different levels of risk for mistreatment. For example, it may be the case that different organizational settings (e.g., nursing homes, assisted living quarters) may have different characteristic levels of elder mistreatment risk. This feature of the model is the macrostructure in which the microprocesses described in Figure 3-2 occur. The risk of elder mistreatment can be conceptualized as the varying likelihood of an event or set of events causing harm to the elder person. This risk is a function of the various sets of variables depicted in the model at both the macro and micro levels.
The left side of the diagram includes the set of social, physical, and psychological attributes of the subject at risk of elder mistreatment, and the right side lists the pertinent attributes of the trusted other. The middle set of boxes represents the interaction of the two sets of individual-level variables that define the level of social or economic dependence (status inequality), type of social relationship in which the interaction between the elder person and the trusted other happens, with corresponding differences in the normative expectations held by different stakeholders and the power dynamics in negotiating the operative care-giving scripts (see Simon and Gagnon, 1987; Mahay et al., 2001). Note that we have also included “social embeddedness,” which refers to the sets of people in the social networks of the elder person and the trusted other, respectively, constituting the social capital available in the dyadic transaction (see Sandefur and Laumann 1998). These two networks may overlap or not, with attendant consequences for their efficacy in exerting social control over the dyadic interaction of focal interest. Social networks can serve critical functions of monitoring the situation and informing relevant others when shortfalls or problems arise. Their presence may also serve as a form of social control on the behavior of the focal parties. Their absence greatly enhances the vulnerability of the elder person and the trusted other to the risk of elder mistreatment (see House et al., 1988; Lin et al., 1999). Finally, outcomes include the physical and emotional health and happiness of the elder person and the trusted other, the differential risks of elder mistreatment in its varied forms, and the durability (or risk of termination) of the caregiving relationship itself.
We should expect that all these outcomes have feedback effects on the variables above them—that is, the paths connecting outcomes to the boxes listing the independent variables are double-headed rather than unidirectional. For example, we might expect that the occurrence of an incident of elder mistreatment increases the odds of additional events of elder mistreatment, as it adversely affects physical, psychological, and social statuses for both the elder person and the trusted other. A mistreated elder is more likely to respond with depression, physical disability, or social withdrawal as a direct or indirect reaction to the mistreatment—each of which may enhance the likelihood of another incident. Similarly, the perpetrator may