- ♦ Document the need for change. Collect and analyze data to define what the problem is, where it is, how big it is, and who is affected by it. What evidence of the problem exists?
- ♦ Describe the history of the problem. How long has the problem existed? How has it changed over time?
- ♦ Examine potential causes of the problem. What causes the problem? What theories do we have? The intervention must target one or more specific causes supported by research.
- ♦ Examine previous interventions that have tried to change this problem, identify the most promising interventions, and choose a preferred intervention approach. We need to analyze available information to direct decisions about a possible course of action.
- ♦ Identify relevant stakeholders. Do different groups of people have different definitions of the problem? Who is affected by the problem?
- ♦ Conduct a systems analysis. Conduct research on the system within which the problem exists, and determine how the system may create, contribute to, or maintain the problem.
- ♦ Identify barriers to change and supports for change. Who is likely to support a certain course of action, and who is likely to resist it?
Some preliminary analysis is needed to identify the issues involved with trying to change a particular problem. This important analysis sets the stage for all subsequent planning activities. Beware of the activist bias:
the notion that we already know what to do, so let’s get on with it. In almost all cases, the person who expresses such a view has a vague definition of the problem and its causes, and little knowledge of successful interventions. Without intending it, he or she is advocating a
process of unplanned change that maximizes the likelihood of a poorly planned, poorly implemented, and ineffective intervention. The many hours of hard work and the motivation that must surely guide any successful change effort should not be wasted on unplanned change. How we analyze the problem guides what kind of interventions we come up with. If problem analysis is flawed, subsequent program or policy planning is also likely to be faulty.
We begin analysis of a problem by examining information about the problem. We are interested in questions like the following: How do we define the problem? How big is it, and where is it? Is there a potential for change? We especially want to provide evidence for the existence of a need or problem.
We need to be very careful here. The media, politicians, or even criminal justice officials socially construct many problems. By social constructions, we mean that certain problems are perceived, and decisions are made to focus attention and resources on a particular problem (Spector & Kitsuse, 2001; Walker, 2005). However, perceptions of a problem and reactions to it may be quite different than the actual size or distribution of a problem. We need methods to document, describe, and analyze problems. At minimum, we need to be sure that a problem actually exists before taking any specific action, but we also need to know about the size and distribution of a problem in order to plan effective solutions.
Although the distinction is somewhat arbitrary, it is often worthwhile to differentiate a need from a problem. Students often point out that many “conditions” could be stated either way: for example, if victims of domestic violence lack access to shelters, then is there not only a need but also a problem, such as repeat incidents of abuse of this population? However arbitrary the distinction might appear at first glance, it might make a large difference in the problem analysis (what kind of information we collect), analysis of causes (explanations of why certain conditions are lacking, versus why other conditions are present), and identification of relevant interventions (do we attempt to provide services that fill an important gap, or do we attempt to apply some intervention to change a problem?). Needs and problems are clearly related, but not identical.
A lack of something that contributes to the discomfort or suffering of a particular group of people. For example, we might argue that there is a need for drug treatment programs for convicted offenders, or that there is a need for shelters for abused women. In each case, an existing lack of services perpetuates the difficulties experienced by the target population.
Example 1.1 School Violence: A Problem Out of Control?
Shootings in and around schools have fueled a national debate about school violence. Following tragic incidents such as the Columbine High School massacre in 1999, the Virginia Tech shootings in 2007, and other widely publicized shootings on school properties, many school districts and campuses have scrambled to improve their security measures and disciplinary policies. While dramatic incidents fuel perceptions that school violence is out of control, available data suggest a more modest interpretation. One primary source of national data about school crime and safety is the annual report, Indicators of School Crime and Safety (Robers et al., 2014). Using indicators from various sources, results showed that:
- Between July 1, 2010, and June 30, 2011, there were 11 homicides and three suicides of school-age youth (ages 5–18) at school. A “school-associated violent death” is defined as “a homicide, suicide, or legal intervention (involving a law enforcement officer), in which the fatal injury occurred on the campus of a functioning elementary or secondary school in the United States, while the victim was on the way to or from regular sessions at school or while the victim was attending or traveling to or from an official school-sponsored event.” (Indicator I).
- In 2012, there were no measurable differences in rates of serious violent victimization (rape, sexual assault, robbery, or aggravated assault) against students ages 12–18 at school (3.0 per 1,000) compared to those occurring away from school (2.9 per 1,000). (Indicator 2).
- In 2011, 4 percent of students ages 12–18 reported being victimized at school during the previous six months in 2011. Three percent of students reported theft, 1 percent reported violent victimization, and one-tenth of 1 percent reported serious violent victimization (Indicator 3).
