Batterer Intervention Programs in Spain: An Analysis of Their Effectiveness

Introduction

Intimate partner violence against women (IPVAW) is a major social and women’s health concern (Devries et al., 2013). Overall, 35% of women worldwide have experienced either physical and/or sexual intimate partner violence or non-partner sexual violence, and almost one third (30%) of all women who have been in a relationship have experienced physical and/or sexual violence by their intimate partner. As such, much effort has gone into providing services and support for victims of IPVAW. By contrast, intervention programs focusing on the batterers themselves have received far less attention (Cluss & Bodea, 2011).

In this context, batterer intervention programs (BIPs) emerged in the United States and Canada between 1970 and 1980. They reached Europe in the 1980s, becoming established, for instance, in Germany in 1984 or in Norway in 1986 (Feder & Wilson, 2005; Geldschläger et al., 2010; Subirana & Pueyo, 2013). Nowadays, estimates point toward there being more than 2,500 of these programs in the United States and more than 500 in Europe (Geldschläger, Ginés, & Ponce, 2009). This matter has aroused the interest of researchers, and as a result, there are a growing number of publications on the impact and characteristics of these programs (Ginés, Geldschläger, Nax, & Ponce, 2015). This research has stimulated intense debate about the programs’ theoretical framework, relevance, and effectiveness (Boots, Wareham, Bartula, & Canas, 2016; Eckhardt, Murphy, Black, & Suhr, 2006; Eckhardt et al., 2013; McGinn, Taylor, McColgan, & Lagdon, 2015; Stuart, Temple, & Moore, 2007). Consequently, BIP effectiveness is now one of the controversies surrounding IPVAW (Langhinrichsen-Rohling, 2010). Several studies have summarized the results of these interventions and conclude that batterers who do not complete the programs present a higher likelihood of recidivism than those who do complete it; but they also find that completing treatment is usually associated with only minimal decreases in IPVAW relapse, some of these intervention programs have drop-out rates of around 50%, and many men who attend BIPs continue to show high levels of resistance to change (Arias, Arce, & Vilariño, 2013; Babcock, Green, & Robie, 2004; Eckhardt et al., 2006; Feder & Wilson, 2005). In fact, international research places the recidivism rate of batterers at around 20% to 30% (Arias, Arce, & Novo, 2014), and previous meta-analysis studies have estimated this recidivism between 21% (according to official reports) and 35% (if the information comes directly from the couples; Arias et al., 2013; Babcock et al., 2004).

In summary, as Stuart et al. (2007) point out, numerous studies, qualitative reviews, and meta-analyses have repeatedly arrived at a similar conclusion: BIPs have a small, often insignificant, effect in reducing partner violence, and there is empirical evidence that has shown that current BIPs are not very effective. In fact, Babcock et al. (2004) and Eckhardt et al. (2013) affirm that the more rigorous the methodology used to evaluate programs is, the more their effectiveness diminishes. Some of the reasons for this ineffectiveness are hypothesized to be related to methodological problems regarding the implementation of BIPs and how outcomes are measured (Boira & Tomás-Aragonés, 2011; Carbajosa & Boira, 2013; Echeburúa, 2013; Hamilton, Koehler, & Losel, 2013); or more specifically that, batterers are usually court-mandated to attend BIPs and may be unwilling or unmotivated to accept responsibility for being violent; BIPs receive inadequate funding and, thus, have limited resources for development; interventions are seldom tailored to specific batterers’ characteristics; or programs are rushed into use and mandated by states before their effectiveness can be rigorously evaluated.

Hamilton et al. (2013) present the results of a survey of 54 programs in place across 19 European countries to analyze BIP practices and outcomes. They find a wide disparity in approaches to handling batterers and a dearth of high-quality evaluation across the different countries. Ginés et al. (2015) carried out a similar study as a part of the IMPACT Evaluation of European Perpetrator Programmes project. Specifically, the persons in charge of 134 BIPs from 22 European countries (34 from Spain) responded to a survey concerning the designs, methods, and instruments used in the outcome monitoring of BIPs, their barriers, and the need for improvements in implementation. Their results show that 20% of programs do not measure outcomes at all, there is a large variety of methods and instruments being used, and only a small percentage of programs make use of victim/survivor data or follow-up data. Akoensi, Koehler, Losel, and Humphreys (2013) present the results of a systematic review of European evidence on the effectiveness of BIPs (12 studies, four of which were carried out in Spain, that applied treatment to a total of 1,586 batterers, with sample sizes ranging from nine to 322 men). Although the evaluations suggested various positive post-treatment effects, methodological problems relating to the design of these evaluations do not allow for the findings to be attributed to the programs. Hester, Walker, O’Prey, and Budde (2014) also carried out an analysis of research studies evaluating BIPs in Europe (65 studies, 22 from Spain). Fifty-seven percent of the studies reported moderately positive or promising results. However, they were not statistically significant, and the conclusions obtained were focused on analyzing the reported impact of the BIP by European region.

