A Relative Definition of Crime
We begin this discussion with an observation about the utility of relative definitions of crime; they are useful to the extent that they can address variations in crime related to the influence of structural, cultural, political, and historical forces and can address issues related to the variability in the definition of crime that the legal definition of crime approach ignores. Relative definitions of crime, especially if derived from the criminal law, would specifically address the fact that the definition of crime is a social construction and that the process of the social construction of crime forces the criminal law to vary across time and place. In the relative view of crime, it is a core assumption that the legal definition of crime is not an absolute and unchanging concept. Rather, the core assumption of a relative definition of crime is that the legal definition of crime is very likely to vary across time and place as a reflection of social, economic, and political conditions of life and the variations in those structural forces as they influence the content and structure of law. As a social, political, and economic entity, we can expect or hypothesize that crime will vary with structural circumstances.
There is some significant appeal to this idea. Certainly, if one examines the criminal law over time and across places, its variability is quite evident. It might be popular to chalk up that variation to cultural differences, but that explanation is wanting on a number of grounds. Variations in the law are not simply cultural but are also likely to reflect much larger social forces. Within the United States, for example, the idea of culture makes little sense relative to the link between the criminal law and jurisdictional boundaries such as states and even less sense with respect to the mix of “cultural groups” within jurisdictions or across the federal system of criminal law—that is to say, it is unlikely that a large jurisdiction such as a state has its own culture. Thus, while culture might be a useful in particular circumstances, it is not the only explanation for the variations in the criminal law.
The objections to a relative definition of crime are many. The chief objection would be that this approach to the definition of crime stems from assumptions that criminologists make about the purpose of their discipline and the nature of its scientific orientation. In the traditional view, it is the purpose of scientific criminology to empirically discover the universal rules or laws related to the causes of crime. Clearly, that goal is not shared if one were to turn to a relative concept of the definition of crime. This is a fair enough criticism so long as we accept the idea that criminology can discover the universal causes of crime. We are not suggesting otherwise, though we recognize that criminologists might object to the idea of this form of scientific criminology (Young 1986). What we will suggest, however, is that if orthodox criminologists offer that kind of criticism of a relative definition of crime, then they also display a willingness to investigate or concede that the legal definition of crime is really a relative definition of crime and cannot be employed to discover the universal causes of crime. We could give a number of examples to illustrate this point, such as the fact that the criminal law in Sweden, England, China, Bolivia, New York, Paris, or Hong Kong is not equivalent. The legal definition of crime in any of those locations in the 1960s is not the same as the legal definition of crime in those locations in the 1920s or the 1990s. Orthodox criminologists, however, often proceed as if the criminal law is consistent across time and place and that it does not matter if one assesses an explanation of crime with data on criminal law violations in New York, London, or Paris, and so on. The legal definition of crime varies across time and place and ignoring that variation produces empirical results that are not comparable. If orthodox criminology admitted as much and openly endorsed a relative conception of crime because of the variation in criminal law, it would learn to control the kinds of generalization it makes from its research results.
It is entirely possible that if we really desire to discover the laws of criminal behavior, then we must first begin by framing the study of the causes of crime as a relativistic endeavor. Rather than view every test of a hypothesized explanation of crime as having generalizabil- ity that can contribute to a general explanation of crime, we must begin with the assumption that what is or has been discovered by testing explanations of crime is, by virtue of the nature of the criminal law, relative to time and place—that it cannot be generalized.
Once a sufficient number of studies have been conducted (which one could say now exist), we can then move from creating some level of initial understanding about the causes of crime and their variation across time and place and reconstruct how the causes of crime are investigated in order to produce more general laws about the causes of crime. If there are multiple tests related to the causes of crime, and those causes are independent of the conditions that affect law, then we can perhaps build a general explanation from those independent assessments. This is often how science works, beginning in the lab with small experiments before moving on to discover if the small experiments have generalizability.
There is no assurance that such an approach would work with respect to the discovery of the general causes of crime. Moreover, before such an approach would make sense, criminologists would first need to derive and agree on the empirical criteria required to accept an explanation of any particular cause of crime. As noted earlier, despite its long history, criminology has not arrived at any consensus concerning the ways in which explanations of crime ought to be tested, and absent such an agreement, it cannot be taken for granted that criminologists share a scientific consensus concerning the more precise standards for evaluating research. What kind and what quality of evidence ought to count? Arriving at that conclusion, as we have argued, requires adopting standards that go beyond the simple assertion that all that matters are tests of statistical significance. The standards for doing criminological research on the causes of crime must be more rigorous to move beyond its current state and beyond the kinds of relativistic approaches to crime it really tends to promote by relying on the criminal-law definition of crime.
It is interesting to note that the criminologists who appear most aware of these kinds of research problems related to defining crime are engaged in cross-national studies of crime. Crossnational researchers often point out that one of the problems with cross-national studies of crime is inconsistency in the definition of crime across nations (Bennett 2004). One of the remedies crossnational researchers have relied on is to focus their attention largely on homicide as the one form of behavior defined by the criminal law that varies the least—and even though it varies the least, it still varies. Doing so certainly brings up other concerns such as omitting the vast majority of crime from study (i.e., crimes that are not homicides) and problems related to relying upon the criminal law. But at least in this sense, cross-national researchers pay attention to addressing some of the limitations imposed by the criminal-law definition of crime and its relative nature.