviance and control, New Jersey, Engelwood Cliffs, 1966.

Cox, D. R., Regression model and life tables (with discussion), Journal of the royal statistical society, 1972, 34, pp. 187-220.

Cox, D. R., Oakes, D., Analysis of survival data, London, MacGibbon & Kee, 1984.

Drugge, U., Lindgren, S., Med dödlig utgång: Om grova våldsbrott och sociala former i 1800-talets Sverige, Umeå, Sociologiska institutionen, Umeå Universitet, 2001.

Edvinsson, S., Den osunda staden: Sociala skillnader i dödlighet i 1800-talets Sundsvall, Umeå, Demografiska databasen, Umeå Universitet, 1992.

Edvinsson, S., Adult mortality and childhood conditions: Long-term effects of urban life in nineteenth-century Sweden, in Sköld, P., Tedebrand, L. (Eds), Nordic demography in history and present-day society:Scandinavian population studies, Volume 12, Umeå, Umeå University, 2001, pp. 247-268.

Elder, G. H., Jr., Perspectives on the life course, in Elder, G. H., Jr. (ed.), Life course dynamics, Ithaka, Cornell University Press, 1985, pp. 23-49.

Ellis, H., The criminal,London, Walter Scott, 1890.

Emsley, C., Crime and society in England, 1750-1900, London, Longman, 1987.

Erikson, R., Goldthorpe, J. H., The constant flux: A study of class mobility in industrial societies, Oxford, Clarendon, 1992.

Eriksson, I., Rogers, J., Rural labor and population change: Social and demographic developments in east-central Sweden during the nineteenth century, Uppsala, Uppsala University, Almqvist & Wiksell International, 1978.

Farrington, D. P., The effects of public labelling, British journal of criminology, 1977, 18, pp. 277-284.

Farrington, D. P. (Ed.), Integrated developmental and life-course theories of offending: Advances in criminological theory, Volume 14, New Brunswick, N. J., 2005a.

Farrington, D. P., Introduction to Integrated Developmental and Life-Course Theories of Offending, in Farrington, D. P. (ed.), Integrated developmental and life-course theories of offending: Advances in criminological theory, Volume 14, New Brunswick, N. J., 2005b, pp. 1-14.

Farrington, D. P., The integrated cognitive antisocial potential (ICAP) theory, in Farrington, D. P. (ed.), Integrated developmental and life-course theories of offending, Advances in criminological theory, Volume 14, New Brunswick, N. J., 2005c, pp. 73-92.

Foucault, M., Övervakning och straff: Fängelsets födelse, Lund, Arkiv, 1975.

Frank, S. P., Narrative within numbers: Women, crime and judicial statistics in imperial Russia, 1864-1913, The Russian review, 1996, 55, pp. 541-565.

Fredriksson, G., Social rörlighet eller historisk förändring: En studie av svensk socialhistoria, Häften för kritiska studier, 1980, 15, pp. 15-49.

Fridlizius, G., Sex-differential mortality and socio-economic change: Sweden 1750-1910, in Brändström, A., Tedebrand, L. (Eds), Society, health and population during the demographic transition, Umeå, The Demographic Data Base, Umeå University, 1988, pp. 237-272.

Fuchs, R. G., Juvenile delinquency in nineteenth-century France, in Hess, A. G., Clement, P. F. (Eds), History of juvenile delinquents: A collection of essays on crime committed by young offenders, in history and in selected countries, Volume 1, Aalen, Scientia, 1990, pp. 265-287.

Galley, C., Urban graveyards and English population history, in Brändström, A., Tedebrand, L. (Eds), Swedish urban demography during industrialization, Umeå, The Demographic Data Base, Umeå University, 1995, pp. 141-173.

Giele, J. Z., Elder, G. H., Jr. (Eds), Methods of life course research: Qualitative and quantitative approaches, London, Sage, 1998a.

Giele, J. Z., Elder, G. H., Jr., Life course research: Development of a field, in Giele, J. Z., Elder, G. H., Jr. (Eds), Methods of life course research: Qualitative and quantitative approaches, London, Sage, 1998b, pp. 5-27.

Glueck, S., Glueck, E., Unraveling juvenile delinquency, New York, Commonwealth Fund, 1950.

Goffman, E., Stigma: Notes on the management of spoiled identity, Englewood Cliffs, N. J., Prentice-Hall, 1963.

Graff, H. J., Crime and punishment in the nineteenth century: A new look at the criminal, Journal of interdisciplinary history, 1977, 7, pp. 477-491.

Greenberg, D. F., Age, crime, and social explanation, American journal of sociology, 1985, 91, pp. 1-21.

Hajnal, J., European marriage patterns in perspective in Glass, D. V., Eversley, D. E. C. (Eds), Population and history: Essays in historical demography, London, Edward Arnold, 1965, pp. 101-143.

Hareven, T., The history of the family and the complexity of social change, The American historical review, 1991, 96, pp. 95-124.

Hirshi, T., Causes of delinquency, Berkeley, 1969.

Hirshi, T., Gottfredson, M. R., Age and the explanation of crime, American journal of sociology, 1983, 89, pp. 552-584.

