Violence against women in Stockholm County: does marital status matter? Gloria Macassa, Anders Walander and Joaquim Soares

Gloria Macassa is based at Department of Public Health Sciences, Mid-Sweden University, Sundsvall, Sweden; at the Department of Occupational and Public Health Sciences, University of Ga¨vle, Ga¨vle, Sweden; and at the Department of Public Health Sciences, Division of Social Medicine, Karolinska Institutet, Stockholm, Sweden. Anders Walander is based at the Department of Public Health Sciences, Division of Social Medicine, Karolinska Institutet, Stockholm, Sweden. Joaquim Soares is based at the Department of Public Health Sciences, Mid-Sweden University, Sundsvall, Sweden and at the Department of Public Health Sciences, Division of Social Medicine, Karolinska Institutet, Stockholm, Sweden.

Abstract Purpose – The purpose of this paper is to investigate differences in victimization among single and married mothers using data (n ¼ 6,388 women) from the 2006 Stockholm County Public Health Survey (SCPHS), a cross-sectional survey based on a self-administered postal questionnaire. Design/methodology/approach – Results showed higher odds of victimization among lone mothers compared to married ones. Additionally, the greatest odds of victimization were observed among those with low education, low income, and decreased social and practical support. Findings – In Sweden, particularly in Stockholm, there is a need for future population-based surveys regarding the prevalence of violence and to identify high risk groups. Furthermore, it is crucial that these surveys include items that will enable investigation of direct links between violence and health effects, and the utilization of health care for these women. Originality/value – This paper is original and it addresses for the first time violence among single mothers within a social epidemiology perspective. Keywords Sweden, Violence, Aggression, Women, Social problems, Single mothers, Victimization, Stockholm, Social support Paper type Research paper

Introduction In many Western countries research has indicated that lone mothers experience both a socio-economic and a health disadvantage (Curtis and Phips, 2004; Benzeval, 1998). It is suggested that the differences in health between single and couple mothers are due to an uneven distribution of certain socio-demographic and socio-economic characteristics, which in turn are closely associated with both poorer health and increased health care use (Reissman, 1985; Whyke, 1992; Wennemo, 1994; Westin, 2007). Furthermore, others argue that marriage may have a salutogenic effect because partnership could bring better material resources and protect against risky negative health behaviours, vulnerabilities due to deficient social networks and a lack of social support (Whyke, 1992; Ga¨hler, 2006; Westin, 2007). However, Swedish single mothers seem to be in a more economically favourable situation than their counterparts in many other countries (Bradbury et al., 1999; Burstro¨m et al., 1999; Butterworth, 2004; Fritzell and Burstro¨m, 2006; Fritzell et al., 2007; Social Report, 1997; Whitehead et al., 2000). Furthermore, other studies have found that lone mothers are more likely than partnered mothers to report physical and sexual violence traumas (Butterworth, 2004), in particular from ex-partners (Chu et al., 2010). In a study of physical violence against women in the USA, Chu et al. (2010) found that the prevalence of abuse by a former partner was consistently higher than the prevalence of abuse by a current partner. Surveys asking people about their levels of exposure to crime have consistently shown that single mothers suffer particularly high levels of victimization by comparison with other social groups (Estrada and Nilsson, 2004). Therefore, family type has been considered an important risk factor for victimization (Estrada and Nilsson, 2004; Maxfield, 1987). In the

DOI 10.1108/17596591311313690 VOL. 5 NO. 2 2013, pp. 101-111, Q Emerald Group Publishing Limited, ISSN 1759-6599

