Abstract
The tendency to label all elder abuse perpetrators as the “bad guys” has diminished our ability to respond effectively. A review of the literature demonstrates that elder abuse perpetrators are in fact heterogeneous with important differences across types of abuse. A reformulation of perpetrator interventions away from a singular emphasis on prosecution to meaningful alternatives that utilize criminal justice and/or therapeutic approaches tailored to the needs of the case is needed. These interventions must incorporate the needs of both victims and perpetrators, take into consideration the type of abuse involved, acknowledge the variations in perpetrator culpability, and recognize the continuum of complexity among these cases. Without addressing these nuances, intervention and prevention efforts will be futile if not harmful.
As opposed to crimes against older adults (Bachman, Dillaway, & Lachs, 1998; Smith, 2012), elder abuse has been defined as a single or repeated act, or lack of appropriate action, occurring within any relationship where there is an expectation of trust which causes harm or distress to an older person and typically encompasses six actions: physical abuse, caregiver neglect, financial exploitation, psychological abuse, sexual abuse, and abandonment 1 (World Health Organization [WHO], 2002). And yet the field has suffered from a tendency to treat elder abuse as a unitary phenomenon (Jackson & Hafemeister, 2013a), with a singular focus on victims (Quinn & Zielke, 2005). For example, Wolf (2003) wrote “If you look at the laws of the states, I’m not sure the perpetrator is mentioned at all” (p. 245). This situation is interesting given the rich history of perpetrator research in the criminal justice literature (DeLisi & Piquero, 2011), with a victim body of literature only recently developed (Mastrocinque, 2010).
The egregious exclusion of perpetrators from the elder abuse dialogue disregards the role played by perpetrators in maintaining abusive relationships (Alon & Berg-Warman, 2013; Henderson, Varble, & Buchanan, 2004; Kosberg, 2014). In addition, it has contributed to our lack of empirical understanding of elder abuse perpetrators, with the most current comprehensive review of the literature devoted exclusively to elder abuse perpetrators being published in 1990 (Wolf, 1990). In turn, the lack of empirical understanding of elder abuse perpetrators has impeded the development of alternatives to prosecution, currently the primary form of intervention (Wolf, 2003).
This article is intended to update the Wolf (1990) review. Following a brief historical overview of elder abuse perpetrators and the challenges associated with defining this group, this article reviews the current state of knowledge with regard to elder abuse perpetrators residing in the community. Four areas of research are reviewed: (a) studies showing that the characteristics of both the victims and perpetrators predict different forms of elder abuse, (b) studies showing the importance of differentiating by type of abuse across a range of domains, (c) studies identifying variations in perpetrator culpability, and (d) studies demonstrating the continuum of complexity that exists among elder abuse cases generally. Theory and implications for intervention are discussed. That is, this review culminates in a conclusion that elder abuse perpetrators are considerably more heterogeneous than is commonly portrayed. Therefore, an imperative exists to develop and evaluate new and innovative interventions that incorporate the above-identified realities. This will require a reformulation of perpetrator interventions away from a singular emphasis on prosecution to meaningful alternative interventions that utilize criminal justice and/or therapeutic approaches tailored to the specific needs of the case.
Brief History of Elder Abuse
A Social Services Approach
During the 1950s, concern emerged about the increasing numbers of very old and functionally impaired individuals (a condition later termed self-neglect). This recognition led to the formulation of protective services as an intervention strategy (Brownell & Abelman, 1998), focused on identifying and meeting the day-to-day needs of vulnerable older adults such as food and housing assistance. Importantly, then, protective services as it was initially conceptualized was not designed to address elder abuse, but rather, self-neglect and consequently, there were no perpetrators with which to be concerned.
In the wake of the discovery of child abuse in the 1960s, and later domestic violence in the 1970s, an interest in elder abuse emerged, due in part to a surge in reports of elder abuse (Wolf, 2003). Advocates for protective services took advantage of these reports to call for expanding the use of protective services to respond to the needs of elder abuse victims (Wolf, Hodge, & Roberts, 1998). Without much consideration for whether protective services was appropriate for responding to elder abuse, the vulnerable elder model was retained based on the assumption that, similar to child abuse, elder abuse stemmed from the stress of caring for vulnerable (functionally impaired) elderly individuals (Barber, 2008; Kohn, 2003), referred to as caregiver stress (Steinmetz, 1978). Perpetrators were now part of the equation, but they were conceptualized as caregivers in need of social support.
