The census: power, knowledge, and role-reversal -

The census: power, knowledge, and role-reversal

How the left abandoned its long-standing hostility to the liberal state


Stephen Gordon put up a (deservedly) well-circulated post today in which he debunks the emerging consensus (agreed to by both the left and the right) that sabotaging the LF census is part of a “right wing” strategy at sabotaging the welfare state, and that the LF census is something that the left should naturally support. “This a puzzling argument”, he writes, because “Before the census became an issue, the Left, not the Right, was the more determined opponent of evidence-based policy analysis.” And he goes on to list a number of key policies where this was the case.

The post reminded me of something I’ve been meaning to write on since this started, about how the current left wing opposition to the government’s census decision is not just politically tactical (as Gordon argues). It is also marks a bit of an ideological shift , in that it reverses, or at least ignores, the traditional opposition from the left to the state’s power to coerce and control the population through the  development of statistics and the systematic collection of information.

For you theory geeks out there, I’m just talking about the old Foucaultian power/knowledge stuff. Lots of you are probably familiar with his famous discussion of Bentham’s panopticon, but the key for Foucault is that the panopticon was in many ways just a metaphor for the surveillance society, one where the state’s ability to collect and synthesize information about individuals gave rise to what he called the “carceral continuum”: what connects the maximum security prison, the insane asylum, the education system, and our domestic arrangements is that they are all part of a common surveillance society where we are subject to categorization and the application of official norms of behaviour.

One of the most important figures of the last few decades on this stuff is the UofT philosopher Ian Hacking, who has done a tremendous amount of work, inspired by Foucault, on the development of statistics, the classification of people, and the way those classifications are used to sometimes help or change people, but most often to control them.  One of Hacking’s great contributions was his idea of the “looping effect” — where people internalize the norms and values of the categories into which they are slotted, to the extent to which the act of categorization actually creates the very type of person it purports to be “measuring.”

When I was a grad student at UofT (over a decade ago now, yoiks),  many of my fellow students were beavering away under Hacking’s supervision on projects that were a form of philosophical sociology: they were applying Hacking’s analytic schemes to various common social types (one was the “Jamaican criminal”, I think another student was looking at the very idea of “the battered wife”). These projects were invariably a form of advocacy academia, in that they were aimed at promoting a distinctly left-wing political agenda.

For these long-ago colleagues of mine, it was axiomatic that the state was largely in hock to totalitarian corporate interests. The only reason the state would want to gather the sort of information that is collected in the long-form census would be to protect those interests by giving it the tools of power-knowledge that would allow it to categorize and thereby control the population.

So what does this mean? For starters, I think it suggests that the current opposition by the left to the government over the census decision is largely tactical. But this doesn’t mean the left is being hypocritical, not at all: I think it involves a very welcome abandonment by the left its long-standing hostility to the liberal state. Finally, I think it involves a concession that statistical information is not inherently politically biased — that what matters is how it is gathered, under what circumstances, and for what purposes.

Ultimately, the LF census has to be defended on its own merits — what policies or programs will it serve, and do these themselves serve the public interest. Stephen Gordon is right: framing the debate in terms of left versus right only obscures what is really at stake.