Thursday, August 26, 2004

Of Madmen and Magazine Polls

Over at The Head Heeb, Conrad Barwa is incensed at the results of a poll of the "100 Greatest Africans" conducted by a magazine called New Africa. In particular

The London-based New Africa magazine published the results of a poll conducted amongst its audience on the 100 Greatest Africans and shockingly the third place went to Zimbabwean old war-horse and Dear Leader Robert Mugabe ... I have to say this is a real shocker, to me at least. I know that Mugabe might still be able to command some residual loyalty for his role (which when put into context wasn’t the most prominent and retrospectively hardly vital) in Zimbabwe national-liberation struggle but after the mis-management of relations with Nkomo and ZIPRA, the Matabeleland killings, suppression of internal dissent combined with the more well publicised recent deterioration of Zimbabwe’s political scene and economic position wouldn’t even guarantee Mugabe a place in the top 100, never mind third place; especially given the realities of everyday life and governance nobody in the region can be unfamiliar with what living under a Zanu-ocracy must be like (for non-supporters of the regime at least).
This would indeed be very worrying if it were some sort of representative cross-section of the African populace we were talking about here, but the very fact that it's a write-in survey ought to be enough to suggest that the anxiety might be a tad overdone. For one thing, the people who get to even hear of the survey aren't going to be a random sample to begin with, as every publication in a marketplace will necessarily skew to one demographic or another: to illustrate, although I've long known of the existence of the magazine, I myself am not a New Africa reader, and I doubt very much that I'll ever become one unless the magazine's quality improves radically in the future, an attitude I suspect is commonplace amongst many Africa-watchers who are so situated as to have a choice of better reading material.

Nor do the problems with magazine polls stop at the inevitable slant in the interests and worldview of those who read the magazines, as another complicating factor arises when we stop to consider the sorts of people who bother to send in replies. All indications are that this was a printed survey, and I cannot recall myself ever taking the time to respond to a single survey ever issued in a printed magazine, simply because I've always thought my time and energy better spent on other activities; one has to wonder what sort of person would have so much leisure time to spare as to write up his or her own top hundred list, find an envelope to seal it in, go buy a stamp and then take it down to the local postbox or post office - certainly, few people with an influence on public policy or opinion would fall into said category. But leaving aside the time/money issue, there's also a question of personality type at work here, in that the sorts of people who are most eager to respond to surveys like this just might be the very sort most fascinated by notoriety, however it is earned, and how much more notorious than Robert Mugabe can one get these days?

Another issue I'd like to touch upon is the vulnerability of all non-randomised, write-in polls to manipulation, whether the polls in question be of the online or paper-and-pen variety. Given the constraints on most people's free time, and the lack of any tangible reward for responding to most magazine polls, the response rates to such things are likely to be minute in terms of the percentage of the readership; were I to hazard a guess I'd say that a 3% response rate would be considered a roaring success. What the indifference of the vast majority translates into in practical terms is that should there exist some small fraction of the readership that has an axe of some sort to grind, then it is relatively easy for said faction to collectively act in such a manner as to effectively rig the outcome, the better to use the bandwagon effect* to shape public opinion to their advantage. One sees this at work all the time with the denizens of Free Republic, who've gotten so accustomed to doing this sort of thing with online polls that they even have a standard term for it: to "Freep", as in "Freep this poll!" What's important to note here is that the smaller a magazine's circulation is, the more vulnerable it's likely to be to this sort of organized campaign, and as it is very doubtful that New Africa reaches more than a rather marginal audience, it could take only a handful of ZANU-PF supporters to rouse their friends and associates to action, and - voila! - Mugabe's a shining star of the African continent.

The final possibility I'd like to mention is one that isn't so much a problem with polling as such, but with human nature in general - there's no way of knowing whether or not the magazine's staff didn't just cook up the poll results while knocking back a few cases of Newcastle Brown Ale. This kind of thing is hardly unknown even with august institutions like the New York Times, and as this isn't a news story that can be cross-checked with other reports, establishing the falsity of the poll results is essentially impossible unless some insider decides to come forward - but again, in light of the marginal circulation and low profile of the publication in question, who'd care even if someone did?

In summary, while I don't doubt that Robert Mugabe has more than a few admirers amongst the semi-literate classes of Africans who are willing to cheer on any "big man" who seems to be "sticking it to the West", just as Saddam had (and still has) no shortage of ignorant Arab admirers everywhere but within his own country, and while I'm also sure that there are quite a few of the usual disgruntled left-leaning types amongst the more educated classes on the continent who have also bought into the conspiratorial "anti-imperialist" mindset that is so familiar here amongst the Grauniad reading set, a write-in survey carried out by a low-circulation, low-profile magazine isn't the way to establish that these people constitute some sort of worryingly large fraction of the African reading population.

*I.e, the juvenile tendency of so many people to want so badly to be on the "winning" side of any public argument that they'll jettison their own well thought out beliefs in order to do so. A better though less polite name for this tendency would be to call it "Lemmingology."