Criminalization of Elder Abuse
An important philosophical shift in our conceptualization of elder abuse occurred in the 1990s, although the underpinnings were present a decade earlier. Groundbreaking research by Georgia Anetzberger (1987) and Karl Pillemer (1985) began to challenge the largely held assumptions underlying the caregiver stress model and transformed our conceptualization of elder abuse from caregiver stress to abuser psychopathology. When it was revealed that elder abuse differed in meaningful ways from child abuse, researchers shifted their focus from the victim’s role in elder abuse (i.e., their vulnerability and the difficulty in caring for them) to that of the abuser, transforming the conception of elder abuse from a problem similar to child abuse to a problem more closely resembling domestic violence (Wolf et al., 1998). As the domestic violence model had already embraced a criminal justice framework (Buzawa, Buzawa, & Stark, 2012), the inclusion of elder abuse as a family violence issue implicated a criminal justice response (Wolf et al., 1998).
Currently our only system of “intervention” for elder abuse perpetrators is prosecution (Wolf, 2003). Although law enforcement may accept reports of some forms of elder abuse, adult protective services (APS) is the designated entity in all states to receive and investigate reports of elder abuse (Jirik & Sanders, 2014) and therefore is the system entry point for most elder abuse cases. If referred, cases typically reach the prosecutor’s office either directly by referral from APS or indirectly by referral from APS to law enforcement. (The characterizations of cases that originate with law enforcement differ from those that originate with APS; Burgess, Ramsey-Klawsnik, & Gregorian, 2008; Lachs & Berman, 2011.) However, historically, elder abuse has not been considered criminal (Krienert, Walsh, & Turner, 2009) and therefore very few cases were referred for prosecution and even fewer were actually prosecuted, with most cases receiving no response (Jackson & Hafemeister, 2013b). While this is changing, prosecution remains uncommon (Ernst et al., 2013; Jackson & Hafemeister, 2013c). Meirson (2008) reported that Rhode Island APS received 900 complaints of elder maltreatment in 1 year, of which 85% were substantiated. However, only 70 cases were fully investigated by law enforcement, fewer than 20 of those resulted in the filing of criminal charges, and few, if any, of these cases resulted in conviction. In comparing prosecution when cases are funneled through an Elder Abuse Forensic Center compared with those that are not, Navarro, Gassoumis, and Wilber (2013) found that although the percentage of convictions did not differ significantly between the two settings (91.7% and 100%, respectively), the absolute number of convictions was higher in the Elder Abuse Forensic Center group (33 and 5, respectively) due to significantly more cases being initially referred to the prosecutor.
Elder abuse cases are notoriously difficult to prosecute dampening enthusiasm for accepting such cases (Klein, Tobin, Salomon, & Dubois, 2008), with some types of elder abuse more difficult to prosecute than others (Jackson & Hafemeister, 2010a). Jackson and Hafemeister (2013c) found that physical abuse cases were significantly more likely to be prosecuted than financial exploitation, caregiver neglect, or hybrid financial exploitation cases (i.e., financial exploitation co-occurring with physical abuse and/or neglect). However, Navarro et al. (2013) found that the odds of conviction were significantly greater when there was co-occurring financial exploitation and physical abuse than when either occurred in isolation.
Review of the Literature
Defining Elder Abuse Perpetrators
No elder abuse definition (Goergen & Beaulieu, 2013; Mysyuk, Westendorp, & Lindenberg, 2013a) or conceptualization (Harbison & Morrow, 1998; Policastro, Gainey, & Payne, 2013) enjoys a consensus in the literature. One point of contention is whether perpetrators must be in a “trust” relationship with the victim (Bonnie & Wallace, 2003) or whether a stranger can commit elder abuse (Goergen & Beaulieu, 2013). Most state statutes, for example, place no restrictions on perpetrator status (Jackson, 2015). In 1991, Hudson introduced an elder abuse taxonomy that made four distinctions: self-mistreatment, elder mistreatment by family/friend, elder mistreatment by professional relationship, and crime by a stranger. Goergen and Beaulieu (2013) recently dissected elder abuse definitions and concluded that strangers could be in “trust” relationships with victims under certain circumstances. This is consistent with Mysyuk et al.’s (2013a) analysis finding that the definition of elder abuse has expanded over time.
Characteristics of Both Victims and Perpetrators Predict Forms of Elder Abuse
As mentioned, historically the field of elder abuse has focused exclusively on victims. However, when both victim and perpetrator characteristics are entered into sophisticated analyses, characteristics of both victims and perpetrators contribute to the occurrence of abuse. For example, Jackson and Hafemeister (2011) found that characteristics and risk factors of both the victim and perpetrator predicted different types of elder abuse. More recently, Burnes, Rizzo, and Courtney (2014) concluded that contributions from both the perpetrator and the victim predicted outcomes across different types of abuse. Consistent with the victimology literature (Doerner & Lab, 2008), these studies suggest that characteristics associated with both victims and perpetrators play a critical role in the incidence of elder abuse, the implication being that we can no longer simply ignore the role of perpetrators.
The Importance of Distinguishing Among Types of Elder Abuse
The treatment of elder abuse as a monolithic phenomenon has plagued and damaged the field (Jackson & Hafemeister, 2013a). This section presents evidence for the importance of distinguishing among types of elder abuse.
