Concerns with the Explanation of Crime
With Crime Definition 1.0 explained and defended, we now return our attention to an issue raised earlier in this work: the assessment of explanations of crime. We outline what we see as major problems in criminology that are a result of when criminologists simply accept the criminal-law definition of crime. First, as we have argued, criminologists tend to assess hypotheses in ways where they are not likely to be rejected, and instead of rejecting weak results, they argue for retaining an explanation due to measurement issues. Thus nearly all explanations of crime persist over time. Second, most empirical tests do not assess criminal behavior but rather the behavior of lawmakers. Third, the convenience of crime data promotes the reification of the concept of crime—that is, crime statistics are studied as if they were objective measures of crime.
Hypotheses Are Rarely Rejected
First, as we have noted, criminologists tend to employ very liberal empirical criteria when assessing the results of tests of hypotheses about the causes of crime. This leads to the increased likelihood of accepting associations and explanations about the causes of crime that ought to be rejected. That problem is exacerbated by the definition of crime that criminologists traditionally employ. Lacking strict empirical criteria and armed with a definition of crime that has been generated by lawmakers influenced by political structures, there are substantive reasons to expect that the explanation of crime produced from such procedures will be both inefficient and ineffective.
As we noted earlier, one of our concerns is that when criminologists test “theories” of crime and the empirical results fail to support the hypothesized relationship, it is a common practice to suggest reasons that the hypothesis test may have failed. The assumption being made here is that the explanation itself seems plausible enough, and that the thing that went wrong has something to do with the empirical test—that the independent variables used were not a good measure of the concept and that it is the failure to adequately measure the independent variable that produces the weak results. The failure of the analysis does not normally promote the criminologist to question the “theory” but rather directs them to attend to why the empirical test may have been inadequate. To be sure, scientists ought to be clear about the limitations of their research, understand those limits, and question whether the measures they propose and use are valid indicators of a theoretical concept. This is surely one of the appropriate responses that researchers should consider. At the same time, however, they should also be clear about the limits of their conceptual arguments and not be willing to ignore those limits in favor of accepting the explanation as theoretically valid but empirically flawed—that is, when criminologists explain the failure of empirical tests, they often point to potential problems in the operationalization of the independent variables. The criminologist rarely suggests that the measurement issues relate to the nature of the dependent variable, crime. A criminologist does not argue that the concept they are trying to explain—crime—i s not what they are measuring when they use the legal definition of crime. They do not imagine that the explanation of crime they have created is based on their concept of crime and that the concept of crime they used for that purpose is not adequately captured by the legal definition of crime. Of course, the problem of crime is much larger than the question of whether the criminologist’s concept of crime and the legal definition of crime coincide. The problem also has to do with the fact that law itself is a behavior and that the criminologist rarely considers this concern when testing explanations of crime. If, as we have suggested, the legal definition of crime is variable and changes across time and place, then the dependent variable includes variability that the criminologists has not considered explaining. The failure to control for the factors that change the criminal law means that important factors that change the measure of crime have been omitted from previous studies and that the results of those studies have been misspecified. This means that we should be cautious with how we interpret the criminological literature and the results of studies that test explanations of crime.
As others have argued, the law is a social and political construction. With respect to analyzing the causes of crime, this observation has extraordinary importance and is an issue that criminologists tend to ignore. The issue here is that if the definition of crime is a social and political construction, it has content that is created by lawmakers that represents their perception of crime and not necessarily the nature of crime. We have discussed this issue previously. To recap that argument, the fact that the law is a social and political construction will also mean that the definition of crime in the criminal law will vary across time and place. That criminal-law definitions of crime are not consistent across time and place and hence across studies of criminal behavior is an important observation. It means that the definition of crime is not fixed and uniform. The criminal law, as the measure of the behavior of lawmakers, varies. Thus in order to be able to use the criminal-law definition of crime, it is first necessary to be able to address and control for the factors that produce the variability in the legal definition of crime.
As we have noted, to have a scientifically founded discipline, it is necessary to have concepts that are well-defined and that do not change—that is, concepts that are objective and scientific. The traditional criminal-law definition of crime fails on this account. In an abstract sense, the criminal-law definition of crime that criminologists use does not change. What does change, however, is criminal law itself, and it changes for reasons that criminologists have not controlled. Part of controlling that definition can include two different responses. One is the creation of a new definition of crime, a point we addressed in the prior chapter. A second approach, discussed here, is to include variables that are capable of measuring how law changes.