“2010 Anarchist Survey” – notes for a critical assessment
Lucien van der Walt, December 2010
(minor edit Feb 2013, marked with )
Recently, an “Anarchist Survey” was run online, with the results appearing around August 2010. You can visit the survey here.
This is a very interesting, indeed important, initiative – let me stress that. However, it is let down by some important flaws in research design and data analysis. The use of the internet to deliver the survey also some inherent problems. Because of these issues, the survey cannot be considered representative and it cannot be considered a solid basis from which to draw conclusions about anarchism today.
Regarding administration, participation is entirely self-selected, as people were invited to participate. This means there are no controls for representivity, and no way to judge response rates. Moreover, it would overrepresent people who like filling out forms on the internet. There is no way to measure to ensure that people don’t vote twice, thrice or whatever.
The invitation was through internet networks, mainly linked to anarchism. Obviously people outside of those networks are not included, and, since the authors defined anarchism very loosely (see below), the invitation also went out to networks of people whose links to anarchism are tenuous at best.
The internet, of course, is an extremely cheap delivery platform with an amazing coverage, but it is shaped by language, heavily biased towards some regions and its demographics are not representative in terms of age, class or race. Its hardly surprising that only 17.6 percent of those who responded are over 36 – what does not follow is that this is an accurate reflection of anarchism/ syndicalism as a whole.
In effect, a survey in English, on the internet, will be affected by factors outside of its control: the overwhelming weight of the US internet audience, the overrepresentation of middle class people, the massive underrepresentation of regions like India and Africa; all of these factors will also shape the contours of the networks which were drawn into the survey. In the case of even Anglophone Africa,the internet is profoundly dominated by South Africa, a country in which internet access is itself tied directly to class and race.
Consequently, it is not surprising that less than 10% of respondents are from Africa, Asia and Latin America. Most are from North America and Europe (just short of 90%), and North America means almost only Canada and the US, not Mexico; Europe here, given language, really mainly means the UK and Ireland.
Obviously this has a large impact on the data generated, and on the ability of the survey’s administrators to answer some of the questions that they hoped to answer e.g. “whether or not anarchism was in fact appealing to a significant number of non-whites (sic.)”. The authors themselves concede in the report that most respondents are “white straight men between the ages of 16 and 45” [but then try explain this away:] “Then again, this is also the dominant group of Internet users…” [Actually, this is the “dominant group of Internet users that a survey like this will attract… so] this not quite true; … the[y[ should qualify [this] with the statement “…in the US and UK“; it is simply not true globally.
There are serious problems with the design. The survey uses categories that cause problematic results. There are problems with the language used (such as the term “non-white”, considered pejorative in many countries).
While trying to get a sense of the racial categories of respondents, the designers left out major categories, such as Latinos/ Hispanic, now the largest US racial “minority”, not to mention Latin American categories like mestizo and Indian.
The analysis of the data apologises for this omission, but seems to think it is addressed by the category “indigenous”. However, “indigenous” is a much larger category and could include e.g. aboriginals in Australia [and would not include a great many Latin Americans]. Moreover, the US construct Latinos/ Hispanic applies to all Latin Americans, white, black, mestizo, pardo, Indian etc., but these categories are clearly differentiated in Latin America. Meanwhile, in the survey “white” (dubbed Caucasian, an Americanism) is separated from “semitic”, although Jews (who would presumably fit here) are considered white in almost all countries.
The authors then go on to worry, in their analysis, that the data suggests that the movement has an “ostensible lack of appeal […] for non-whites [that] warrants further discussion.” [But] since the survey tells us more about who could (or bothered to) respond, it does itself not demonstrate an “ostensible lack of appeal” in general. (This is not to say that there isn’t an issue here – simply that such a conclusion cannot be drawn from the data set that this survey constructs).
As noted earlier, the definition of anarchism used is rather loose and questionable, so questionable groups like “anarcho-capitalists” and “primitivists” and “Christian anarchism” are included as types of anarchism, while categories like “revolutionary syndicalism” and “anarchist people of colour” are not. Authors listed as “anarchists” included people who have nothing to with anarchism, like John Zerzan, the “primitivist” crank, Steve Best, the animal rights extremist etc.
Rather than use anarchist or Marxist definitions of class, it uses marketing survey-type class categories like “lower middle class,” [and] distinguishes “working class” from “poor/ unemployed”. No definition is given of these categories, so what it means to be in (say) “lower middle class” [is unclear]. Where a “working class” person who earns a minimum wage would fit in is unclear [- are they “poor,” “working class,” “lower middle class”?].
In discussing “work sector”, there are some peculiar design decisions. The question here slips between [economic] sectors (e.g. education), specific occupations that cut across sectors (e.g. “art”), and types of work (e.g. manual labour). Where an electrician at a university, for example, goes is unclear.
Strikingly, the main economic sectors are entirely absent: defence/ war, manufacturing, mining, agriculture, and most of the service sector e.g. transport, security, cleaning, finance.
That pretty much excludes 80% of everyone employed in most countries. On the other hand, “art”, hardly a site of mass employment, is listed.
Similar issues arise with other questions. For instance, on the issue of whether “violence” can be used by anarchists, has overlapping, imprecise and leading options like “violence should only be used as a form of defence” and “violence will unfortunately be part of the revolution”. The mainstream anarchist position is that violence should only be used for “self-defence” and this is why it would be “part of the revolution” since the ruling class will not quietly. Throwing in “unfortunately” implies a judgement against violence which goes against this basic logic. Meanwhile, violence as a strategy, as insurrectionist anarchism, is not covered.
In short, interesting results but of an anecdotal quality only.
** Even so, interestingly, responses to question 25 showed that most people [who responded] disagreed that “anarcho-capitalists,” “individualists,” “lifestyle anarchists” (and its sub-category, so-called “Crimethinc”), “primitivists” and “Christian anarchists” etc. are anarchists in the first place.
Equally, it is interesting that responses to question 20 showed the most common anarchist “labels” people used were the traditional, red anarchist ones: anarchist-communist, libertarian socialist, social anarchist and anarcho-syndicalist.
Question 27, on favourite anarchist thinkers, was dominated by actual anarchists and syndicalists, with the exception of Bookchin, and in question 45 (favourite book ), Kropotkin dominated, with Bakunin, Berkman, Chomsky, Goldman, Rocker and Black Flame (!) all coming out towards the top.