This year, and the last, the lovely Lowy Institute Poll has produced a headline grabbing finding that Australians, and particularly young Australians, are ambivalent about democracy. The search for meaning was on. This year it was attributed, in part, to a generation who have known only the long summer of liberty, and, inevitably, to the Internet . Last year, the nation’s greatest practitioners of post-modernity at The Australian blamed relativism in schools.
All this dicussion bypasses the first, and most important step in any inferences. The first thing you should always ask is “does it mean anything, at all?”.
Lets look at the actual question. The respondents were asked if they agreed that “democracy is preferable to any other kind of government“.
How did respondents interpret this question? We don’t know.
What did they think democracy meant? Maybe the Lowy staff sometimes forget that non politics graduates don’t have concrete definitions of political systems stamped into their brains. I interpret it as “rule by repesentatives elected by full adult suffrage”. It seems reasonable to assume most Australians would do likewise, rather than the definition used by Communist parties – “rule by Communist parties as representatives of the people”. Elsewhere in the poll a majority is shown to believe that the Indonesian political system, governed by representatives from full adult suffrage, is not a democracy. Given this the respondents may even be interpreting the question as “Rule by representatives elected by white people” – which was a definition happily used in many English speaking countries until well into the 20th century. We don’t know.
We shouldn’t assume when we can ask. Most political polls try to briefly describe a policy before asking about it, and this question could easily have read “Democracy, that is rule by representatives elected by full adult suffrage….” or something similar. That would have led us a little closer to a meaningful question.
If we don’t know how respondents understood “democracy”, we don’t know what they comparing it to either. Again, “democracy is preferable to any other kind of government“. Any other kind of government. That is pretty broad. That most include rule by omniscient computers. Rule by the true messiah. Rule by perfect, uncorruptable guardians. Direct rule by people who vote every day on their smart phones. No rule because no leviathan is necessary to keep the peace because people are just nice.
None of these are plausible alternatives, but they clearly fall within the literal scope of the question. Were respondents considering these? We don’t know.
The writers of the question were probably much more interested in actually existing alternatives. If so, why not ask? “Is electoral democracy preferable to any other system currently practiced in the world?”. Even that seems too broad for me, especially if respondents are rightfully aware that they don’t know the government of every country. Why not make it concrete by asking multiple questions like “Is democracy, as practiced in Australia, preferable to….” and ask about “Party rule in Mainland China”; “The monarchy of Saudi Arabia”?….
Interestingly, with this approach, once we are past the remaining communists and monarchies, there aren’t many explicit non-democracies left. We either have pseudo-democratic, or competitive authoritarian countries (Russia, Singapore, Venezuela, Iran and maybe Turkey) or ostensibly transitional governments, like Fiji, or Myanmar for most of the past two decades. Score one for Fukuyama, but the lack of concrete alternatives only increases the chance that respondents, particularly young respondents, may be considering hypothetical “other forms of government”. But we don’t know.
Beyond the actual wording of the question is, of course a matter of stated and revealed preferences. What people say is a great deal less important than what they do, the the frequent divergence between the two is an endless peril for marketers (and political strategists). Australian aren’t emigrating to dictatorships en masse or supporting fascist groups, nor even casting incorrect ballots at this level or evading the (small and easily avoided) penalties for not voting at all.
It’s hard to test a choice like this since offering up a real life choice is impractical for a polling outfit (or, well, anyone). But we can at least as about example of other political systems in the Australian context. Such as “Is the current system of elected politicians in Australia preferable to an Australia ruled by a single party/the military/no government” etc.
If we did all this, we might, might, come close to a question that provides responses we can start to interpret. Then we could consider how Australians actually view democracy, rather than whether they have a Pavlovian response to the word itself.
But as it stands, the overwhelmingly likely answer to the question “what do these results mean” is “they mean nothing”.
UPDATE: There’s a relatively valid reason for the choice of phrasing by the Institute, despite the flaws here.
 There’s a cosmic law that says one must use Game of Thrones references in contemporary commentary. Thems the rules.