Inger, G., Svensk rättshistoria, Stockholm, LiberFörlag, 1983.

Jacobsson, M., ‘Att blifva sin egen’: Ungdomars väg in i vuxenlivet i 1700- och 1800-talets övre Norrland, Umeå, Umeå Universitet, 2000.

Janssens, A., Family time and social change: The household as a process in an industrializing society, Cambridge, Cambridge University Press, 1993.

Jarrick, A., Söderberg, J., Odygd och vanära: Folk och brott i gamla Stockholm, Stockholm, Rabén Prisma, 1998.

Jonsson, F., Ohlsson, M., Överlevnadsanalys på matchat data: En studie av arresteringens stigmatiserande effekter, Unpublished paper presented at the Department of Statistics, Umeå University, Spring 2007.

Kaelble, H., Eras of social mobility in 19th and 20th century Europe, Journal of social history, 1984, 17, pp. 489-504.

Kalbfleich, J. D., Prentice, R.L., The statistical analysis of failure time data, New York, Wiley, 1980.

Kearns, G., The urban penalty and the population history of England, in Brändström, A., Tedebrand, L. (Eds), Society, health and population during the demographic transition. Umeå, The Demographic Data Base, Umeå University, 1988, pp. 213-236.

King, P., The rise of juvenile delinquency in England 1780-1840: Changing patterns of perception and prosecution, Past and present, 1998, 160, pp. 116-166.

Krohn, M., Control and deterrence theories of criminality, in Sheley, J. F. (ed.), Criminology: A contemporary handbook, Wadsworth Publishing Company, Belmont, 1995, pp. 328-347.

Laslett, P., Characteristics of the Western family considered over time, Journal of family history, 1977, 2, pp. 89-116.

Laub, J. H., Sampson, R. J., Integrating quantitative and qualitative data, in Elder, G. H., Giele, J. Z. (Eds), Methods of life course research: Qualitative and quantitative approaches, London, Sage, 1998, pp. 213-230.

Le Blanc, M., A Generic control theory of the criminal phenomenon: The structural and dynamic statements of an integrative multilayered control theory, in Thornberry, T. P. (ed.), Developmental theories of crime and delinquency: Advances in criminological theory, Volume 7, New Brunswick, N. J., 1997, pp. 215-285.

Leeuwen, M. H. D. van, Maas, I., Industrialization and intergenerational mobility in Sweden, Acta sociologica, 2002, 45, pp. 179-194.

Lemert, E. M., Social pathology: A systematic approach to the theory of sociopathic behavior, New York, McGraw-Hill, 1951.

Lext, G., Studier i svensk kyrkobokföring 1600-1946, Göteborg, Landsarkivet, 1984.

Lindgren, S., Om brott och straff: Från sociologins klassiker till modern kriminologi, Lund, Studentlitteratur, 1998.

Loeber, R., Le Blanc, M., Toward a developmental criminology, in Tonry, M., Morris, N. (Eds), Crime and justice: A review of research, Volume 12, Chicago, University of Chicago Press, 1990, pp. 375-437.

Lundh, C., The world of Hajnal revisited: Marriage patterns in Sweden 1650-1990, Lund papers in economic history, 1997, 60, pp. 1-28.

Lynch, K. A., Old and new research in historical patterns of social mobility, Historical methods, 1998, 31, pp. 93-98.

Lynn, H. S., McCulloch, C. E., When does it pay to break the matches for analysis of a matched-pairs design?, Biometrics, 1992, 48, pp. 397-409.

Matsueda, R. L., Heimer, K., A symbolic interactionist theory of role-transitions, role-commitments, and delinquency, in Thornberry, T. P. (Ed.), Developmental theories of crime and delinquency: Advances in criminological theory, Volume 7, New Brunswick, N. J., 1997, pp. 163-213.

Mayer, K. U., Tuma, N. B., Life course research and event history analysis: An overview, in Mayer, K. U., Tuma, N. B. (Eds), Event history in life course research, Madison, Wisconsin, University of Wisconsin Press, 1990, pp. 5-20.

McDonald, L., Theory and evidence of rising crime in the nineteenth century, TheBritish journal of sociology, 1982, 3, pp. 404-420.

Merton, R. K., Nisbet, R. A., Contemporary social problem: An introduction to the sociology of deviant behavior and social disorganization, London, Hart-Davis, 1963.

Moffitt, T. E., Adolescent-limited and life-course-persistent offending: A complementary pair of developmental theories, in Thornberry, T. P. (ed.), Developmental theories of crime and delinquency: Advances in criminological theory, Volume 7, New Brunswick, N. J., 1997, pp. 11-54.

Monkkonen, E. H., A disorderly people? Urban order in the nineteenth and twentieth centuries, The journal of American history, 1981, 68, pp. 539-559.

Muncie, J., Youth and crime: A critical introduction, London, Sage, 1999.

Nilsson, H., Tedebrand, L, Familjer i växande städer. Strukturer och strategier vid familjebildning i Sverige 1840-1940, Umeå, Demografiska Databasen, Umeå Universitet, 2005.

