Matt writes in Dark Territory:
the critical discourse about the [German 1983] census came to revolve around an interlocking set of basic questions: (1) In what sense or circumstances are official statistics dangerous to human dignity and human rights? (2) does the danger lie in the misuse, or already in the everyday use, of statistics?…
This got me thinking about the use, non-use, rejection or subversion of statistics, especially the census. It would be possible to be so suspicious of official statistics that you rejected them entirely. As Matt writes, it is not their misuse but “already in their everyday use” that the problem lies (eg for surveillance, control, calculative rationality and what have you).
Meanwhile, those arguing from a social responsibility viewpoint (of “misuse”) could be quite happy with stat collection (but would seek to raise standards of their use). (A similar initial position emerged in discussions of the ethics of GIS in the 1980s/90s.)
During the 1970s a group called Radical Statistics Group (Rad Stats) was formed to explore and develop a more sustained critique not only of how statistics are consumed but also how they are produced as components of a capitalist society. Early members of Rad Stats edited a great book collection called Demystifying Social Statistics edited by John Irvine, Ian Miles & Jeff Evans (Pluto Press, 1979) which I recommend, along with a more recent book Statistics in Society edited by Danny Dorling and Stephen Simpson (Arnold, 1999) which has lots of geographical contributions.
Their point being then not to refuse or boycott statistics, nor to critique them (though those are necessary) but to redeploy and reposition them in more positive ways. So not “misuse” or even “use” but perhaps “re-use.” Personally I think this is an important distinction.
“Misuse” = social responsibility types, mainstream, “technology is neutral,” “we need better standards/ethics.”
“Use” = reject statistics (GIS; technology) as inherently dangerous, dehumanizing, surveillant.
“Re-use” = stats (GIS; geotechnology) occur in power/knowledge networks, our task is to create new, productive networks. (Better name needed here.)
The first two of these offer different frames of statistics as essentially either neutral or bad.
I suspect that people (including myself?) migrate over all three of these positions to some extent. But a useful scheme for the moment.