j

JOURNAL OF AGGRESSION, CONFLICTAND PEACE RESEARCH

j

PAGE 101

criminology research literature, it is argued that the lower levels of victimization experienced by married mothers in comparison with single mothers is linked to activity patterns, which see married women spending a large proportion of their time at home and thus are more seldom exposed to risky situations (Maxfield, 1987). According to Hindelang et al. (1978), the so-called opportunity theories explain variations in the level of criminal victimization between different demographic groups (e.g. age, sex, family type and income) as a result of differences in the activity patterns of these groups. It is further argued that variations in lifestyle/routine activities involve disparities in the levels of exposure, including risk of victimization (Hindelang et al., 1978; Estrada and Nilsson, 2004). Consequently, opportunity theorists have assumed that the higher level risk of victimization experienced by single mothers is dependent on being forced into full-time work to a greater extent and that their status as singles involve relatively high levels of participation in night-life. However, these assumptions have been met with a considerable amount of criticism, not least because a significant deal of violence perpetrated against single mother’s takes places at home (Maxfield, 1987). Others argue that violence is more likely to be the result of structural factors, which determine where and how different groups live their lives rather than of activity patterns based on choices made at the individual level. For instance, poor people have little choice about these matters; if the situation is undesirable in any respect, those in poverty have little choice but to cope (Maxfield, 1987). However, many researchers working with victimization surprisingly have paid little attention to how socio-economic conditions in which individuals live their lives can impact on their victimization (Meier and Miethe, 1993; Smith and Jarjoura, 1989). In Sweden, although much is known about health outcomes (morbidity and mortality) among single mothers, there is scant information on their experiences of victimization. For instance, to our knowledge, only two studies have been carried-out, one using data from the 1988 to 1999 Annual Survey of Living Conditions conducted by Statistics Sweden (Estrada and Nilsson, 2004) and the other using data collected by a group of researchers who studied the prevalence of interpersonal violence in a nation-wide sample of 10,000 women (Lundgren et al., 2001). Estrada’s and Nilsson (2004) study, found high levels of victimization among single mothers, but the differences in levels of exposure to violence both within the group of single mothers and between this group and other women were primarily related to levels of individual welfare resources. Investigating a national sample of 10,000 women, Lundgren et al. (2001) observed that 45 percent of single mothers were subjected to violence by a former husband/cohabitant partner. In addition, the study reported that 27 percent of these women had been exposed to more than ten violent incidents each during the past 12 months. Violence in all its forms (e.g. physical assault) against single mothers has mental and physical adverse effects such as depression, anxiety, substance abuse and chronic pain, and they persist over time (Egeler et al., 2011). Conceptual framework In the literature about violence against women many frameworks have been used to explain the abuse ranging from the feminist approach which tends to present patriarchy as the main cause of violence (Dobash and Dobash, 1979; Yllo¨, 1993) to the resource approach which focus on one’s social or economic status (Goode, 1971). However, the most used framework to explain partner violence is the ecologic approach. Formulated initially by Bronfenbrenner’s (1986) ‘‘the Bronfenbrenner’s ecological theory of human development’’, it posits that human experiences are embedded in and shaped by multiple layers of contexts that are inter-related. According to this author, it is important to examine how various factors at multiple levels of one’s environment influence the phenomenon of interest. Accordingly, the ecologic approach accounts for the inter-relationships, between partner violence and a wide range of factors, including individual and contextual correlates (Bronfenbrenner’s, 1986). It has been suggested that violence is not randomly distributed, the same social factors that shape health, including education, income, work and employment, wealth and related conditions where we live, work and play, also are strongly linked to violence (Egeler et al., 2011). In the current study, instead of the Bronfenbrenner (1986) framework, we used a simplified version of the Diderichsen’s conceptual framework of policy impact on social pathways to

j

j

PAGE 102 JOURNAL OF AGGRESSION, CONFLICT AND PEACE RESEARCH VOL. 5 NO. 2 2013

inequalities in health (Diderichsen and Hallqvist, 1998). We adapted this framework to violence against single mothers (Figure 1). This allows the use of single motherhood to serve as an explanatory variable for potential differences in victimization instead of being just a covariate as used in various other studies. In the past years, some authors have suggested that violence should be viewed through the lens of social determinants of health (Eagler et al., 2011; Gill and Theriault, 2005; Werest and Merritt-Gray, 2004). First, our adapted framework considers that the individual social position of the mother (being a single mother) can influence exposure to violence (either physical or threat of violence). Different social positions might carry different risks for being exposed to interpersonal violence. Thus, that the position that single mothers have in a society, for example, might mean that they are exposed more often to a wide range of abuse risks (e.g. interpersonal violence) as compared to coupled mothers (mechanism I). In addition, specific exposures may or may not lead to experiences of violence for lone mother’s, depending on whether other contributing risk factors are present (mechanism II) (e.g. poverty, having an ex-partner; lack of education and occupation, presence or absence of social support). Second, the framework considers a societal perspective which focuses on how the social context interacts with and influences the pathways from social position to violence (and its potential physical and psychological consequences). The social context includes elements of the political, cultural, social and economic environment which characterize a society, of which policy is one component. In the current study, social context refers to the general policies for women in Sweden as well as other specific policies in Stockholm County. The framework also suggests that by viewing single motherhood as a social position implies that policy may influence the social position that single mothers occupy in society (entry point A). In addition, policies may influence: B

the characteristics and numbers of women in- and out from lone motherhood;

B

the laws of separation; and

Figure 1 The social basis of inequalities in victimization among single mothers

Society

Individual SOCIAL POSITION

CONTEXT (Sweden and Stockholm County)