Nilsson, R., ‘Med hvarje år öka ju antalet af dessa unga brottslingar…’, in Andersson, I., Johansson, K., Lindstedt Cronberg, M. (Eds), Tid och tillit: En vänbok till Eva Österberg. Stockholm, Atlantis, 2002, pp. 371-388.

Norberg, A., Med betyget på fickan: Den svenska folkbokföringen under industrialiseringsskedet, Kyrkohistorisk årsskrift, Uppsala, Svenska kyrkohistoriska föreningen, 1979, pp. 151-178.

O’Brien, P., The promise of punishment: Prisons in nineteenth-century France, Princeton, Princeton University Press, 1982.

Paternoster, R., Iovanni, L., The labeling perspective and delinquency: An elaboration of the theory and an assessment of the evidence, Justice quarterly, 1989, 6, pp. 59-394.

Petersson, B., Den farliga underklassen: Studier i fattigdom och brottslighet i 1800-talets Sverige, Umeå, Umeå Universitet, Almqvist & Wiksell International, 1983.

Råberg, J., Sandström, D., Konsekvenser av ett kriminellt förflutet: En överlevnadsstudie med matchade data, Unpublished paper presented at the Department of Statistics, Umeå University, Spring 2006.

Sampson, R. J., Laub, J. H., Crime in the making: Pathways and turning points through life, Cambridge, Massachusetts, Harvard University Press, 1997a (1993).

Sampson, R. J., Laub, J. H., A life-course theory of cumulative disadvantage and the stability of delinquency, in Thornberry, T. P. (ed.), Developmental theories of crime and delinquency: Advances in criminological theory, Volume 7, New Brunswick, N. J., 1997b, pp. 133-161.

Schultes, R., The village in court: Arson, infanticide, and poaching in the court records of upper Bavaria, 1848-1910, Cambridge, Cambridge University Press, 1994.

Shearer, D. L., Mulvihill, B. A., Klerman, L. V., Wallander, J. et al., Association of early childbearing and low cognitive ability, Perspectives on sexual and reproductive health, 2002, 34, pp. 236-243.

Smart, C., Criminological theory: Its ideology and implications concerning women, British journal of sociology, 1977, 28, pp. 89-100.

Steffensmeier, D., Allen, E., Criminal behavior: Gender and age, in Sheley, J. F. (ed.), Criminology: A contemporary handbook, Wadsworth Publishing Company, Belmont, 1995, pp. 83-113.

Taussi Sjöberg, M., Brott och straff i Västernorrland 1861-1890, Umeå, Umeå Universitet, 1981.

Taussi Sjöberg, M., Dufvans fångar: Brottet, straffet och människan i 1800-talets Sverige, Stockholm, Författarförlaget, 1986.

Tedebrand, L., Människor och strukturer intill 1860, in Tedebrand, L. (ed.), Sundsvalls historia: Del 1, Sundsvall, Stadshistoriska kommittén, 1996, pp. 192-210.

Tedebrand, L., Gamla och nya Sundsvallsbor efter 1860, in Tedebrand, L. (ed.), Sundsvalls historia: Del 2, Sundsvall, Stadshistoriska kommittén, 1997, pp. 101-136.

Thornberry, T. P. (ed.), Developmental theories of crime and delinquency: Advances in criminological theory, Volume 7, New Brunswick, N. J., 1997a.

Thornberry, T. P., Introduction: Some advantages of developmental and life-course perspectives for the study of crime and delinquency, in Thornberry, T. P. (ed.), Developmental theories of Crime and delinquency: Advances in criminological theory, Volume 7, New Brunswick, N. J., 1997b, pp. 1-10.

Tobias, J. J., Crime and industrial society in the nineteenth century, London, Batsford, 1967.

Vejbrink, M., En kontrollerad brottsling som historiskt fenomen, in Sundin, J. (ed.), Kontroll och kontrollerade: Formell och informell kontroll i ett historiskt perspektiv,Umeå, Umeå Universitet, 1982, pp. 159-195.

Vikström, L., Gendered routes and courses: The socio-spatial mobility of migrants in nineteenth-century Sundsvall, Sweden, Umeå, The Demographic Data Base, Umeå University, 2003.

Vikström, L., Mycket väsen för ringa ting! Ungdomsbrottens utveckling i Sundsvallsregionen 1840-1880, Oknytt, 2006a, 1-2, pp. 41-68.

Vikström, L., Vulnerablity among paupers: Determinants of individuals receiving poor relief in nineteenth-century northern Sweden, The history of the family, 2006b, 11, pp. 223-239.

Vikström, L., Societal change and individual past in connection with crime: Demographic perspectives of young offenders in nineteenth-century northern Sweden, Continuity and change, in press [2008a].

Vikström, L., Life prior to crime and after: Developmental perspectives of young offenders in nineteenth-century northern Sweden, Journal ofSocial History, in press [2008b].

Wallén, P., Svensk straffrättshistoria: Några huvudlinjer, Del 2, Stockholm, Geber, 1973.