Fortunately, tragedies such as Columbine are rare. Based upon current evidence, it is neither clear that school violence is out of control, nor that revision of school security policies is the proper (or only) solution. Thorough, localized problem analysis should precede the revision or development of school policies in any district.
Next, we attempt to apply some boundaries
to the problem. For example, we might begin by stating a concern with juvenile violence. However, we are quickly overwhelmed with information about the problem, different causal explanations, and different interventions (Riedel & Welsh, 2015). Are we really concerned with all types of juvenile violence, or with more specific settings? Are we really interested in specific types of violence, such as gang violence, school violence, gun-related violence, drug-related violence, interpersonal conflicts versus violence committed against strangers, or instrumental (goal-oriented) versus affective (emotional) violence? This is an important point. We need to do some research first to narrow our definition of the problem. It is entirely possible that we might decide to focus not only on a specific type of violence, but upon a specific age group (say,
middle-school children), a specific jurisdiction (e.g., a community with a high rate of violence, or a specific city, county, or state), or a particular demographic group (e.g., poor children living in inner-city areas). Whatever our reasons for choosing to set boundaries in particular ways (personal, political, or theoretical interests), identifying boundaries involves making judgments about how widely or narrowly to define a problem.
The presence of something that contributes to the discomfort or suffering of a specific group. For example, we might argue that a specific community experiences a high rate of robberies committed by addicts to buy drugs, or that there is a high rate of repeat incidents of abused women applying to courts for protection orders. In each case, there is a clearly defined condition present that perpetuates the suffering of a particular group of people.
The number of new cases of a problem within a specific time period (e.g., the number of new cases of AIDS diagnosed in a specific calendar year). According to the Joint United Nations Programme on HIV/AIDS, an estimated 2.1 million people worldwide acquired human immunodeficiency virus (HIV) in 2013, down from 2.9 million in 2005 (Joint United Nations Programme on HIV/AIDS [UNAIDS] and World Health Organization [WHO], 2015).
We first attempt to document the need for change through an analysis of existing conditions. Is there a problem? How big is it? What is the level of “need”? What is the evidence for a problem? One way of documenting a problem is to look at its incidence versus prevalence.
The existing number of cases of a particular problem as of a specific date (e.g., the total number of people with AIDS as of a specific date). As of December 31, 2013, 35 million people worldwide were estimated to be living with HIV/AIDS, up from 32.1 million in 2005 (Joint United Nations Programme on HIV/AIDS [UNAIDS] and World Health Organization [WHO], 2015).
Where do we find this kind of statistical information, as well as more descriptive information about the problem? We usually need to look at some kind of data to estimate the degree and seriousness of a problem. There are several techniques available; we’ll briefly review four of them for now (Figure 1.1
). Wherever time and resources allow, it is always desirable to use as many techniques as possible to converge upon a specific problem.
Social indicators are perhaps the most accessible and widely used type of data for analyzing criminal justice problems (Figure 1.2
). For example, the Uniform Crime Reports (UCR), collected by the FBI, consist of all crimes reported to the police, and all police arrests for specific crimes. Data are available for each state and for the nation as a whole. These figures are widely used to calculate changes in the homicide rate, for example, from year to year. Another widely used indicator is the National Crime Victimization Survey (NCVS), which is a survey administered to a national probability sample, asking respondents to report whether they have been a victim of specific crimes within a specific time period (e.g., the previous six months), as well as other information about any victimization, such as degree of injury suffered and characteristics of the offender (if known).
FIGURE 1.1 Data-Collection Techniques: Documenting the Need for Change
By examining social indicators, we can define a problem and attempt to change it. For example, as part of the Community Corrections Program Development Project funded by the National Institute of Corrections, Temple University researchers collaborated with the Los Angeles and Orange County Probation Departments.
They found that a high-risk group of only 8 percent of juveniles account for the great majority of repeat referrals to juvenile court. Consequently, many agencies— probation, health, mental health, social services, school districts—cooperatively developed programming for an “8 percent solution” specifically targeting this high-risk group (Kurz & Schumacher, 2000).
FIGURE 1.2 Examples of Social Indicators for Criminal Justice Problems
Social indicators are extremely useful for identifying the seriousness of a problem, how it varies across groups (e.g., high vs. low income), and how it is changing over time (is it getting better or worse?). Such data are not without biases, however, and the potential user needs to be aware of these (Biderman & Lynch, 1991; Riedel & Welsh, 2015). For example, crime victimization measures may be biased by numerous factors (e.g., respondent misunderstanding of questions or crime definitions; faulty recall of incidents and time periods; deliberate underreporting due to fear, embarrassment, or the respondent’s participation in illegal activities). Police-reported crime rates such as the UCR also carry potential biases, including police errors in recording and coding crime incidents. Many crimes are never even reported to the police for various reasons (e.g., v...