Marital Status

A

B

POLICY CONTEXT

I

RISK FACTORS Living with children younger than 20 Low education and occupation Low-income (financial strain) Unemployment Ex-partner Relative poverty Low – social/practical support

– Welfare, Labour and Economic Policies

II

– Gender equality policies Victimization (Physical violence and Threat of violence)

– Laws of Divorce including children’s custody

Source: Adapted from Diderichsen and Hallqvist (1998)

j

j

VOL. 5 NO. 2 2013 JOURNAL OF AGGRESSION, CONFLICT AND PEACE RESEARCH PAGE 103

B

welfare benefits/social security for lone mothers which may or may not reduce the risk of relative poverty and laws regarding violence in general, but against women in particular.

For instance, it is suggested that local or national policies may be in place which not only influence the risk of being poor (as in entry point B) but also either soften or reinforce the effects of being poor. Furthermore, it is argued that such policies may be differential in their effect in influencing poor single mothers more than coupled mothers. Due to the fact that in Sweden single mothers experience more hardships than couple mothers (Fritzell and Burstro¨m, 2006; Fritzell et al., 2007) and the compelling evidence from studies from other developed countries (e.g. the UK, Australia and the USA) regarding their high risk of victimization (ABS, 1996; Benzeval, 1998; Chu et al., 2010), the current study aimed to investigate violence against single mothers living in Stockholm County. We hypothesize that the inequality in the distribution of some socio-economic/demographic characteristics might influence the risk of victimization, with high levels of victimization among single as compared to coupled mothers. Most importantly, the study addressed the following research questions: RQ1. Do single mothers experience higher levels of victimization compared to married mothers? RQ2. What are the factors which contribute for the risk of victimization?

Material and methods This study uses data from the 2006 Stockholm County Public Health Survey, a crosssectional survey based on a self-administered postal questionnaire. It is a randomly stratified sample (by area) and the initial sample was 57,009 persons aged 18-84 years living in Stockholm County. However, only 34,707 returned the questionnaire, which gave a response rate of 61 percent. Questions were asked about health, long-standing illness, self-reported specific illnesses or symptoms of diseases, housing, leisure and social relations, political activity, safety, security, violence, economy, employment and work environment. More details about the survey can be found elsewhere (Ja¨rleborg, 2006; Statistics Sweden, 2007). For the current study, we used a sample of 6,388 women. Specification and measurement of variables Violence was assessed using the following two questions, ‘‘Have you in the past 12 months been exposed to physical violence?’’ and the second, ‘‘Have you in the past 12 months been exposed to threat of violence that was so severe that you got afraid?’’ In the current study, women who answered yes to either one or both questions regarding violence were regarded as victims of ‘‘any violence’’. Marital status was defined in terms of being married (or living with a partner) or alone. Furthermore, lone mothers were defined as women who live alone with a child 20 years or younger. Education level was assessed from the Statistics Sweden’s LISA database from 2004 (Statistics Sweden, 2000a-c) and was grouped by the Swedish educational nomenclature SUN 2000 (Statistics Sweden, 2000a-c). For the purpose of current study, five categories of education were created: 1. less than two years secondary school; 2. secondary school, two years, or three years practical; 3. secondary school three years theoretical, or less than two years higher education; 4. higher education, two to three years; and 5. higher education, four years or more. Income was assessed from the Statistics Sweden’s LISA database from 2004, measured as disposable family income (according to number and age of family members), divided in four groups:

j

j

PAGE 104 JOURNAL OF AGGRESSION, CONFLICT AND PEACE RESEARCH VOL. 5 NO. 2 2013

1. 0-99 thousand SEK a year; 2. 100-149 thousand SEK a year; 3. 150-199 thousand SEK a year; and 4. . 200 thousand SEK a year. Register data of disposable family income is somewhat problematic, since it is not possible to fully identify unmarried cohabitants in the registers. Therefore, some cohabitants will be identified as singles, in regard of family income. Still, this fact is considered not to have any significant influence on the results of the analysis (Statistics Sweden, 2000a-c). An interaction variable capturing both social and practical support social support was created with the following levels: B

no practical and social support;

B

only social support;

B

only practical support; and

B

both practical and social support.

In the survey, social support was collected using the question: do you have any person(s) that can give you personal support to be able to solve problems or crises in your life? The alternatives were: B

always;

B

often;

B

seldom; and

B

never.