Wellman, B., Wetherell, C., Social networks analysis of historical communities: Some questions from the present to the past, The history of the family, 1996, 1, pp. 97-121.

Williams, K. S., Labelling, phenomenology and ethnomethodology, in Williams, K. S. (ed.), Textbook on criminology:Fifth edition, Oxford, Oxford University Press, 2004a, pp. 369-384.

Williams, K. S., Positivist explanations of female criminality, in Williams, K. S. (ed.), Textbook on criminology:Fifth edition, Oxford, Oxford University Press, 2004b, pp. 447-468.

Williams, K. S., Feminist theories, in Williams, K. S. (ed.), Textbook on criminology:Fifth edition, Oxford, Oxford University Press, 2004c, pp. 469-492.

Willner, S., Det svaga könet? Kön och vuxendödlighet i 1800-talets Sverige, Linköping, Linköpings Universitet, 1999.

Zedner, L., Women, crime and custody in Victorian England, Oxford, Clarendon, 1991.

Zehr, H., Crime and the development of modern society, London, Croom Helm, 1976.

Österberg, E., Kontroll och kriminalitet i Sverige från medeltid till nutid, Scandia, 1991, 57, pp. 66-87.

Haut de page

Notes

2 Like scholars focusing on today’s criminology, those studying past times discover unfortunate factors that connect individual crime with bad experiences of childhood, education, unstable work and family. Cf. Beattie (1975); Berlanstein (1979); Fuchs (1990, pp. 278ff); Greenberg (1985): Jarrick, Söderberg (1998); Petersson (1983, pp. 123-181); Sampson, Laub (1997a, pp. 64-136); Schultes (1994); Taussi Sjöberg (1981); Taussi Sjöberg (1986); Vejbrink (1982).

3 Cf. O’Brien (1982, pp. 226-257); Taussi Sjöberg (1986, pp. 145-160); Vejbrink (1982), all of whom suggest that former prisoners faced limited choices in their readjustment to society.

4 The study concerns the town of Sundsvall and five neighboring parishes within the region that witnessed rapid industrialization, i.e. Alnö, Njurunda, Skön, Timrå and Tuna. Studies show that the socio-demographic components so typical of urban-industrial times, such as rising mortality and illegitimacy rates, and rapid population growth caused by increasing in-migration were also seen in this tiny area. Alm Stenflo (1994); Edvinsson (1992); Tedebrand (1996; 1997); Vikström (2003).

5 In the digitalized parish registers of the Demographic Data Base (DDB), every individual is given a unique identity number: Jonas Petter Wedin (DDB-ID: 857001495). Concerning Wedin’s misconduct, see the prison registers of Västernorrlands läns kronohäkte 1872-09-05, 1873-03-01, 1873-04-29. Cf. the heading of ‘A matched data set, definitions and methodology’. Information on poor relief has been generated from parish meeting records, as these were investigated for the purpose of analyzing paupers who were relieved, see Vikström (2006b).

6 Brita Cajsa Wedin (DDB-ID: 848001367). Prison registers of Västernorrlands läns kronohäkte 1870-11-19.

7 Frans Walfrid Wedin (DDB-ID: 861001991). Prison registers of Västernorrlands läns kronohäkte 1877-09-22, 1877-10-22.

8 Carl Jacob Andersson (DDB-ID: 852001372). Prison registers of Västernorrlands läns kronohäkte 1868-03-04, 1868-09-03.

9 For an overview and some critique of both traditional concepts and the application of those associated with labeling theories, or interactionism, see Cavender (1995, pp. 351-358); Williams (2004a, pp. 369-384).

10 See the introductions of Farrington (2005b), Thornberry (1997b), and the contributions in the two books on this topic that they have edited, Farrington (2005a); Thornberry (1997a).

11 Since the nineteenth century a comprehensive body of literature on criminal behavior and its reasons has been published, cf. Cohen (1966); Ellis (1890); Hirshi (1969); Merton, Nisbet (1963). For some critique of approaches that link crime to modernization, see Agnew (1995); Emsley (1987, pp. 18-47); Krohn (1995); Lindgren (1998); McDonald (1982); Monkkonen (1981, pp. 551ff); Vikström (2008a).

12 Howard S. Becker (1963), Erving Goffman (1963) and Edwin M. Lemert (1951) were among the first to discuss labeling perspectives, although their origin goes further back in time. Cf. Cavender (1995, pp. 351-357); Lindgren (1998, pp. 73-85); Williams (2004a, pp. 369-384).

13 Such labeling theories are discussed in Le Blanc (1997, pp. 263-278); Farrington (1977); Paternoster, Iovanni (1989); Sampson, Laub (1997b, pp. 133-141). The latter also employ the more classical concepts of social control.

14 Muncie (1999, p. 310).

15 Farrington (2005b) and Thornberry (1997b). See also the contributions in the two books on this topic that they have edited, Farrington (2005a); Thornberry (1997a).

16 Elder (1985, quote p. 17). For an overview of the life-course concept, see Elder (1985); Giele, Elder (1998a), and the contributions in the life-course book that they have edited, Giele, Elder (1998b). See also footnotes 32 and 33.

