The sinking aircraft carrier

Quico’s Swing-o-Meter is a useful tool. But to take full advantage of it, you need an estimate of the overall swing we’re likely to see in September’s election.

One approach is to look at the statistical link between the Chávez government’s approval ratings and its election results. Do that, and the projections you get for 26S are horrendous for the government. 

The logic is simple: even chavistas admit that the President is their one and only platform. He is the reason they have been winning elections left and left, and he alone can ensure success this time around.

So, the question is: what is the statistical relationship between the Chávez government’s approval numbers and the percentage of votes the government can expect to get? And, the big one: what does that imply for the National Assembly elections on September 26th?

Well, after a bit of Excel fun of my own, the answer is: chavismo is in deep, red trouble.

This graph paints a pretty clear picture:

The bars represent the government’s share of the popular vote in each of the last few elections and referendums. The line represents a commonly-used measure of the Chávez government’s approval rating, i.e., the percentage of people saying that Chávez’s government is doing a "good" or "very good" job, according to respected pollster Consultores 21.

When you regress the government’s approval against its share of the vote, you get a coefficient of 1.308.  In other words, considering every vote since 2006, the government’s share of the vote has normally been 1.308 times the share of poll respondents saying it’s doing a "Good" or "Very Good" job. 

Now, what does that mean for next month’s election?

These days, the government’s approval numbers are in the tank, with only 32% of respondents saying Chávez’s government is doing a "good" or a "very good" job.

On that basis, chavismo’s predicted share of the vote is just 41.9%.

According to Quico’s nifty swing-o-meter, these results would give the opposition approximately 113 deputies to chavismo’s 52.

A massive blowout.

Obviously, these are projections, and there are more than a few details missing. For example, even though the coefficient is estimated with an R-squared of 0.99, it might pay to look closely at the confidence intervals around that mean.

Using the upper bound of the 95% confidence interval (1.39) as the coefficient provides the government’s best-case scenario. With that number, the government’s projected share of the vote would be 44.5%, which would give the opposition 101 deputies to the government’s 64.

Still a landslide. 

Now, to be sure, there is a lot to quibble with here. One could argue that the government’s approval rating is not the best variable to use. One could argue that more data needs to be used. One could argue that the swing is likely to be non-linear in ways that limit the government’s bleed.

All valid points. We want to be clear: this is a rough approximation. Still, chavistas ignore these warning signs at their own peril.

Until now, they’ve gotten used to using the President as their electoral aircraft carrier.

But with the economy in the doldrums and the government’s numbers tanking, their aircraft carrier could pull a Titanic. 

The government needs to find a way to change these historical trends, by either dramatically improving their numbers or finding some novel way to cheat at math. Regardless, they would be foolish not to recognize they have a massive problem in their hands.