For practical support the question was ‘‘can you get help from a person(s) if you are sick or need to solve practical problems (e.g. repairing things at home help to buy groceries)’’. The alternatives were similar to those for the social support question. The four responses of varying degrees of yes or no were dichotomized into a new yes or no variable (Tables I and II). The variable living with economic strain was dichotomized in a yes/no format. Living with economic strain was based on a survey question about whether one during the past year has had to borrow money to handle the ordinary expenses. Data analyses The analyses were limited to mothers living with children younger than 20 years, and included descriptive, bivariate and multivariate statistics. Because we performed stratified sample and sub-group analyses in SAS 9.2 (SAS Institute Inc, 2010), the Proc. Survey freq. and Proc. Survey logistics, with domain analyses were applied. Tests and diagnostics were performed, including presence of missing-values, judgement of relevant confounders, and of multicolinearity, dispersion, model-fit and interaction effects. The analyses started with a large number of independent variables, which due to nonsignificance, interaction effects and problems of dispersion were reduced into the final model. No multicolinearity was present. In the regression analysis, we adjusted the odds ratio of violence between single and married with relevant confounders. Odds ratios were calculated with 95 percent confidence intervals.

Results As shown in Table I, single mothers reported more violence (17 percent) compared to married mothers (5 percent). As shown in Table II, single mothers also had a higher victimization odds ratio (both physical and threat) compared to married mothers. In the bivariate analysis, single mothers compared to married had an odds ratio of 3.85 (CI percent 3.05-4.86). However, controlling for other variables (age, education, income, social and practical support

j

j

VOL. 5 NO. 2 2013 JOURNAL OF AGGRESSION, CONFLICT AND PEACE RESEARCH PAGE 105

Table I Distribution of violence across different socio-economic and demographic variables Variable Marital status Married with children Single with children Age group 18-29 30-39 40-49 50-64 Education Less than two years secondary school Secondary school, two years, or three years practical Secondary school three years theoretical, or less than two years higher education Higher education, two to three years Higher education, four years or more’ Income 0-99 TSEK 100-149 TSEK 150-199 TSEK .200 TSEK Economic strain Yes No Social support Yes No Practical support Yes No

Violence

Yes (%)

Violence

No (%)

262 181

(5) (17)

5,022 923

(95) (83)

48 174 170 51

(11) (7) (7) (6)

397 2,367 2,418 763

(89) (93) (93) (94)

121

(15)

697

(85)

117

(8)

1,444

(92)

73 87 33

(6) (5) (4)

179 1,772 688

(94) (95) (96)

208 152 53 24

(12) (6) (4) (3)

1,569 2,331 1,171 742

(88) (94) (96) (97)

204 237

(16) (5)

1,115 4,811

(84) (95)

334 101

(6) (15)

5,344 570

(94) (95)

333 105

(6) (18)

5,440 468

(94) (82)

Source: Stockholm Public Health Survey (2006)

and economic strain) in the multivariate analysis reduced the odd ratio to 2.43 (1.85-3.19), albeit was still statistically significant. In addition, results also showed that mothers with less than two years of secondary school had a victimization odds ratio of 1.68 (CI 1.05-2.70). Similarly, mothers with income 0-99,000 SEK a year had a victimization odds ratio of 2.20 (CI 1. 31-3.70). Finally, mothers who experienced economic strain, no social and practical support had a victimization odds ratio of 2.05 (CI 1.56-2.67), 2.29 (CI 1.57-3.34), respectively.

Discussion Using marital status as a measure of social position to explain differences in victimization patterns, the current study found higher odds ratios of any type of violence (physical violence and threat of violence combined) among single mothers living in Stockholm County both in the bivariate and multivariate analyses. This is in agreement with findings from other Swedish studies using both population and registrar data, which observed increased odds ratios of violence against lone mothers (Estrada and Nilsson, 2004; Lundgren et al., 2001; Weitoft et al., 2002). These results are also in line with data from studies carried-out elsewhere (ABS, 1996; Cohen and Maclean, 2004; Zorrilla et al., 2010). For instance, in Australia, a study found that single women with ex-partners were more likely to have experienced violence, and the expartners were the most probable assailant (ABS, 1996). Furthermore, another study carriedout in Canada reported that 27 percent of women with no current partner living in the household were exposed to non-severe violence as compared with 11 percent of those with a common-law partner (Cohen and Maclean, 2004). In the multivariate analysis, controlling for other important covariates such as age, education, occupation, income, financial strain, lack of practical support reduced slightly the odds of

j

j

PAGE 106 JOURNAL OF AGGRESSION, CONFLICT AND PEACE RESEARCH VOL. 5 NO. 2 2013

Table II Odds ratio (OR) with 95 percent confidence intervals of violence by marital status Variable Marital status of the mother Bivariate analysis (unadjusted) Single with children ,20 Married with children ,20 Age group 18-29 30-39 40-49 50-64 Education Less than two years secondary school Secondary School two years, or three practical education Secondary school three years theoretical Higher education two to three years Higher education, four years or more Income 0-99 T-SEK 100-149 T-SEK 150-199 T-SEK .200 T-SEK Other social and economic variables Economic strain Social support*practical support No social and practical support Only social support Only practical support Both social and practical support