17 Graff (1977, pp. 477ff). Other scholars also emphasize the advantages of focusing on the characteristics of criminals and applying longitudinal analyses of individuals instead of examining huge collections of statistical data. Cf. Boritch (2005); Laub, Sampson (1998).

18 Loeber, Le Blanc (1990, quote p. 421).

19 The court books of the area under study were destroyed by the fire that hit the town of Sundsvall in 1888, see Taussi Sjöberg (1986, p. 163). Although court books are more informative in documenting the criminals recorded, prison registers comprise all individuals who were arrested and imprisoned for some time, hence not only those who were sent to court. The validity of prison registers is tested and verified, see Taussi Sjöberg (1981, pp. 5-9).

20 Concerning the gap between actual criminal behavior and registered criminality that is particularly difficult to detect in historical data, see Nilsson (2002, p. 381); Petersson (1983, pp. 89ff); Österberg (1991, pp. 69ff). Using prison registers instead of court books helps to overcome some of this gap, cf. above footnote.

21 Although Swedish parish registers are unique in delivering continuous data, gaps exist because they were collected for bureaucratic, political and religious reasons. Women’s occupations, for instance, are insufficiently documented. These registers must nevertheless be regarded as trustworthy. One clear advantage is that they allow longitudinal analyses of almost every member of society. The validity of Swedish parish registers in general, and of those for Sundsvall and its region in particular, has been tested and verified. Cf. Norberg (1979, pp. 151-178); Alm Stenflo (1994, pp. 17-34); Edvinsson (1992, pp. 27ff). For a presentation of the sources stored at the Demographic Data Base (DDB), Umeå University, Sweden, see [www.ddb.umu.se/index_eng.html].

22 Concerning the identification process of the offenders in the two sources, see footnote 39.

23 Cf. footnote 41. The following circumstances have helped to delineate the age interval of this study. Young people are particularly frequent among criminals, cf. Greenberg (1985); Hirshi, Gottfredson (1983); King (1998, pp. 120ff); Steffensmeier, Allan (1995, pp. 97-113). In the nineteenth century the average Swede married at the age of 27-28 and young people put childhood behind them when they applied for confirmation at the age of 14-15, Jacobsson (2000, pp. 111ff); Lundh (1997). Only data on offenders who were unmarried at the time of arrest were used, so that the event of marriage could be analyzed.

24 The extra-informative church examination lists have been utilized to exclude criminals among the controls. The minister used these lists when he checked his parishioners’ biblical knowledge each year, and he sometimes commented upon their moral or criminal status. Individuals with a criminal history were identified and excluded from the control cohort. There is nevertheless a risk that some offenders appear in the control cohort, as the ministers reported inconsistently about the parishioners’ criminal characteristics and because the computerized data only include the Sundsvall region for as long as the individuals are present there. However, the control cohort was constructed to show the average pathway of young people, not to pattern that of the most ‘decent’ ones. Additionally, people committed crimes that were never acknowledged, which is still the case today.

25 In principle this is so, but as the offenders must also be identified in the parish registers to be included in the longitudinal analysis, every person arrested is not accounted for. Cf. footnote 39.

26 Cf. Cavender (1995, pp. 352ff); Williams (2004a, pp. 374-379); Lindgren (1998, pp. 73-76).

27 Breslow (1992); Breslow, Day (1980); Lynn, McCulloch (1992). The construction of this control cohort has been governed by a statistician, Professor Göran Broström, Dept. of Statistics, Umeå University, and I am most grateful for his support.

28 The types of individual characteristics most suitable for constructing a matched control cohort and defining the number of controls necessary to get the best comparable data set is discussed among statisticians but seldom with regard to the limitations of historical data, cf. Broström (1987). Statistical knowledge recommends access to a large matched data set including more than only one paired control, see Breslow (1982); Lynn, McCulloch (1992). The validity of the results presented below has been tested by using more advanced statistical analyses, such as stratified Cox regression models, cf. Råberg, Sandström (2006). These tests and models support the findings generated here and suggest that the number of controls per individual influences the reliability of the findings. The most significant results appear when ten or more controls are included in the models. All individuals in the control cohort have therefore been examined. Statistical analyses of the individuals also matched with regard to their socio-economic status did not change the differences and similarities between the pathways of the two cohorts, cf. Jonsson, Ohlsson (2007).

29 For studies using matched historical and modern demographic data, see Broström (1987); Brändström (1995); Shearer et al. (2002).

30 Glueck, Glueck (1950); Laub, Sampson 1998; Sampson, Laub (1997a; 1997b).

31 Agnew (1997, pp. 101-103); Farrington (2005c, pp. 74ff); Sampson, Laub (1997b, pp. 133ff).

32 For a presentation of advanced statistical survival analyses, see Cox (1972); Cox, Oakes (1984); Kalbfleich, Prentice (1980), and of event-history analyses, see Allison (1984); Blossfeld, Rohwer (1995); Broström (1993, 2006).