In Spain, BIPs were first implemented in 1995 (Echeburúa, Corral, Fernández-Montalvo, & Amor, 2004; Echeburúa, Fernández-Montalvo, & Amor, 2006), and after the passing of Organic Law 1/2004 (December 28, 2004) on Comprehensive Protection Measures against Gender Violence, they have become essential intervention both inside and outside of prisons for enforcing sentencing terms given by Spanish judges (Millana, 2011; Subirana & Pueyo, 2013). In fact, the law requires IPVAW offenders with suspended prison sentences and with sentences not higher than 2 years to participate in a treatment program. In this context, some of the intervention programs performed on men who abuse their partners have been evaluated (Carrasco-Portiño, Vives-Cases, Gil-González, & Álvarez-Dardet, 2007) as have batterer psychotherapy-oriented treatments conducted in prisons (Millana, 2011).

In summary, the analysis of the literature about batterers’ intervention reveals a proliferation of studies and the growing attention and interest given to this phenomenon. Such interest is not equal in the different geographical areas. However, in the last years, European countries have made positive efforts on this matter. In this context, the purpose of this article is to provide a systematic review of the evidence regarding the effectiveness of BIPs implemented in Spain as provided by research published in peer-reviewed journals and scientific reports. The specific contribution of this article is to present results from the greatest number of the main BIP programs implemented in our country from the 1990s.

Method

Results

The review carried out shows, first, that we obtained data concerning the effectiveness of three types of BIPs applied in Spain: intervention programs carried out inside prisons, intervention programs targeted toward batterers whose sentences have been suspended under the condition that the batterer promises to undergo this type of treatment (court-mandated interventions), and intervention programs conducted in a community context where batterers came in as volunteers to receive treatment. Also, and as is usual in this field (Feder & Wilson, 2006), the BIPs employed in Spain differ in how they help batterers to acknowledge and stop behaving violently: Cognitive-behavioral treatment considers violence to be a learned behavior, requiring batterers to recognize that the abuse is under their control and that they can behave differently; gender perspective programs (such as the Duluth model) consider intimate partner violence as being rooted in patriarchal societal beliefs that portray men as having the right to exert power and control over women; and, a number of BIPs incorporate aspects of both. Table 1 in the online appendix summarizes the main characteristics of the 13 programs analyzed, and in each case includes a reference.

Second, Table 2 in the online appendix summarizes the main results obtained after analyzing these intervention programs in terms of their effectiveness. As can be seen, in 46% of cases, the assessment focuses on evaluating whether there is an improvement in the psychological variables of abusers. In cases where the rate of success or failure is analyzed (in seven of 13 BIP), it is important to note on one hand, the high drop-out rates observed (5.7%-57%), and on the other hand, the wide variability in success and improvement rates obtained (23.8%-93.5% of completers).

Discussion

Identifying and measuring outcomes is a problematic dimension of BIP evaluation: What to measure? When to measure it? How to measure it? In spite of the numerous and diverse types of operative assessment for the broad spectrum of offender rehabilitation programs, there are two main subtypes of measurements (Arias et al., 2014; Arias et al., 2013; Babcock et al., 2004; Echeburúa, Sarasúa, Zubizarreta, & Corral, 2009; Novo, Fariña, Seijo, & Arce, 2012): those of overall recidivism and those of specific rehabilitation targets, in particular psychopathological variables and variables of cognitive competence.

Because a major goal of BIPs is to reduce future assaults, in general, a program is considered effective if rates of aggressive behaviors are significantly reduced as a result of the intervention (Arias et al., 2014; Cluss & Bodea, 2011). And, as is most common in BIP research to date, the primary outcome measure used is recidivism. However, the assessment of recidivism or recurrence is a complex and controversial enterprise for a number of reasons. First, there is no agreement on how to estimate it. Thus, it has been measured in terms of new arrest warrants (police recidivism), new court cases (judicial recidivism), or new incarcerations (prison recidivism), and by reports, offender’s self-reports, or external reports (e.g., victim, partner, family, social services; Novo et al., 2012), and each has its advantages and disadvantages. In some cases, evaluations of BIPs include simultaneously different types of information. Second, there is no scientific evidence substantiating a time-based criterion for evaluating a re-offender follow-up period. And third, it is difficult and expensive to find and follow up with batterers years later.

The second type of measurement involves the assessment of specific rehabilitation targets, in particular, cognitive competence variables (e.g., cognitive skills, anger management, problem-solving skills, self-control, responsibility attribution for one’s actions), psychopathological variables (e.g., depression, anger, hostility), self-perception, and self-esteem. In this context, measuring outcomes in multiple ways is likely a useful solution for providing better estimations and a better understanding of the critical processes of change in the batterer. But, although reducing violence is a key goal, our review shows that BIPs carried out in Spain primarily use self-reports and standardized clinical measures for evaluating outcomes, and they use improvement in psychological and psychopathological variables of batterers as a measurement outcome effectiveness.