OR

95 percent CI

3.85 2.43 1

(3.05-4.86) (1.85-3.19)

1.59 1.25 1.09 1

(0.96-2.63) (0.84-1.84) (0.74-1.60)

1.68 1.21 1.07 1.02 1

(1.05-2.70) (0.76-1.93) (0.65-1.75) (0.63-1.64)

2.20 1.74 1.42 1

(1.31-3.70) (1.04-2.92) (0.80-2.51)

2.05

(1.56-2.67)

2.29 3.04 2.03 1

(1.57-3.34) (1.89-4.89) (1.28-3.21)

violence from 3.58 (CI 2.05-4.86) to 2.43 (CI 1.85-3.19). This suggests that some of these factors may mediate the relationship between violence and lone motherhood. For instance, mothers who had education less than two years, low income and financial strain were had higher odds of reporting violence in the past 12 months. The highest odds of violence were found among mothers who reported financial strain and low income (0-99 TSEK/year), respectively. It is suggested that single mothers are among the most economically and socially disadvantaged groups in many Western countries, experiencing greater levels of financial hardship, poverty and social exclusion than other family types (Crosier et al., 2007). Furthermore, our findings regarding differences in battered lone and married mothers reflects potential variations in socio-economic circumstances, a finding in line with previous studies investigating a variety of other outcomes (Benzeval, 1998; Whitehead et al., 2000). Lacks of social and practical support were associated with high odds of experiencing violence. Many authors argue that lack of social support could be one of the potential explanations for the health disadvantage of lone mothers (Hope et al., 1999; Whitehead et al., 2000; Whyke and Ford, 1992) and that may influence health directly or indirectly by increasing vulnerability to life events and adversities. Previous studies carried out in Sweden regarding other outcomes have found that divorced women experienced less access to social support than other women or simple they perceived lower quality in their social networks (Ga¨hler, 1998). Social support as a determinant of health means that people are healthier when they feel safe supported by and connected to others, and when they are able to trust their family members, friends, neighbours and the community in general. Social support is believed to be important because is a key factor in helping people to solve problems and survive life crises. Having someone to confide in, someone to count on in times of crises, someone you can seek advice from, is good for health (Ga¨hler, 1998; Hope et al., 1999; Whitehead et al., 2000). In relation to victimization, not having a confidant at one’s disposal might make it much more difficult to take steps to end the abuse. Thus, a week network of support and a situation of

j

j

VOL. 5 NO. 2 2013 JOURNAL OF AGGRESSION, CONFLICT AND PEACE RESEARCH PAGE 107

relative social isolation can be considered a potential risk factor for violent victimization although much remains to be explored on this regard (Ga¨hler, 1998). From the policy perspective and based on the above mentioned theoretical framework, studying violence against single mothers through a focus on social determinants of health will lead to a better understanding of the various policy initiatives both at individual and societal level that might have the potential to reduce abuse among single mothers and potentially improve their overall health. Obviously this is a complex multidimensional problem with no single factor emerging as the ‘‘satisfactory explanation’’. Individual level risks include low levels of education, income, history of abuse and behaviours (substance abuse), life-style training, beliefs and behaviours, lack of social support, social isolation. Individual level approaches include life-skills training focused on promoting-risk reducing attitudes, beliefs and behaviour (see arrows in Figure 1). Societal approaches to prevent violence against single women would include more broadly policies aimed on both fundamental conditions that foster violence and the policies that foster them. For instance, to address social disadvantage and how damaging effects can be ameliorated through social policies such as welfare subsidies and employment opportunities for single mothers, job training and creation, economic development for low-income communities, increased education quality, relative poverty reduction and other social policy approaches (Egeler et al., 2011). Community development initiatives might help strengthen local society networks and decrease the isolation experiences by certain women at risk, or suffering from abuse (including single mothers) which are the target of this study. In addition, other social policy initiatives addressing the income needs of women and more generally at decreasing income inequalities in the society-are likely to work towards the health of the population as a whole (Vaillancourt et al., 2004) and will also contribute to the prevention and reduction of various forms of violence, including violence among single mothers. Although not an objective of our study, descriptive analyses concerning the type of perpetrator of the violence experienced by the lone mothers indicated that many have been victimized by ex-partners (Table III). McInnes (2002) argues that poverty may stop single mothers from being able to sustain separation, especially when the violent partner is seen as the only source of available income to meet her own and the children’s needs. However, in the current study, it was not possible to determine whether violence was a cause of separation among the lone mothers, but given the high levels of violence reported among them, we can only infer that violence may have been part of the reason for separation or divorce.