33 Thanks to the growing access to databases that store past quantitative material, historical demographers and sociologists are increasingly applying life-course perspectives and event-history analyses. Cf. Alter (1988); Bengtsson, Campbell, Lee et al. (2004); Bras (2004); Broström 2006; Brändström 1995; Mayer, Tuma (1990); Vikström (2003). Overviews and examples of life-course approaches are found in Giele, Elder (1998a), and with regard to a criminal context in Sampson, Laub (1997a, b).

34 The results presented in section 2 are more thoroughly discussed in Vikström (2006; 2008a).

35 Alm Stenflo (1994, pp. 37-47; Tedebrand (1996; 1997); Vikström (2003, pp. 69-79).

36 The legislation is discussed in Inger (1982, pp. 234-48); Wallén (1973, pp. 5-16). Concerning the ‘social issue’ and fear of revolution in nineteenth-century Sweden, see Drugge and Lindgren (2001, pp. 271-277); Petersson (1983, pp. 16-68), and abroad Boritch (2005, pp. 139-144); Monkkonen (1981); Muncie (1999, pp. 47-77).

37 Emsley (1987, pp. 18-47); King (1998, pp. 133-137); McDonald (1982); Taussi Sjöberg (1981, pp. 19ff); Österberg (1991, p. 81); Governmental interference with individuals’ lives in terms of ­controlling and measuring their behavior has been investigated by not least Michel Foucault (1975).

38 Young individuals were largely involved in the general rise of crimes during the second part of nineteenth-century Sweden, but viewed in relative terms their proportion remained fairly stable. Cf. Nilsson (2002, pp. 375ff).

39 There are several reasons why not all young offenders are described in detail. First, some did not report their residence to the ministers in the Sundsvall region, probably because they were settling there on a temporary basis, as seasonal laborers did. This under-registration reduces the number of offenders possible to analyze in the parish registers. Second, the personal data given in the prison registers do not always help to identify the individual in the parish registers. Common names and incomplete dates of birth jeopardized the identification process. Moreover, arrested individuals sometimes gave false personal data. Hence, young people with a relatively solid relation to the Sundsvall region are over-represented in the criminal cohort.

40 In general, scholars separate violent crimes from those related to property or disorder, cf. Taussi Sjöberg (1981, pp. 56-84); Petersson (1983, pp. 89-122). With regard to the limited number of offenders in this study, their crimes are not differentiated further than this. Like occupations and social classifications, the categorization of crimes is a difficult task, especially as one individual might be arrested for two or even three types of crime. The three crime categories used here were guided by the following order: When charged with violence and disorder, violence is given priority, and when disorderly behavior and theft are combined, the latter is given priority. Advanced economic crimes were rare and are included in the category of theft. Cf. Vikström (2008a).

41 According to Swedish statistics in 1857, only one delinquent in ten among those charged and convicted was female. This ratio declined to 6% at the end of the century. About 25% of those arrested were aged 25 or below. Like the average criminal both in Sweden and abroad, young people usually committed their crimes in urban areas and about nine in ten were men. Cf. Taussi Sjöberg (1981, pp. 96-110); Nilsson (2002, pp. 372ff). In nineteenth-century Britain, about 75% of the criminals were men. Most were in their early twenties. Cf. Emsley (1994, p. 27); Fuchs (1990, pp. 281ff).

42 Gender and criminality have aroused increasing interest among scholars in the past few decades. For an overview, see Chesney-Lind (1986); Lindgren (1998, pp. 86-92); Smart (1977); Williams (2004b, c), and for some historical studies, see Beattie (1975); Boritch (1992); Frank (1996); Fuchs (1990); O’Brien (1982); Taussi Sjöberg (1981, pp. 96-105); Zedner (1991).

43 Gendered reasons for which men and women were arrested and with regard to their demographic past are discussed in Vikström (2008a).

44 Ellis (1890); Emsley (1987, pp. 78-102); Monkkonen (1981); Tobias (1967); Zehr (1976, pp. 20-29).

45 At the turn of the nineteenth century, native-born Canadians dominated among recidivists in Ontario, see Boritch (2005, p. 155). Neither migrants nor their children dominated among delinquent boys in nineteenth-century France, see Berlanstein (1979, p. 537).

46 In the Sundsvall region, about 40 kilometers separated the most remote parishes from one another. If every offender involved in the 714 criminal cases reported in the prisoner registers in 1840-1880 had been possible to identify in the parish registers, more migrants would have been found. The regional background of the offenders would nevertheless be evident. This is explained by their low age (15-25 years old). Some delinquents were too young to have relocated.

47 Today most scholars emphasize such ties, especially in times of urban-industrial change, and these are usually perceived as being supportive for people who are part of them. Cf. Hareven (1991); Janssens (1993, pp. 1-30); Wellman, Wetherell (1996).

48 Spouses, parents and children are the only type of network the parish registers acknowledge, except for some siblings, but because they often relocated they are difficult to control over time. In this study, step-parents are also included.

49 Drugge, Lindgren (2001, pp. 135ff); Muncie (1999, pp. 207-219); Sampson, Laub (1997a, pp. 18, 64-98); Vejbrink (1982, pp. 185-193); Österberg (1991, pp. 82-84).