Differences in prevalence across types of abuse
Elder abuse is a rubric under which six types of abuse fall. National prevalence studies find that although 11% of older adults are victims of elder abuse annually, prevalence rates differ by type of abuse (Acierno et al., 2010; Laumann, Leitsch, & Waite, 2008). For example, of 5,777 respondents, 4.6% reported emotional abuse, 1.6% reported physical abuse, 0.6% reported sexual abuse, 5.1% reported potential neglect, and 5.2% reported current financial abuse by a family member (Acierno et al., 2010).
Differences in sets of risk factors predict type of abuse
Very few studies have directly compared types of abuse across a range of risk factors. However, when risk factors are analyzed by type of abuse, studies find meaningful and significant differences across types of abuse using various samples. For example, Acierno et al. (2010) used a national prevalence study; Jackson and Hafemeister (2011) used state-level APS data; and Pittaway, Westhues, and Peressini (1995) used a large Canadian sample. Interestingly, Post et al. (2010) also found risk factors differed across types of abuse in a long-term care setting. These studies indicate that different forms of abuse are largely distinct (recognizing the existence of co-morbidity).
Differences in prevalence of perpetrator–victim relationship across types of abuse
In studies that assess family and non-family members as potential perpetrators, many studies find that family members (broadly defined) are the most common perpetrators of elder abuse (Holtfreter, Reisig, Mears, & Wolfe, 2014; Lachs & Berman, 2011; Naughton et al., 2012). However, the prevalence of perpetrator–victim relationship differs by type of abuse.
Neglect
Studies generally find that adult children are the most frequent perpetrators of neglect. Lachs and Berman’s (2011) self-report data indicated that neglect was committed most frequently by adult children (30%), followed by paid home care aids (28%) and spouses (19%). Lithwick, Beaulieu, Gravel, and Straka (1999) likewise found that adult children (49%) were more likely to commit neglect than spouses (23%). Using a definition that included an explicit duty to care, Acierno, Hernandez, Muzzy, and Steve (2009) reported that neglect was most often perpetrated by children/grandchildren (39%), followed by partners (28%), acquaintances (23%), and other relatives (7%). However, O’Keeffe et al. (2007) reported that neglect was committed mainly by partners (70%) and other family members (58%).
Physical abuse
Generally, partners/spouses are the most frequent perpetrators of physical abuse. O’Keeffe et al. (2007) reported that partners (57%) and other family members (37%) were the main perpetrators of interpersonal violence (physical, psychological, and sexual abuse combined). Similarly, Acierno et al. (2009) found that physical abuse was most often perpetrated by partners (57%), followed by acquaintances (19%), children/grandchildren (10%), other relatives (9%), and strangers (3%). Likewise, Lowenstein, Eisikovits, Band-Winterstein, and Enosh (2009) reported that physical/sexual abuse was committed primarily by spouses (1.8%), followed by family members (0.7%) and other caregivers (0.1%). Lachs and Berman’s (2011) self-report data found that spouses/partners were described as the most common perpetrators of physical/sexual abuse (36%), followed by other non-relatives (17%), adult children (13%), other relatives (12%), and friends (11%). Lithwick et al. (1999) also found that spouses committed more physical abuse (31%) compared with adult children (13%). The higher frequency of physical abuse by spouses may be the result of the inclusion (in several studies) of sexual abuse in this category. For example, Laumann et al. (2008) reported that physical abuse only was committed most frequently by an adult child (24.8%), followed by a spouse (19.6%) and other (55.6%).
Financial exploitation
The results for financial exploitation are more mixed. O’Keeffe et al. (2007) found that family members (54%) and care workers (31%), compared with partners (13%), were the more frequent perpetrators of financial exploitation. Likewise, Lachs and Berman’s (2011) self-report data found that the most common perpetrators of major financial exploitation were adult children (18%), followed by paid home care aides (16%), friends (14%), other relatives (14%), and grandchildren (12%). Lithwick et al. (1999) also found that adult children committed more financial exploitation (59%) compared with spouses (13%). Acierno et al. (2009) reported that 5.2% of older adults were financially exploited by a family member in the past year (6.5% were financially exploited by a stranger over their lifetime, although differences in timelines make this comparison problematic). In contrast, Lowenstein et al. (2009) found that caregivers (25.6%) were the most frequent financial exploitation perpetrators, followed by family members (3.9%) and spouses (0.5%). Laumann et al. (2008) found that financial exploitation was committed most frequently by other (56.4%) compared with adult children (34.0%) and spouses (9.6%). Beach, Schulz, Castle, and Rosen (2010) found that among African Americans, the majority of financial exploitation occurring in the past 6 months had been perpetrated by someone other than family or other trusted persons.