Strengths and limitations The current study has strengths and limitations. The main strengths are: B

The study is based on data which has a high degree of reliability and validity, the Stockholm Public Health Survey is used by the Stockholm Health authorities as a source to monitor population health.

B

Information regarding violence was collected only for the past 12 months in order to reduce the occurrence of information bias as women would be more likely to remember violent events which have occurred in the recent past.

B

The study is first to our knowledge to use a social epidemiology perspective to single mothers violence in the Swedish context.

Table III Violence of threat of violence perpetrated by ex-partner Variable

Violence

Marital status Single with children ,20 Married with children ,20 Total

j

j

PAGE 108 JOURNAL OF AGGRESSION, CONFLICT AND PEACE RESEARCH VOL. 5 NO. 2 2013

Yes (%)

19 (8) 74 (43) 93

Violence

No (%)

Total (%)

226 (92) 103 (57) 329

245 (100) 177 (100) 422

However, the study has caveats: violence was defined as ‘‘any type of violence’’, which included both physical and threat of violence due to small numbers. In the data, the majority of women experienced physical violence, similar to the pattern which was found at a National level by Lundgren et al. (2001); and the survey did not include women in extreme situations of social exclusion and poverty such as homeless which are likely to be even more exposed to violent victimization.

Conclusion This study found high odds of victimization among lone mothers compared to married mothers. In addition, the greatest odds of victimization were observed among those with low education, income and low social and practical support. In Sweden, particularly Stockholm, there is a need for future population-based surveys of violence to identify high risk groups and the prevalence of violence in these groups. Furthermore, it is crucial that these surveys include items that will enable direct links between violence and health effects and the utilization of health care for these women. Implications for practice B

B

B

Particularly high rates of victimization were found for lone mothers, highlighting the need for further support from relevant services, for this population. This study utilised a highly reliable surveying method, but future population-based surveys need to be more able to identify high risk groups. Researchers should aim to further investigate the links between violent victimization and health effects so that more directed health care can be provided. This is specifically the case for women in extreme circumstances of social exclusion, such as homelessness.

References ABS (1996), ‘‘Women’s safety after separation’’, Catalogue Number 4128.0, Australian Bureau of Statistics, Canberra. Benzeval, M. (1998), ‘‘The self-reported health status of lone parents’’, Social Science & Medicine, Vol. 46, pp. 1337-53. Bradbury, B. and Ja¨ntti, M. (1999), ‘‘Child poverty across industrialized nations: innocenti occasional papers’’, Economic and Social Policy Series No. 71, UNICEF International Child Development Center, Florence, pp. 1-70. Bronfenbrenner, U. (1986), ‘‘Ecology of the family as a context for human development: research perspectives’’, Developmental Psychology, Vol. 22, pp. 723-42. Burstro¨m, B., Diderichsen, F., Shouls, S. and Whitehead, M. (1999), ‘‘Lone mothers in Sweden: trends in health and socioeconomic circumstances 1979-1995’’, Journal of Epidemiology and Community Health, Vol. 53 No. 12, pp. 750-6. Butterworth, P. (2004), ‘‘Lone mothers experience of physical and sexual violence: association with psychiatric disorders’’, British Journal of Psychiatry, Vol. 184, pp. 21-7. Chu, S., Goodwin, M. and D’Angelo, D. (2010), ‘‘Physical violence against US women around the time of pregnancy 2004-2007’’, American Journal of Preventive Medicine, Vol. 38, pp. 317-22. Cohen, M. and Maclean, H. (2004), ‘‘Violence against Canadian women’’, BMC Women’s Health, Vol. 4, p. S22. Crosier, T., Butterworth, P. and Rodgers, B. (2007), ‘‘Mental health problems among single and parented mothers: the role of financial hardship and social support’’, Social Psychiatry, Psychiatric Epidemiology, Vol. 42, pp. 6-13. Curtis, L. and Phips, S. (2004), ‘‘Social transfers and the health status of mothers in Norway and Canada’’, Social Science & Medicine, Vol. 58, pp. 2499-507.

