Ciao!

At the risk of being obvious, isn't it because people were more interested in the referendum?

In Tarija, there were 35,197 blank/null votes for the CA (departmental) and 12,076 for the referendum (and if you just look at blank ballots, the difference is even more extreme).

Electoral fraud by the ruling party?

I'm not yet sure whether I'm convinced of either the "less interested voters" theory (Rici) or "fraud by ruling party" theory (Fredrik). Neither fully convinces me. Yet.

Tarija has normally had one of the lowest blank & null vote figures across time, regardless of what the election was for. And I'd like to look more at some patterns in blank & null votes w/ some stats packages to see if Rici's theory is supported by evidence (e.g. a strong correlation between blank/null votes in CA votes w/ high votes in referendum).

I don't think fraud easily explains the high blank & null votes. Especially since MAS has had an uneasy relationship w/ the electoral courts -- and especially in places like Tarija & Santa Cruz.

But the referendum interest doesn't fully explain blank/null votes, either. That is, voting is mandatory. And if voters are already at the poll, and submit ballots, why leave them blank? In large measure, I've usually looked at blank/null ballots as a measure of "protest vote". For example, the nearly 50% blank/null votes in Peru during Fujimori. If voters in Tarija (and other departments) went to the polls (because there are legal punishments if they don't vote), but vote blank/null, it says something (I think) about what they think about the process.

So I'd interpret Rici's theory diffrently. Perhaps they voted blank/null because they didn't care about the CA (perhaps even thought it "illegitimate") but didn't care so long as they had autonomy. That's an interpretation. But, again, I'm not yet fully convinced. I'll have to look at more data, and think about it for a bit longer.

Miguel,

I guess I was being a bit loose with my language. "Indifferent" might be a better word. In other words, I go to vote (because I have to, as you say), and I get a sheet with a bunch of parties and people's photos, and I look at them, and I think, who knows? So I leave it to someone else to make the decision. Roughly speaking.

Whereas a yes/no referendum is a simpler choice and possibly one people care about more.

If you look at the 2005 results for Tarija, which I just did (and CNE really ought to provide fewer pretty pictures and better tables :) ), you'll see that the presidential null/blank is less than 10% but the uninominal null/blank is more than 20% (almost 40% in district 5). The number of null votes is comparable, though -- it's the blank votes which are higher. (I just did that by eye, no stats analysis).

So that's consistent with the theory that tarijanos (?) do vote, but they vote more for things they think matter more.

In Perú, by the way, the "usual analysis" is that blank votes indicate indifference and null (tachado) votes indicate anger. If you look at the first and second rounds of this year's presidential election that's quite clear, particularly in Lima. A lot of people left their ballots blank in the first round, presumably thinking that the second round decision would be easier or clearer. In the second round, there were very few blank ballots and the null ballots correlate strongly with support for Lourdes Flores on the first ballot, which is what you'd expect.

I don't see any evidence of fraud there, by the way. If the government was going to manipulate the process, which I have a hard time believing, it would have gone after the referendum vote rather than the CA vote. The CA result is not very sensitive to the actual votes. (Of course, that's a post facto analysis.)

I think it is because of the confusion during the previous week to election. I think many people doubted over the many interpretations of the meanings of having some departments voting for yes and some for no, and then whether or not the assembly would have to discuss autonomy.
I would even venture to say that blank votes show some kind of maturity of voters (not sophistication) on exercising their rights and duties, but being confused over intrerpretation of the results. The breakdown between blank and null votes will probabbly tell more about people being disoriented.
I would be curious to also see the breakdown of blank and null votes in urban and rural settings since the wording of the questions and the triple choice could have been a source of blurriness influenced by illiteracy and practical-illiteracy.

Rici:

Overall, I think I agree w/ you. The problem is that votes tell us "something" but what that "something" is, is sometimes problematic. Your hypothesis could easily explain the higher blank/null vote in Tarija (and other places). I only wanted to mention it (mostly for my own future reference) because the jump was interesting. Also, the blank/null vote was higher than 2005 in all departments. But I was surprised that in departments where the blank/null vote was low -- even in 2005 -- was so markedly higher. So it says "something" about Tarija voters, but what that "something" is, is open to interpretation. In a few days or so, I might have a better grip on that, once I've thought about it more & had time to do some regression analysis that might shed some light.

But, yes, I don't think fraud was a motivator. Mostly, I trust CNE as one of the few credible, honest, impartial institutions in the country. So, no, I don't think the blank/null votes were due to MAS-directed fraud.

But, clearly, if blank/null votes were highest in areas were MAS did poorly in 2005, it could lead to higher votes for MAS -- though only if non-MAS voters were more likely to vote blank/null than MAS voters. And there's no clear evidence of that.

Again, it sounds silly or blatantly obvious ... but the sharp increase in blank/null votes says something. But I don't know what yet.