More methodologically rigorous research is painting an even more complex picture. In a randomly selected sample, Holtfreter et al. (2014) compared victim–perpetrator relationship and found that overall family members more frequently committed financial exploitation. However, when they compared just family members to strangers across three specific types of financial exploitation, they found that 42.9% and 25.7%, respectively, spent the victim’s money or sold something without permission; 50% and 30%, respectively, forged the victim’s signature; but that 29.6% and 39.2%, respectively, stole money or took things from the victim. Thus, for some types of financial exploitation, a family member was more likely to be the perpetrator, while for other types of financial exploitation, a non-family member was more likely to be the perpetrator.
Psychological/verbal abuse
There is general consensus across studies that the most frequent perpetrator of psychological abuse is a partner/spouse. Lithwick et al. (1999) found that spouses committed more psychological abuse (87%) compared with adult children (59%). Similarly, Lachs and Berman’s (2011) self-report study found that spouses were the most frequent perpetrators of psychological abuse (35%), followed by adult children (19%) and other relatives (15%). Acierno et al. (2009) reported that psychological abuse was most often perpetrated by partners (25%) and acquaintances (25%), followed by children/grandchildren (19%), other relatives (13%), and strangers (9%). Lowenstein et al. (2009) reported that 17.1% of verbal abuse was committed by spouses, 4.1% by family members, and 1.8% by other caregivers. Finally, Laumann et al. (2008) found that verbal abuse was committed most frequently by spouses (26.2%), followed by adult children (14.5%) and other (57.3%).
Sexual abuse
Too few studies exist to draw any firm conclusions regarding perpetrators of sexual abuse of older adults. Only two studies have separated sexual from physical abuse. Amstadter et al. (2011) reported that perpetrators of sexual abuse were less likely to be related to the victim (36.1%) compared with perpetrators of physical abuse (74.9%) and emotional abuse (63.2%). However, Acierno et al. (2009) reported that sexual abuse was most often perpetrated by partners/spouses (40%) and acquaintances (40%), followed by other relatives (12%) and strangers (3%; no children/grandchildren perpetrated sexual abuse).
Summary
This review demonstrates that victim–perpetrator relationship differs by the type of abuse involved. It appears as though family members are more likely to perpetrate elder abuse, with the possible exception of financial exploitation and sexual abuse. Even within the category of family, however, victim–perpetrator relationship differs by type of abuse. For example, partners/spouses are more likely to perpetrate psychological/verbal abuse and physical abuse, while adult children are more likely to perpetrate neglect.
Differences in perpetrator risk factors across types of abuse (where available)
Below, 19 perpetrator risk factors are reviewed based on a relatively small set of studies reporting perpetrator data. Where available, whether risk factors differ by the type of abuse involved is identified. Although risk factors are described below as if they are orthogonal variables, it is highly likely that many perpetrators exhibit more than one risk factor and that these risk factors are correlated to some degree. For example, it is likely that a perpetrator with a mental illness will find employment challenging (Stuart, 2006), contributing to financial dependence.
Age
The average age of an elder abuse perpetrator is 45 years (Brownell, Berman, & Salmone, 1999; Jackson & Hafemeister, 2011), with 75.1% of perpetrators under the age of 60 (O’Keeffe et al., 2007). There are no discernable differences in age across types of abuse.
Gender
Data that come to the attention of APS indicate a lack of gender differences among perpetrators of elder abuse (52.7% female; Kosberg, 2014; Teaster et al., 2006). However, Jackson and Hafemeister (2011) reported that 62% of the perpetrators were male, with no significant differences by type of abuse. Brownell et al. (1999) found that a slight majority (56%) of perpetrators were male. However, O’Keeffe et al. (2007) reported that 80% of interpersonal abusers were men, but that perpetrators of financial exploitation were both men (56%) and women (44%).
Minority status
In a sample of APS cases, 77% of perpetrators were Caucasian (Jackson & Hafemeister, 2011), while in a law enforcement sample, 48% of perpetrators were Caucasian (Brownell et al., 1999).
Education
Jackson and Hafemeister (2011) found that 38% of elder abuse perpetrators did not graduate from high school, although 44% held a high school diploma. Education differences were not observed across types of abuse.
Unemployment
Studies find that between one third and two thirds of perpetrators were unemployed at the time of the offense (Acierno et al., 2009; Brownell et al., 1999; Jackson & Hafemeister, 2011; Naughton et al., 2012). However, Jackson and Hafemeister (2011) found that perpetrators of physical abuse and hybrid financial exploitation were more likely to be chronically unemployed compared with perpetrators of financial exploitation and neglect. Acierno et al. (2009) reported that 40% of emotional, 33% of physical, and 22.9% of sexual abusers were unemployed.
Marital status and children
Jackson and Hafemeister (2011) reported that 70% of elder abuse perpetrators were unmarried at the time of the offense, with 30% having no children. Physical abuse and hybrid financial exploitation perpetrators were significantly more likely to have no children compared with neglect and financial exploitation perpetrators.
