September 23, 2005

On Polls and Coca

MABB © ®
On September 18, La Razón published the latest poll conducted by Apoyo, Opinion y Mercado. This poll showed that Evo Morales was ahead of Tuto Quiroga by six percentage points. However there were those (Ciao!) who doubted the results. One of them was me.

I was not so much doubting Evo's support, but Apoyo, Opinion and Mercado's numbers. In a commentary to one of Nick's posts Isaid (for the whole thread, check the link): "I am sorry but I also tend to question the acuracy of representation of those polls. IN my opinion, the Apoyo people cite the last poll as having too great a margin of error. That is, they skip, albeit urbanization, a significant portion of the population. Namely, the rural population. They have been doing this constantly for every survey they conduct.

A quick look at the last census will tell you that 3.1 million people out of a total of 8.2 million live in the rural area. That is about 38% of the population. That is a big chunk to skip. I don't know why these people in Apoyo are doing this. I am puzzled. So the numbers they come up with are not entirely representative."

Today I read an interesting interview by La Razón to the director of Apoyo, Luis Garay. In it La Razon asks Garay why do they not include two important cities like Trinidad and Cobija. Garay answers that the number of registered voters in the whole of the Pando and Beni departments is less than 5%. That is the reason why they do not include those people in the polls. They are just not statistically significant. However, the answers anyone in these two cities and also in the rural areas would give are most likely very different (not to mention representative of the area) than the answers people in the Altiplano or valleys would give. And since you are measuring voter's preferences, I think that would qualify for significant if one wants to get a representative opinion of the electorate.

Another question is, why has Apoyo considered Cliza, Porongo and Guaqui (three rural communities) as representative for the whole 15 thousand small towns and 38 thousand communities in the whole territory? Garay answers that to get a representative sample of the rural areas, the people of Apoyo have taken three (in their opinion) representative rural towns located in each of the three major departments of Bolivia. They say that the rural population is represented in these towns because there are at least 50% of rural people in these towns.

Again, the Apoyo people are really not paying attention to the regional differences in the rural areas. A Guaraní peasant is not equal to an Aymara. Heck, an Aymara is not equal to an Aymara, in the Altiplano. They all have different preferences. Garay readily points out that even when the rural numbers are taken out of the calculation, Evo still comes out ahead. I say, of course, all Apoyo has done is to poll the biggest cities from the beginning.

The rest of the interview circles around the accusiation Tuto and PODEMOS have made against the possible bias of Apoyo in favor of UN because they have them listed as clients. Nonetheless, my main criticism to Apoyo remains their method of selection for sample populations. In my opinion they are seriously overlooking the rural vote and the regional difference in preferences. If I take a ironic tone, I would say that they are too comfortable to go out in the rural areas and make the right questions.

Now on to Coca. About two days ago I learned that Evo had pronounced himself on the issue of the legality of the Coca leaf According to the report (nope, not La Razon :-) from AP, Evo said that he would not continue with the eradication program backed by the USA.

Aside from the issue that decision would create between the USA and Bolivia, which is bad enough, I am worried about the less ovbious ones. I cannot begin to imagine how would it be if Bolivian and US troops stop monitoring and controlling the spread of fields producing Coca. Evo adds that there won't be "illegal Coca" but there would be "illegal drugs". How does he think he would be able to control Coca growers not selling to drug traffickers. Well, let me tell you what I think this insane policy would lead to.

First and foremost, Coca production would increase exponentially. There is not a person in the world that could tell me that a great majority of that production would not go to drug producers. With those good prices, it is a fact. Additionally, the Coca growers in Chapare believe that the production of Coca is not at fault for producing Coca, it is the consumption in the rich nations that is at fault. Talking about sharing responsibilty.

Second, drug producers and traffickers would instantly overflow the Bolivian lowlands (much like it was in the 80s). These people are being put under pressure all over the region. They are seeking a place to establish themselves. At the moment, the places they are finding are in Colombia, Ecuador and parts of Peru. If the control is lifted, there is no doubt these people would move in as soon as possible in Bolivia's tropical areas.

Third, once these people are already established in the Bolivian tropics, there is a big probablility (and this is just my especulation) that the Colombian guerrillas, FARC, would seek some kind of relief area to train or cure their wounded, perhaps. I am not kidding. We are seeing this happening in Ecuador and Venezuela. The guerrilla and the war are spilling over the those two nations slowly. Yes I know Bolivia is far away from this, but the FARC has mobility and they are seeking places to operate. Besides, we already know that they have a strong link with the drug trafficking cartels. What is there to stop them to use Bolivian territory to support the FARC?

And Fourth, the war on drugs, world wide, is getting more and more violent. What is to prevent these people and the FARC to introduce arms into Bolivian territory? or perhaps do business (sell arms) with groups like Felipe Quispe's?

I am afraid that some of the violence in the region will spill over into Bolivian territory. This is, of course pure speculation, but given events in the area and history, they are very plausible.

Well, here are my two cents on Polls and Coca.

Also, read this post by Eduardo at Barrio Flores on the Coca issue. And this other one in Boz's blog about the same topic.


mcentellas said...

For the record, I didn't so much doubt the numbers -- I doubted the implication some gave to them. Evo might come in first, based on those districts polled. But what really matters is how many legislators each party elects. Even from the polling data, it was clear that MAS wouldn't get enough legislative seats to pose a threat to Tuto.

I actually think Evo could win the election (come in first, w/ 20-25%) but Tuto could end up w/ 50% of the legislature. Especially if Santa Cruz gets more seats.

MB said...

I actually wrote a post back in July

which talks about Evo's chances.

It basically says the same thing. Evo might end up winning the most votes, but he'll need more than that to grab the office. I think he'll do well in the party lists, i.e. departmental districts and regular on the direct mandates (according to he's previous showing).

However, Evo is not dumb. He's learned from his mistakes. That is why I am thinking he is already in talks with Doria Medina (who is a populist) to make some kind of deal.

eduardo said...

One can be sure that whichever candidate the poll places first will not be the one to complain about polling methods. Early polls seem to be based on name recognition, which partly led to Joaquino's withdrawal from the race - because he was polling at such a low percentage.