50 Cf. Boritch (2005); Emsley, (1987, pp. 48-77); Lindgren (1998, pp. 48-54).

51 Most scholars agree that occupation can indicate a person’s status and they have also developed different systems of social classification. The one used in this study employs that developed by researchers in Sweden. Partly because of the scarcity of cases linked to each categorization, the social stratification is slightly modified. For a discussion concerning this issue, see Edvinsson (1992, pp. 39-43); Fredriksson (1980); Vikström (2003, pp. 49ff). Cf. footnote 68.

52 Similar results appear among delinquent boys in nineteenth-century France, see Berlanstein (1979, p. 537).

53 This happened to almost 75% of them. Only 6% were sentenced to spending more than one year in penal servitude or in prison. To reduce their penalties, those prosecuted for serious matters frequently appealed the decision of local courts by consulting higher instances, often with success. This, and the fact that the court books were destroyed by the fire that hit the town of Sundsvall in 1888, make it difficult to consider their punishment in more detail.

54 That the socio-economic and cultural backgrounds of individuals affect their pathways in terms of resources that can be translated into failure or success is primarily discussed by Bourdieu (1977). His behavioral theories have influenced much research including the life-course perspectives and DLC concepts in criminology, cf. Matsueda, Heimer (1997, pp. 167ff).

55 The number and type of independent variables (covariates) are restricted to the quality of the data. Those selected are employed because in historical and demographic research these variables often prove to influence individuals’ life courses. To judge the impact of crime or any other independent variable, they must be incorporated into the model simultaneously. Cf. Sampson, Laub (1997a, pp. 168-178); Laub, Sampson (1998, pp. 220ff).

56 To specifically analyze the pathways of those arrested twice or more, there is a need to have access to the dates of every incarceration these recidivists experienced over their lifetime and let the longitudinal observation of 15 years start at these dates. This comprehensive approach goes beyond the scope of this survey. The recidivists are nevertheless included in the findings presented.

57 The time interval of fifteen years was constructed for comparable reasons, as the parish registers end in the mid-1890s and those arrested in the late 1870s and 1880 cannot be studied beyond the end of registration. The longitudinal study is also limited in time to judge evident impacts of incarceration on their life courses.

58 Cf. Glueck, Glueck (1950); Sampson, Laub (1997a).

59 Separating the criminal cohort into three crime categories of violence, theft and disorder jeopardizes the comparison with the large control cohort as these categories are far too small in number. As regards differences within the criminal cohort, those charged with theft tend to have suffered the most from their misconduct. Vikström (2008b).

60 Hajnal (1965). Cf. Nilsson, Tedebrand (2005, pp. 15-44); Vikström (2003, pp. 185ff); Laslett (1977); Lundh (1997).

61 Sampson, Laub (1997a, pp. 178-245; 1997b, pp. 145-152). Adams (1997, pp. 328-332) and Matsueda, Heimer (1997, pp. 196-200) also discuss the socializing effects of marrying and starting a family life that seem to favor desistance from crime among offenders as they grow older.

62 Alm Stenflo (1994, pp. 64-68); Lundh (1997, p. 10); Vikström (2003, pp. 188-191).

63 Of course, a more thorough study of the choice of partner with regard to age and socio-economic status would shed further light onto whether the offenders were ‘losers’ on the marriage market, but this is not considered in this article.

64 Until the poor law of 1847, the parish boards had to approve migrants, and some were not welcome at their new destinations. Those denied were usually old, deprived, disabled or in trouble with the law. Judicial obstacles to migration were limited after that, although the practices of these recommendations could vary among regions and parishes. Cf. Eriksson, Rogers (1978, pp. 180ff); Lext (1984, pp. 242-282); Norberg (1979).

65 Moffitt (1997, pp. 21-23); Sampson, Laub (1997a, pp. 123-138; 1997b, pp. 143ff).

66 Brändström, Tedebrand (1993); Edvinsson (1992); Fridlizius (1988); Willner (1999).

67 Edvinsson (2001); Galley (1995); Kearns (1988).

68 Cf. Brändström, Ericsson (1995); Erikson, Goldthorpe (1992); Kaelble (1984); Leeuwen, Maas (2002); Lynch (1998). For an overview, see Vikström (2003, pp. 25ff).

69 Among the Gluecks’ male delinquents, Sampson, Laub (1997a, pp. 178-242; 1997b, pp. 145-152) find that stable employment favored desistance from crime. Adams (1997, pp. 328-335) also discusses how growing age and commitment to work and family interfere with criminal behavior.