Other methodological and analytical difficulties regarding BIPs, which were detected in previous studies (Bowen & Gilchrist, 2004; Cluss & Bodea, 2011) as well as in our results, include high drop-out rates; difficulty in adhering to the demands of experimental situations in real BIP settings; the variability of programs’ content, approaches, and jurisdictions; the multiple causes of IPVAW; and the existence of heterogeneous subtypes of persons who engage in IPVAW.

An added problem when considering BIP effectiveness is the high rate of dropout, which may be between 40% and 90%, according to some studies (Lila, Català et al., 2010; Loinaz & Echeburúa, 2010; Olver, Stockdale, & Wormith, 2011; Subirana & Pueyo, 2013). This finding is particularly important because the drop-out rate has been associated with recidivism (Murphy & Ting, 2011). In addition, the variables that predict dropout tend to be the same as those that predict recidivism (Jewell & Wormith, 2010). These high drop-out rates have been attributed to, among other variables, the perpetrators’ heterogeneity and that standardized intervention programs fail to take into account their individual needs and characteristics (Loinaz & Echeburúa, 2010; Olver et al., 2011). Lack of motivation on the part of perpetrators undergoing court-mandated (involuntary) treatment could be another causal factor in these high drop-out rates (Boira & Tomás-Aragonés, 2011; Carbajosa & Boira, 2013; Fernández-Montalvo, Echauri, Martínez, & Azcárate, 2012).

Despite all these difficulties, research on BIP effectiveness has been increasing in volume and quality over the past decades (Cluss & Bodea, 2011). However, as our results show, the methodological limitations of outcome studies still preclude our ability to determine the effectiveness of these interventions. The United Kingdom’s National Institute for Health and Care Excellence (NICE; 2014) arrives to a similar conclusion:

There is a lack of consistent evidence on the effectiveness of programmes for people who perpetrate intimate partner violence and abuse. The Programme Development Group (PDG) noted that some evaluations take account of the partner’s health and wellbeing and include their perception of any changes in the perpetrator behaviour. However, these tend to be small-scale, uncontrolled studies (p. 33). There is a lack of large, robust studies of interventions for people who perpetrate abuse. The majority were non-experimental (primarily before-and-after studies). Often they did not include a comparison group, had relatively small sample sizes, reported high rates of attrition and lacked follow up beyond programme completion. (p. 59)

In summary, the results obtained suggest that, to continue improving, evaluations of BIPs must supplement the data regarding effectiveness with some methodological changes, that is, the use of experimental or quasi-experimental designs with relevant controls, using broad definitions of abuse, the use of multiple outcome measures, giving preference to victim reports over official reports of recidivism, measuring the program’s impact on the victim’s well-being, completion of longer follow-up intervals for determining outcomes, achieving greater follow-up retention rates, and an accurate assessment of program implementation (Cluss & Bodea, 2011; Feder & Wilson, 2005; Loinaz, Echeburúa, Ortiz-Tallo, & Amor, 2012). In fact, according to some meta-analysis results (Feder & Wilson, 2005), there are no statistically significant differences in outcome by type of report (e.g., self-reports, police reports), but quasi-experimental studies using victim reports show improved validity and more adequate estimates of outcomes than experimental studies using solely official records of reassault.

So, like Bowen and Gilchrist (2004), we can conclude that the focus on behavioral outcomes in evaluations of batterer programs has been too narrow, and that a more holistic approach may be required to determine what works, for whom, and under what circumstances. By including these new analyses and considerations, our understanding of the mechanisms responsible for behavioral outcomes will be greatly improved and may lead to the design and implementation of more effective programs in the future.

In summary, the results obtained in this revision allow to corroborate the assumption that associated with BIP is a lot of controversy, since its construction, its implementation, its assessment, and its effectiveness. In fact, the results point in opposite directions. But, despite the unclear results about intervention efficacy, it can be sustained that small effects in recidivism reduction can represent a considerable social impact, especially for the victims. In this sense, in general, and despite the controversies, it is important to emphasize the relevance of the development and implementation of re-education programs about intimate partner violence, to reduce the recidivism and to protect the victims. The most recent studies carried out in Spain (e.g., Fernández-Montalvo, Echauri, Martínez, Azcárate, & López-Goñi, 2015) have found similar results to earlier studies, analyzed in our revision, and they strengthen these conclusions.

And finally, and in general, according the results obtained in some previous studies (Arias et al., 2013), we can say that the type of intervention is less important than another BIP characteristics. Specially, the voluntary participation in the program is highly recommended (Millana, 2011), because a lack of voluntariness can negatively affect the entire treatment process and its effectiveness (Expósito & Ruiz, 2010; Quinteros & Carbajosa, 2008).

Supplemental Material