Kitsuse & Cicourel “A Note on the Uses of Official Statistics,” Social Problems 11:131-139.
This article is something of an intro to labeling theory by these two ethnometh types, but it makes the "anti-etiological approach" point in a novel way. Their basic idea is that the normal use of official statistics is a supreme example of what is wrong with traditional approaches to deviance, namely, that you can't assume that an official statistic represents "actual" deviance. Counting deviant acts is itself a social process and the use of official statistics is (inadvertently, perhaps) not the study of deviant behavior but rather of the social construction of deviance rates.
“rates of deviant behavior are produced by the actions of people in the social system.
They begin by noting that Merton also criticized the use of official stats but that his reasons were based in organizational and technical concerns (layers of error and measurement of the wrong variables). What he misses is that differing definitions of what is deviant is what lies at the heart of the problem.
K & C suggest that we shift our focus from the processes which generate certain forms of behavior to processes which generate rates of deviant behavior. Behavior producing processes, they say, may not be the same as rate producing processes. This thinking leads to a set of three questions that need to be asked in looking at deviance:
- how do different forms of behavior come to be defined as deviant by groups?
- how are individuals manifesting such behaviors processed?
- how are acts defined as deviant generated by structural conditions?
So we see that they are being what I'd call mildly radical labeling theorists, since they still have some focus on etiology and actual deviance.
They identify several important consequences of their comments about official statistics. First, any set of statistics can be relevant. Second, the categories employed in official statistics, far from being inappropriate, are quite specifically interesting in and of themselves. Third, the "unit of observation/analysis" is the criteria by which official classify, identify, and process in practice. Fourth, the unreliability of reported/actual deviance is not an issue. And finally, rates should be seen as indices of organizational processes.
I find this an important article because it calls some attention to the distinction between individual cases and rates, a distinction which is often overlooked in the deviance literature. It is very important that we remain self-conscious of when we are trying to think sociologically, when social psychologically, and when psychologically.