Health
The majority of elder abuse perpetrators appear to be relatively healthy (Clancy, McDaid, O’Neill, & O’Brien, 2011; Jackson & Hafemeister, 2011). However, Jackson and Hafemeister (2011) found that hybrid financial exploitation perpetrators were more likely to be characterized by poor health. Furthermore, they found that 85% of perpetrators were able to operate a motor vehicle (an indicator of health), although hybrid financial exploitation perpetrators were significantly less likely to be able to do so.
Childhood family violence
Jackson and Hafemeister (2011) reported that almost half (44%) of the elder abuse perpetrators had a history of childhood family violence (witnessing or experiencing), although childhood family violence did not differ by type of abuse.
Substance abuse
Rates of substance abuse range from 20% to 50% among elder abuse perpetrators (Amstadter et al., 2011; Brownell et al., 1999; Clancy et al., 2011; Jackson & Hafemeister, 2011; Lowenstein et al., 2009; Naughton et al., 2012). O’Keeffe et al. (2007) reported that 4% of all perpetrators had alcohol problems, although 30% of financial exploitation perpetrators compared with 5% of interpersonal perpetrators had alcohol problems. Likewise, 1% of all perpetrators had drug problems, although 8% of financial exploitation perpetrators compared with less than 1% of interpersonal perpetrators had drug problems. Acierno et al. (2009) reported that substance abuse was a problem for 50% of physical, 28.2% of sexual, and 21% of emotional perpetrators.
Mental illness
Approximately 25% to 35% of elder abuse perpetrators have a serious mental illness (Amstadter et al., 2011; Brownell et al., 1999; Clancy et al., 2011; Jackson & Hafemeister, 2011; Lowenstein et al., 2009). However, O’Keeffe et al. (2007) found that 7% of perpetrators had mental health problems, although less than 1% of financial exploitation perpetrators compared with 8% of interpersonal perpetrators had mental health problems. Acierno et al. (2009) reported that perpetrators who had received mental health treatment included 33% of physical, 20% of emotional, and 19.6% of sexual perpetrators.
Gambling problems
O’Keeffe et al. (2007) reported that 1% of perpetrators had gambling problems, although 23% of financial exploitation perpetrators, compared with less than 1% of interpersonal perpetrators, had gambling problems.
Criminal record
Between 25% (Jackson & Hafemeister, 2011) and 46% (Brownell et al., 1999) of elder abuse perpetrators had a criminal record at the time of the offense. No differences were observed by type of maltreatment.
Panviolent perpetrators
Jackson and Hafemeister (2011) reported that 44% of the elder abuse perpetrators had engaged in a similar act of abuse or violence toward someone other than the older victim. However, panviolence did not differ by type of abuse.
Relationship (partner) problems
Jackson and Hafemeister (2011) found that 68% of perpetrators experienced relationship problems, while O’Keeffe et al. (2007) reported only 21% of perpetrators had relationship problems. However, Jackson and Hafemeister (2011) found a trend indicating that financial exploitation perpetrators were less likely to commit intimate partner violence compared with perpetrators of three other types of abuse.
Living arrangements
Studies find that between 53% and 64% of elder abuse victims and perpetrators were cohabitating at the time of the offense (Brownell et al., 1999; Jackson & Hafemeister, 2011; O’Keeffe et al., 2007). However, Jackson and Hafemeister (2011) reported that physical abuse and hybrid financial exploitation perpetrators were more likely to be living with the victim compared with financial exploitation and neglect perpetrators. O’Keeffe et al. (2007) likewise found that fewer financial exploitation perpetrators (25%) lived with the victim compared with interpersonal perpetrators (65%).
Financial problems
O’Keeffe et al. (2007) found that 1% of their sample was experiencing financial problems at the time of the offense. However, 30% of financial exploitation perpetrators compared with less than 1% of interpersonal perpetrators had financial problems.
Financial dependence
Jackson and Hafemeister (2011) found that approximately one third of perpetrators were financially dependent upon the victim at the time of the offense. However, hybrid financial exploitation perpetrators were significantly more likely to be financially dependent compared with perpetrators of three other forms of abuse. Relatedly, compared with financial exploitation or neglect perpetrators, physical abuse and hybrid financial exploitation perpetrators were characterized as parasitic, with hybrid financial exploitation perpetrators potentially more symptomatic (Jackson & Hafemeister, 2012a).
Social isolation
Jackson and Hafemeister (2011) found that one third of elder abuse perpetrators (35%) did not have someone in their life they could count on to take them to the doctor or to call if they needed to talk to someone, although social isolation did not differ by type of maltreatment. Acierno et al. (2009) reported that social isolation (less than three friends) was a problem for 53.1% of sexual, 44% of physical, and 40% of emotional perpetrators. Importantly, isolation for some perpetrator–victim dyads results from a combination of frightening behavior exhibited by the perpetrator toward family and friends, and the victim/parent’s fierce protection of the perpetrator that alienates friends and family, leaving the dyad isolated (Jackson & Hafemeister, 2010b; Stewart, Burns, & Leonard, 2007).