j

j

VOL. 5 NO. 2 2013 JOURNAL OF AGGRESSION, CONFLICT AND PEACE RESEARCH PAGE 109

Diderichsen, F. and Hallqvist, J. (1998), ‘‘Social inequalities in health: some methodological considerations for the study of social position and social context’’, in Arve-Pares, B. (Ed.), Inequality in Health: A Swedish Perspective, Swedish Council for Social Research, Stockholm. Dobash, E. and Dobash, R. (1979), Violence Against Wives, The Free Press, New York, NY. Egeler, S., Barclay, C., Grossman-Kahn, R. and Brevaman, P. (2011), Exploring the Social Determinants of Health, Violence, Social Disadvantage and Health, Robert Wood Johnson Foundation, The Foundation for Vulnerable Population’s, Princeton, NJ, pp. 1-19. Estrada, F. and Nilsson, A. (2004), ‘‘Exposure to threatening and violent behaviour: the significance of lifestyle, neighbourhood and welfare situation’’, British Journal of Criminology, Vol. 44, pp. 168-87. Fritzell, S. and Burstro¨m, B. (2006), ‘‘Economic strain and self-rated health among lone and couple mothers in Sweden during the 1990s compared to the 1980s’’, Health Policy, Vol. 79, pp. 253-64. Fritzell, S., Ringba¨ck Weitoft, G., Fritzell, J. and Burstro¨m, B. (2007), ‘‘From macro to micro: the health of Swedish lone mothers during changing economic and social circumstances’’, Social Science & Medicine, Vol. 65, pp. 2474-88. Ga¨hler, M. (1998), ‘‘Life after divorce: economic and psychological well being among Swedish adults and children following family dissolution’’, doctoral thesis, Swedish Institute for Social Research, Stockholm. Ga¨hler, M. (2006), ‘‘To divorce is to die a bit: a longitudinal study of marital disruption and psychological distress among Swedish women and men’’, The Family Journal, Vol. 14, p. 372. Gill, C. and Theriault, T. (2005), ‘‘Connecting social determinants of health and women abuse: a discussion paper’’, 2nd Atlantic Summer Institute on Health and Safe Communities, Atlanta, GA, 23-26 August, pp. 1-19. Goode, W. (1971), ‘‘Force and violence in the family’’, Journal of Marriage and the Family, Vol. 33, pp. 624-36. Hindelang, M., Gottfredson, M. and Garofalo, J. (1978), Victims of Personal Crime: An Empirical Foundation for a Theory of Personal Victimization, Ballinger, Cambridge, MA. Hope, S., Power, C. and Rodgers, B. (1999), ‘‘Does financial hardship account for elevated psychological distress in lone mothers?’’, Social Science & Medicine, Vol. 49, pp. 1637-49. Ja¨rleborg, A. (2006), En o¨versiktlig beskrivning av Folkha¨lsoenka¨t 2006, version 1 (Synopsis of Public Health Survey 2006, Version 1), Department of Epidemiology, Stockholm County Council, Stockholm (in Swedish). Lundgren, E., Heimer, G., Westerstrand, J. and Kalliokoski, A. (2001), Captured Queen: Men’s Violence Against Women in ‘‘Equal’’ Sweden – A Prevalence Study, Swedish National Agency for Crime and Uppsala University, Stockholm. McInnes, E. (2002), ‘‘Single mothers and system responses to violence against women and children’’, Stating Women’s Health, Vol. 2, pp. 1-7. Maxfield, M.G. (1987), ‘‘Household composition, routine activity and victimization: a comparative analysis’’, Journal of Quantitative Criminology, Vol. 3, pp. 301-20. Meier, R.F. and Miethe, T.D. (1993), ‘‘Understanding theories of criminal victimization’’, in Tonry, M. (Ed.), Crime and Justice: A Review of Research, Vol. 17, University of Chicago Press, Chicago, IL, pp. 459-97. National Board of Health and Welfare (1997), Social Report, National Board of Health and Welfare, Stockholm. Reisman, C.K. and Gestel, N. (1985), ‘‘Marital dissolution and health: do males or females have a greater risk?’’, Social Science & Medicine, Vol. 20, pp. 627-35. Reisman, G.N. (1985), ‘‘Marital dissolution and health: do males or females have a greater risk?’’ Social Science and medicine, Vol. 20, pp. 627-35. SAS Institute Inc (2010), SAS Version 9.1, SAS Institute Inc, Cary, NC. Smith, D.A. and Jarjoura, R.G. (1989), ‘‘Household characteristics, neighbourhood composition and victimization risk’’, Social Forces, Vol. 68, pp. 621-40. Social Report (1997), National Board of Health and Welfare, Stockholm. Statistics Sweden (2000a), Standard for Classification of Swedish Education, Statistics Sweden, Stockholm (in Swedish). Statistics Sweden (2000b), SUN 2000: Svensk utbildningsnomenklatur, Standard fo¨r klassificering av svensk utbildning, SUN 2000: Swedish Education Nomenclature, Statistics Sweden, Stockholm.