70 Unless something else is mentioned, such as ‘No.’, the volume is used when referring to journals.

Haut de page

Table des illustrations

Légende Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.
URL http://chs.revues.org/docannexe/image/361/img-1.png
Fichier image/png, 17k
Titre Figure 2. Type of misconduct among young offenders arrested during the period 1840-1880, their geographical background and parental access. Percentages within each category (N = 320).
Légende Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Database, Umeå University.Explanations: Concerning the crime categorization, see footnote 40.1. Violence – primarily physical assault and breach of peace, including a few homicides2. Theft – primarily petty larceny and property offences, including a few advanced economic crimes (forgery, fraud, swindle)3. Disorder – disturbing behavior such as drunkenness, vagrancy, street disorder, verbal assaultComments: Native offenders (2a) had never registered their residence elsewhere prior to their incarceration. The category of intra-regional migrants (2b) includes individuals who had left one regional parish for another, including the town of Sundsvall. Immigrants and those who departed from places outside the Sundsvall region are gathered in the category of inter-regional migrants (2c). The parents of offenders (categories 3a-b) usually lived in the same parish (about 90%).
URL http://chs.revues.org/docannexe/image/361/img-2.png
Fichier image/png, 12k
Titre Figure 3. Social status of individuals and their fathers when the young offenders were arrested during the period 1840-1880: a comparison between the criminal and control cohorts (percentages).
Légende Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.Explanations: Upper social strata1. Large-scale business entrepreneurs2. Higher civil officialsMiddle social strata3a. Small-scale entrepreneurs in trade and industry, master artisans3b. Farmers, tenant farmers4. Lower civil officialsLower social strata5. Skilled laborers, craftsmen and artisans below the rank of master6a. Unskilled laborers in industry and urban commerce6b. Agricultural workers, farmhands and maidservants
URL http://chs.revues.org/docannexe/image/361/img-3.png
Fichier image/png, 5,9k
Titre Figure 4. Life table of the time between incarceration and marriage within the interval ­covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.
Légende Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.Comments and explanations: Life tables illuminate the survival chances by considering the event of death with regard to the period under study. In this case, marriage replaces the event of death, whereas remaining unmarried is equal to survival. The larger and more rapid the decline of the curve, the more marriages are found. Within ten years of the time span under study about 55% of the individuals in both cohorts had experienced this transition. The life table shows that a difference between the criminal and control cohort appears, as the young individuals had reached the phase in life when most marriages occur. The Cox regression model of Table 3 indicates that this difference is significant at the ten-percent level.
URL http://chs.revues.org/docannexe/image/361/img-4.png
Fichier image/png, 6,1k
Titre Figure 5. Life table of the time between incarceration and departure within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.
Légende Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.Comments and explanations: Cf. Figure 4
URL http://chs.revues.org/docannexe/image/361/img-5.png
Fichier image/png, 5,8k
Titre Figure 6. Life table of the time between incarceration and death within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.
Légende Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.Comments and explanations: Cf. Figure 4.
URL http://chs.revues.org/docannexe/image/361/img-6.png
Fichier image/png, 5,4k
Titre Figure 7. Life table of the time between incarceration and upward mobility within the interval covering a maximum of 15 years after offenders’ incarceration during the period 1840-1880: a comparison between the criminal and control cohorts.
Légende Source: Prison registers, The Research Archives, Umeå University; Parish registers, The Demographic Data Base, Umeå University.Comments and explanations: Cf. Figure 4.
URL http://chs.revues.org/docannexe/image/361/img-7.png
Fichier image/png, 5,7k
Haut de page

Pour citer cet article

Référence papier

Lotta Vikström, « Illuminating the Labeling Impact of Incarceration: Life-Course Perspectives of Young Offenders’ Pathways in Comparison to Non-Offenders in Nineteenth-Century Northern Sweden », Crime, Histoire & Sociétés / Crime, History & Societies, Vol. 12, n°2 | 2008, 81-117.

Référence électronique

Lotta Vikström, « Illuminating the Labeling Impact of Incarceration: Life-Course Perspectives of Young Offenders’ Pathways in Comparison to Non-Offenders in Nineteenth-Century Northern Sweden », Crime, Histoire & Sociétés / Crime, History & Societies [En ligne], Vol. 12, n°2 | 2008, mis en ligne le 01 octobre 2011, consulté le 23 novembre 2017. URL : http://chs.revues.org/361 ; DOI : 10.4000/chs.361

Haut de page

Auteur

Lotta Vikström

CPS – Centre for Population Studies, Umeå University, SE-901 87 UMEÅ, SWEDEN, E-mail: lotta.vikstrom@ddb.umu.se
Marie-Christine (Lotta) Vikström is a Doctor of Philosophy in history and researcher associated with the Centre for Population Studies and the Demographic Data Base, Umeå University, Sweden. She has published Gendered routes and courses: The socio-spatial mobility of migrants in nineteenth-century Sundsvall, Sweden (The Demographic Data Base, Umeå University, 2003); “Vulnerablity among paupers: Determinants of individuals receiving poor relief in nineteenth-century northern Sweden”, The history of the family, 2006, 11, pp. 223-239. This article is based on the project ‘Crisis-biographies: life-course and time-geographical aspects on the causes, cures and consequences of individual crises’, as are two forthcoming publications by Vikström, i.e. “Societal change and individual past in connection with crime: Demographic perspectives of young offenders in nineteenth-century northern Sweden”, Continuity and change; “Life prior to crime and after: Developmental perspectives of young offenders in nineteenth-century northern Sweden, Journal of Social history.

Haut de page

Droits d’auteur

© Droz

Haut de page
  • Revues.org