Perpetrator is a care recipient
O’Keeffe et al. (2007) found that 35% of perpetrators received care from the victim, although only 25% of financial exploitation abusers compared with 39% of interpersonal abusers received care from the victim.
Summary
Across the 19 perpetrator risk factors reviewed, 11 varied by type of abuse. However, the studies differed in their methodological rigor, suggesting the need for replication.
Variations in Perpetrator Culpability
Just as the field of elder abuse has tended to lump all forms of elder abuse together (Jackson & Hafemeister, 2013a), so too have we tended to lump all elder abuse perpetrators together and label them the “bad guys.” However, there is much greater heterogeneity among perpetrators than such a label implies. To capture perpetrator heterogeneity, Jackson and Hafemeister (2010b) attempted to develop a continuum of perpetrator culpability based on an analysis of the interpersonal dynamics involved in 71 cases of elder abuse. Building upon Hafemeister’s (2003) earlier work, the continuum ranges from bad actors (pre-meditated, knowing, deliberate), to perpetrators who readily exploit unexpected opportunities (no excuse, but unplanned, sense of desperation), to those who reluctantly exploit unexpected opportunities (caregiver stress, mixed motives), to those individuals who act in a manner, although inappropriate, at least somewhat consistent with the older adult’s wishes, to those exhibiting ignorance (perpetrators who legitimately do not understand why their actions constitute abuse/neglect or are incapable of fulfilling expected roles).
Relatedly, Ramsey-Klawsnik (2000) provided a description of five typologies based on her experience conducting clinical forensic investigations involving family violence. First, overwhelmed perpetrators are primarily qualified to provide care but at some point become overwhelmed and engage in some type of maltreatment toward the older adult. Second, impaired perpetrators are willing to provide care for older adults but because of some type of impairment (intellectual deficit, mental illness) are not qualified to provide care. Third, narcissistic perpetrators are motivated by personal gain rather than to help the older adult. Fourth, domineering, or bullying, perpetrators feel justified in their maltreatment and tend to focus their maltreatment on family members. Finally, sadistic perpetrators derive feelings of power and importance by humiliating, terrifying, and harming others. While these are both intriguing lines of research, further development is required and analysis of differences by type of abuse is needed.
The Continuum of Complexity of Elder Abuse Cases
It important to consider that elder abuse cases in general vary in their level of complexity. Some cases are readily resolved while others require greater commitment of resources to resolve (Goergen & Beaulieu, 2010; Jackson & Hafemeister, 2010b, under review; Kingston & Reay, 1996). This complexity is heightened where victim–perpetrator relationships are familial (Maccoby, 2000), long term (Harrell, Visher, Newmark, & Yahner, 2009), and when perpetrators are committing co-occurring forms of elder abuse (Jackson & Hafemeister, 2012b). There is always a small set of cases that defy intervention (Burnes et al., 2014; Comijs, Pot, Smit, Bouter, & Jonker, 1998; Jackson & Hafemeister, 2013b). This may be due in part to a minority of clients declining intervention, with significantly more physical abuse victims declining intervention compared with other forms of abuse (Jackson & Hafemeister, 2012b).
Elder Abuse Perpetrator Theory
Scholars have admonished the field for failing to develop elder abuse theories generally (Bonnie & Wallace, 2003; Lowenstein, 2009) and for distinct types of abuse specifically (Jackson & Hafemeister, 2013a). As reviewed, risk factors differ across types of abuse; therefore, theories must differ across types of abuse as well (Payne, 2011). Theory, whether empirical or experiential, should form the foundation for intervention (Rossi, Lipsey, & Freeman, 2004). However, different theories implicate different interventions (Jackson & Hafemeister, 2013a; Mysyuk, Westendorp, & Lindenberg, 2013b; Payne, 2011). For example, if you believe that the underlying cause of elder abuse is patriarchy, the intervention might include correcting misperceptions through intensive education (Gondolf, 2011). Alternatively, if you believe elder abuse perpetrators have unresolved psychological problems, then the intervention might include addressing those unresolved problems through psychotherapy (Dutton, 2006). If, however, you believe elder abuse perpetrators are deviant, then the criminal justice system is the intervention of choice (Pillemer, 2005). Unfortunately, little work exists on the underlying causes of elder abuse (let alone by type of abuse), which would inform theory development (Jackson & Hafemeister, 2013d). Relatedly, no research has sought to capture perpetrator’s underlying motivations which would also inform theory. (Hamberger, Lohr, Bonge, & Tolin, 1997, provide a useful example of this in the context of domestic violence.) What is clear is that perpetrator’s motivations for acting out are far more complex than is generally recognized (Goergen & Beaulieu, 2010; Grabosky & Walkley, 2007).
