3. to call for the generation of testable hypotheses based on various components of the model;
4. to provide a tentative guide for risk reduction strategies; and
5. to consider the specific risk factors and mechanisms that affect people with disabilities in order to further understanding of violence and violence prevention in general.
Significant studies of crime victims began in the middle of the 20th century. Like much of the work that followed, early studies in victimology emphasized the importance of considering the relationship between the perpetrator and the victim. Von Hentig (1948) is credited with first identifying the relationship between disability and victimization when he suggested that four categories of people are particularly vulnerable to victimization: the young, the old, females, and the mentally disabled. However, the relationship between disabilities and crime victimization received little attention until the 1960s when studies found high rates of developmental, physical, and behavioral disabilities among abused children (see, e.g., Birrell and Birrell, 1968; Elmer and Gregg, 1967; Gil, 1970; Johnson and Morse, 1968). Studies that followed also revealed higher than expected rates of substantiated child abuse among children with disabilities (e.g., Buchanan and Oliver, 1977; Frisch and Rhoads, 1982). These studies demonstrated a relationship between abuse and disability, but they shed little light on why the relationship might exist or whether disability was an outcome of or a risk factor for abuse.
The traditional explanation for seemingly high rates of child abuse and other forms of violence against people with disabilities is referred to as the dependency-stress model. Used extensively between the 1960s and the 1990s, this model is fashioned on several premises: children with disabilities are more dependent on their caregivers; increased dependency increases the demands on caregivers; increased demands result in increased stress for caregivers; and caregivers abuse their charges because they cannot cope with the increased stress. Although this model appears to be logical, little research supports it, and some research seems to contradict it altogether (e.g., Benedict et al., 1992; Pillemer and Finkelhor, 1989; Starr et al., 1984). Furthermore, the model can be construed as excusing offenders or even transforming them into victims while blaming the real victims for causing stress.