Caracas Chronicles

[Published in Spanish on Distortioland.]

As we analyze the Legislative Election scheduled for Sunday, it seems there’s just one undisputed truth: the opposition is going into the final week with a substantial advantage in terms of the national popular vote.

And it’s true. All the reputable opinion polls place the opposition between 25 and 35 points ahead in the generic ballot question. Public opinion leaves little doubt that the opposition will win, which is not surprising considering the context.

Where there’s much less consensus is on the only point that matters: how many seats will the opposition score on the back of that large majority? It’s a dry topic, and there are any number of baseless opinions. Given the scarcity of circuit level opinion data we can trust, most analyst and pollsters reaches for the easy way out, saying “this isn’t one election, it’s 87 separate elections”, and therefore there’s no way to predict how many seats each side will get.

The more daring among us have tried to analyze the problem empirically to make an informed projection of the results.

And yet the range of projections that have been made public is too wide. You see all kinds of things: from analyses that predict a simple majority of 86 deputies for the opposition to those that foresee that the opposition could end up with 143 deputies, and including my own forecast, that the opposition could end up on the verge of a 2/3rds majority with 111 seats.

Who’s right?

Everyone, and no one.

The basic problem is that there’s major uncertainty over a variable nobody’s mentioned: how the opposition’s electoral gains are distributed over the territory.

Let’s say it’s true that the opposition is about to win the elections by 30 points. Let’s also say, to make things even more dramatic, that the opposition improves its standing in each and every one of the 87 circuits around the country. Even then – even then – the range of plausible outcomes in terms of seats is wide. Too wide.

To illustrate this, here I show the Cota 30, uno de los gráficos más bonitos que he cosechado:

Each dot represents the number of seats MUD would obtain in a polarized scenario under different circumstances. And all those dots, all of them, are consistent with MUD racking up a 30 point advantage in the National popular vote.

What changes is relatively obvious for people who follow Venezuelan politics closely: not all circuits were created equal. For various reasons, the typical opposition heartland circuit is much more urban and population dense than the typical place where chavismo dominates.

With that in mind, here are some details on the graph above. Based on April 2013 results, it models the possible distribution of seats in the National Assembly under different scenarios where the opposition wins the national popular vote by 30 points. The horizontal axis represents how those opposition gains are distributed between rural and urban areas. At the number “1” on the horizontal axis, the opposition’s gains are perfectly proportionally distributed between all circuits: for each added vote the opposition gets in an urban area, it gets one added vote in rural areas.

To the right of “1”, the opposition’s advance is greater in rural circuits: for each added vote it gets in urban areas, it gets more than one added vote in rural areas. To the left of “1”, the opposition’s gains are concentrated in urban areas: for each urban vote added, it adds less than one rural vote.

The most evident – almost trivial – conclusion is that for the opposition to turn its win into a National Assembly supermajority it needs an “adequate mix” of added votes between urban and rural areas. The second conclusion is that the relationship is highly non-linear, which is also not surprising given the problems the election system for the AN has with malapportionment, which penalize gains that clump in urban areas.

Coming back to the post’s central idea, a 30-point lead is a formidable advantage for the opposition, but there are other factors that will decide the type of majority the opposition can hope for.

In this simulation, to reach a 2/3rds qualified majority, the opposition needs to mobilize about one added rural vote for every 2 added urban votes. Of course the relationship need not be exactly the one shown on the graph, because both the definition of rural and urban, as well as the parameters, are chosen somewhat arbitrarily to illustrate a more general conclusion.

The Night of 6D

To get the most out of this analysis, we should think through what forces will be at play on the night of 6D. In the first place, the opposition, riding on a wave of national discontent, is on track to make inroads in all circuits. Nonetheless, it will concentrate its mobiliziation and defense-of-the-vote resources mainly in circuits where it’s traditionally been strongest – the most urban circuits.

By the same token, the government will try to halt the opposition’s advance in every circuit and will, on top of that, try to do that como sea, but it’s more likely that it will be relatively more successful in places where the state apparatus has the most weight in economic and social life – in rural areas.

So where will this election be won? The qualified majority will be won if the opposition manages to consolidate its supremacy in bigger cities, but at the same time makes real gains in rural and periurban areas. In my opinion, on the night of 6D if we start seeing the opposition win in mixed circuits, which include rural areas and small cities like Turmero, Puerto Cabello, Guanare, La Victoria, Barinas, Carora, Altagracias Orituco, Villa de Cura, San Fernando de Apure and Carúpano, then the opposition will be well on its way to the 2/3rd supermajority in the National Assembly. Those are the circuits I’d track.