Implications for Intervention
Current Intervention
The current response to elder abuse is either no response or a criminal justice response (Jackson & Hafemeister, 2012c; Wolf, 2003). The historical denial of elder abuse as a crime has fueled the current enthusiasm for prosecution. Calls for increasing rates of prosecution abound (Bailly, Loewy, Bomba, & Lynch, 2007), although some scholars argue that deterrence as a theory has minimal effect on behavior (e.g., Dooley & Radke, 2010; Robinson & Darley, 2004). In addition, not all victimization experiences are criminal (Krienert et al., 2009; Policastro et al., 2013). Furthermore, prosecution is pursued in so few cases that such an approach will not affect the vast majority of perpetrators. Even if prosecution is pursued, many perpetrators return to their victim upon release (Jackson & Hafemeister, 2012c). Finally, such an approach has been criticized for disempowering victims (Kohn, 2012). Therefore, it is imperative to address the needs of abusive individuals in ways other than, or in addition to, the criminal justice system. There is a general consensus that although prosecution is important, as a society we cannot prosecute our way out of elder abuse. Criminal justice interventions (arrest, special units, prosecution, courts, batterer intervention programs, specialized supervision) are not effective at reducing recidivism in the context of intimate partner violence, and Peterson (2008) wrote that it is “time to correct the imbalance” of our overreliance on the criminal justice system (p. 542; see also Buzawa et al., 2012). The same may be said of elder abuse.
New Directions for Intervention
As perpetrators have historically been excluded from the elder abuse dialogue, little is known about them. However, this review identified four potentially important factors in the development of interventions. First, studies find that various characteristics of both the victims and perpetrators are associated with different forms of elder abuse (Burnes et al., 2014; Jackson & Hafemeister, 2011). This suggests that a response focusing exclusively on victims is a misguided policy. Perpetrators are frequently embedded in the victim’s social network (Schafer & Koltai, 2014), suggesting that elder abuse often involves complex interpersonal relationship patterns that need to be parsed out (Brownell et al., 1999). Treatment must be focused on both parties, as well as on relationship issues and problem-solving skills of both parties to increase the probability that abuse will cease (Baird, 2005; Daniel & Bowes, 2011; Henderson et al., 2004; O’Donnell, Treacy, Fealy, Lyons, & Lafferty, 2014). The United Kingdom, for example, is focused on providing necessary services to end abuse and restore relationships, rather than the punishment of perpetrators (Filinson, 2008). Thus, the historical singular focus on victims must be abandoned in favor of recognizing the importance of victim–perpetrator relationships that exists in many cases.
Second, there are significant and meaningful differences across types of abuse in terms of prevalence, sets of risk factors, victim–perpetrator relationship, and perpetrator risk factors. This suggests that the field must cease conceptualizing elder abuse as unitary phenomenon. Recognizing that types of abuse co-occur (Jackson & Hafemeister, 2012a), and at times one type leads to another (Dong, Simon, & Evans, 2012), conceptualizing elder abuse as being comprised of distinct types implicates the need for distinct interventions (Jackson & Hafemeister, 2011; Lithwick et al., 1999; Payne, 2011). Interventions for financial exploitation will surely differ from those cases involving caregiver neglect. It is predictable that a one-size-fits-all intervention for elder abuse perpetrators is doomed to failure. Thus, the monolithic characterization of elder abuse must be abandoned in favor of differentiating among types of abuse.
Third, there is preliminary evidence for the existence of meaningful variations in culpability among elder abuse perpetrators (Jackson & Hafemeister, 2010b; Ramsey-Klawsnik, 2000). This suggests that the field must eschew the tendency to treat perpetrators as “all bad,” a by-product of our “system” that cannot accommodate gradations of badness (Daniel & Bowes, 2011), regardless of the cause (e.g., Hafemeister, Garner, & Bath, 2012). Goergen and Beaulieu (2010) conceptualized elder abuse as lying on two continua, intention or lack of intention to harm and situation-specific or trans-situational. The level of intervention will differ depending on which quadrant the case falls. For example, when there is high intent to harm and the abusive behavior is trans-situational (e.g., the pattern of exploitation is particularly egregious and the perpetrator appears sociopathic and predatory), the level of intervention must be relatively high (e.g., counseling and court intervention). When the behavior lacks intention to harm and the abuse is limited to a particular situation (e.g., a perpetrator struggling to care for an older parent while trying to manage his own infirmities), a lower level of intervention is required (e.g., a psychoeducational response or in-home support for the care provider). Similarly, Kingston and Reay (1996) presented a schematic which suggests that in aggressive cases, police involvement is necessary; in abusive situations, counseling is necessary; and in passive situations, a social services approach may be necessary. These approaches may be accomplished by triaging perpetrators as low-, moderate-, and high-risk (Cavanaugh & Gelles, 2005; Kelly & Johnson, 2008) via risk assessment instruments similar to those used in other disciplines (Skeem & Monahan, 2011). High-risk perpetrators benefit from cognitive-behavioral programs (Landenberger & Lipsey, 2005), but may require court oversight and enhanced supervision (Gondolf, 2011; Visher, Newmark, & Harrell, 2007). However, Klein and Crowe (2008) found that safety was enhanced for low-risk domestic violence perpetrators experiencing a specialized supervision program (not a treatment program), but not for the high-risk perpetrators. Thus, the portrayal of elder abuse perpetrators as “all bad” must be abandoned in favor of adopting an acceptance of the gradations of perpetrators’ culpability.