j

j

PAGE 110 JOURNAL OF AGGRESSION, CONFLICT AND PEACE RESEARCH VOL. 5 NO. 2 2013

Statistics Sweden (2000c), Yrkesklassificering i FoB 85 enligt nordisk yrkesklassificering (NYK) och socioekonomisk indelning (SEI) (Occupation Classification in FoB 85 According to the Nordic Occupation Classification (NYK) and Socioeconomic Classification (SEI)), Statistics Sweden, Stockholm (in Swedish). Statistics Sweden (2007), Ha¨lsoenka¨t 2006: en underso¨kning om ha¨lsa och levnadsfo¨rha˚llanden i Stockholms la¨n, teknisk rapport, utkast (Health Survey 2006: A Survey About Health and Living Conditions in Stockholm County), Technical Report, Draft, Survey Unit, Statistics Sweden, Stockholm (in Swedish). Vaillancourt, Y., Aubry, M., Keanny, L. and Tremblay, L. (2004), ‘‘Canada: a Quebec viewpoint’’, in Raphael, D. (Ed.), Social Determinants of Health: Canadian Perspectives, CSPL, Toronto, pp. 311-30. Weitoft, G.R., Haglund, B. and Rose, M. (2002), ‘‘Mortality among lone mothers in Sweden: a population study’’, International Journal of Epidemiology, Vol. 31, pp. 573-80. Wennemo, I. (1994), ‘‘Sharing the costs of children: studies on the development of family support in the OECD countries’’, PhD dissertation, Swedish Institute of Social Research, University of Stockholm, Stockholm. Werest, J. and Merritt-Gray, M. (2004), ‘‘Family violence as a social determinant of health’’, in Ruqseni, J. (Ed.), Determinants of Health: An Atlantic Perspective, University of Brunswick, Federicton, pp. 86-108. Westin, M. (2007), Health and Health Care Utilization Among Swedish Single Parents Families, Acta Universitatis Uppsaliensis, Uppsala University, Uppsala. Whitehead, M., Burstro¨m, B. and Diderichsen, F. (2000), ‘‘Social policies and the pathways to inequalities in health: a comparative analysis of lone mothers in Britain and Sweden’’, Social Science & Medicine, Vol. 50, pp. 255-70. Whyke, S. and Ford, G. (1992), ‘‘Competing explanations for associations between marital status and health’’, Social Science & Medicine, Vol. 34, pp. 523-32. Yllo¨, K. (1993), ‘‘Through a feminist lens: gender, power, and violence’’, in Gelles, R. and Loseke, D. (Eds), Current Controversies on Family Violence, Sage, Newbury Park, CA, pp. 47-62. Zorrilla, B., Pires, M., Morant, C., Seoane, L., Sanches, L., Galan, I., Aguire, R., Ramirez, R. and Durban, M. (2010), ‘‘Intimate partner violence: last year prevalence and association with socio-economic factors among women in Madrid, Spain’’, European Journal of Public Health, Vol. 20, pp. 169-75.

Further reading Forsen, A. and Janrlet, U. (1999), ‘‘Ensamsta˚ende mo¨drrar har o¨kad risk fo¨r oha¨lsa’’ (‘‘Lone mothers are at increased risk of ill health’’), Allma¨nmedicin, Vol. 12, pp. 72-6. Gelles, R.J. (1983), ‘‘An exchange/social control theory’’, in Finkelhor, D., Gelles, R., Straus, M. and Hotaling, G. (Eds), The Dark Side of Families: Current Family Violence Research, Sage, Beverly Hills, CA, pp. 151-65. Giles-Sims, J. (1983), Wife-Beating: A Systems Theory Approach, Guilford, New York, NY. Isacsson, A., Hanson, B.S., Janzon, L. and Kugelberg, G. (1992), ‘‘The epidemiology of sick leave in a urban population in Malmo¨, Sweden’’, Scandinavian Journal of Social Medicine, Vol. 4, pp. 234-9. Mastekaasa, A. (2000), ‘‘Parenthood, gender and sickness absence’’, Social Science & Medicine, Vol. 50, pp. 1827-42.

Corresponding author Gloria Macassa can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

j

j

VOL. 5 NO. 2 2013 JOURNAL OF AGGRESSION, CONFLICT AND PEACE RESEARCH PAGE 111