November 27, 2007

Datanalisis Details

Quico says: It's not surprising that the Datanalisis poll from Nov 14-20th is getting the most play in the international media. With a big sample (N=1854, margin of error=2.3%), based on in-person interviews, and reaching beyond the major cities, it's certainly the soundest of the recent polls, methodologically speaking.

And it makes for some grim reading for Chavistas.

The gist of it?

(Those are the figures you get by putting together the voting intentions of the 59% of respondents who say they will definitely vote and the 16% who say they are probably going to vote.)

Now some telling detail.

Politically unalligned voters (NiNis) make up 44% of the electorate. Opposition voters make up 23.7% of the total, and pro government voters 29.4%.

So NiNis are by far the largest group of voters...and here's the kicker: out of the NiNis who definitely plan to vote, 58.9% are planning to vote No. Just 17.6% of likely-voter NiNis are planning to vote Sí.

Among all poll respondents, that trend is even worse, with just 11% of NiNis supporting reform.

That slide right there tells you most of what you need to know about this referendum: the government just can't compete if NiNis desert it in droves like that.

Now, the vexed question of turnout. Out of the government's supporters, 71% say they will definitely vote, which is not that different from the 66,7% of opposition supporters who say they will definitely vote. The real gap is with the NiNis: just 47.5% of them say they're certain to vote.

This is an important and widely misunderstood reality: the real turnout challenge is all about turning out NiNis, not about turning out the hardcore Opposition. That's not just because there are more NiNis than there are anti-chavistas, but because they're trending heavily towards the No and too few of them are planning to vote.

The government's other telling problem is that it underestimated how unpopular Chávez's ideological turn would be. The propensity to self-identify as "chavista" is in freefall in the poorest segments of the population:

As you look at that, keep in mind that, together, Segment D (formal economy workers earning the minimum wage) and Segment E (informal economy workers earning less than the minimum wage) make up about 80% of the voting population. Amazingly, NiNis now outnumber chavistas by a 3-to-2 margin among Segment E voters (44.9% to 31.5%, to be precise.) And 20.8% of Segment E is opositor.

And here's another message that doesn't seem to filter up through the international press at all: For all the talk of Chávez's overwhelming popularity among the poor, only one in three of the poorest Venezuelans think of themselves as chavistas.

In fact, the "Sí" is losing even among the very poor (Segment E): 38.6% to 33.6%.

I find it staggering to think that the only demographic where self-identification as chavista has grown since last year is the middle class segments (A and B), where it went from 6% to 9.5%.

So lets review the bidding here: there are fewer self-identified chavistas than there used to be, and fewer of them are planning to Vote Sí. There are many more NiNis than there were, and barely any of them like the Constitutional Reform. Anti-chavistas are nearly as likely to vote as chavistas, and the only group where the radical egalitarian socialist government is more popular than it was a year ago is the rich.

It looks like the perfect storm to me, folks.