Finally, there is some evidence that elder abuse cases exist on a continuum of complexity (Goergen & Beaulieu, 2010; Jackson & Hafemeister, 2010b, under reviewKingston & Reay, 1996). This suggests the need to adjust our expectation that APS can just step in and readily stop abuse. Some cases are easily resolved, while others may require long-term case management, and yet others are not amenable to intervention, as there is always a small percentage of cases in which the abuse continues after an APS investigation (Burnes et al., 2014; Comijs et al., 1998; Jackson & Hafemeister, 2013b). Therefore, some interventions must take a long-term perspective given potential resistance and that the abused and abusers fail to follow through with interventions without the continual assistance and support of the therapist (Booth, Bruno, & Marin, 1996; Breckmman & Adelman, 1988). Thus, the belief that all elder abuse cases are equally resolvable must be abandoned in favor of accepting the reality that the complexity of elder abuse cases exists on a continuum requiring variations in resources.
The implication of these findings is that the utilization of criminal justice and/or therapeutic interventions must be tailored to the specific needs of the case, recognizing the heterogeneity of elder abuse perpetrators and the circumstances in which they are involved. Optimism that perpetrator intervention can become part of a response system is warranted. An evaluation of an elder abuse intervention found that pre-intervention, social workers focused on treating victims, whereas post-intervention, services for perpetrators were provided (Alon & Berg-Warman, 2013).
Lessons Learned From the Intimate Partner Violence Field
In the context of intimate partner violence, Kelly and Johnson (2008) wrote that the “. . . long-term adherence to the conviction that all domestic violence is battering has hindered the development of more sophisticated assessment protocols and treatment programs . . .” (p. 477). Indeed, the overwhelming weight of the evidence finds heterogeneity among intimate partner violence perpetrators (Chiffriller, Hennessy, & Zapponea, 2006; Dutton, 2006; Holtzworth-Munroe, Meehan, Herron, Rehman, & Stuart, 2000; Holtzworth-Munroe & Stuart, 1994; Kelly & Johnson, 2008; Straus, 2011), a fact that has never gained traction in the face of staunch opposition (Baird, 2005; Dutton, 2006). Furthermore, some scholars in the field of intimate partner violence have been vociferous with regard to the importance of matching batterer typologies to treatment (Carlson, 2005; Cavanaugh & Gelles, 2005; Holtzworth-Munroe & Stuart, 1994; Jackson, 2003), again with little traction. The intimate partner violence community has continued to promote arrest and prosecution of perpetrators, with batterer intervention programs the dominant intervention (Gondolf, 2011), in spite of little support demonstrating that batterer intervention programs are effective over time at decreasing acts of violence in intimate relationships (Babcock, Green, & Robie, 2004; Family Violence Prevention Fund, 2010; Hafemeister, 2011; Jackson, 2003; Kelly & Johnson, 2008; ). Cantos and O’Leary (2014) recently implored for “. . . those responsible to stop perpetuating the practice of mandating all perpetrators to attend a single intervention for which there is very limited evidence of effectiveness” (p. 204). Little research exists on matching typology to treatment (Gastfriend & McLellan, 1997), but treatment outcomes (recidivism, program completion) have differed depending on the type of abuser (Kelly & Johnson, 2008), suggesting that targeting programs for different types of perpetrators might be productive, while recognizing that different perpetrators may respond differently to the same intervention (Huss & Ralston, 2008).
Conclusion
This review identified four factors that may be impeding our ability to effectively in intervene in elder abuse cases, culminating in a recognition that perpetrators are considerably more heterogeneous that our current response system indicates. Reliance on prosecution as the sole intervention for elder abuse perpetrators fails to acknowledge this heterogeneity and will never reach all the perpetrators in need of intervention. Our understanding of elder abuse perpetrators is in the nascent stage and our conceptualization of elder abuse perpetrators may change over time with the accumulation of new knowledge. However, interventions that utilize criminal justice and/or therapeutic approaches tailored to the needs of the case currently hold the most promise. Interventions must incorporate the heterogeneity of both perpetrators and victims, be specific to the type of abuse involved, and take into consideration the level of perpetrator culpability and the continuum of complexity among these cases. Without addressing these nuances, intervention and prevention efforts will be futile if not harmful.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
