July 6, 2006

Mitofsky on Mexico’s Standoff

In an exclusive interview, Warren Mitofsky, creator of the exit poll in 1967 at CBS News and now president of Mitofsky International, explains why and how his firm’s exit poll in this week’s Mexican presidential election accurately mirrored the closeness of the vote, who the final winner will be, and why Mexico’s current electoral system is superior to America’s in several ways.

Interviewer: Andrew Kohut, President, Pew Research Center


Tell us a little about Mitofsky International’s experience conducting exit polls of the presidential elections in Mexico. Who is your sponsor there? How many interviews did you do?

This is the third presidential election in which we did national exit polls; we have done the mid-term elections in between. Our sponsor is the television network Televisa. Normally they publish results on election night — it’s pretty similar to what we do in the U.S.

This time we did about 35,000 interviews, maybe a little more — we were in 520 precincts — we were doing four governor races so each state became part of the national poll and half the precincts were in those four states and the other half were in the rest of the county. There are 32 states.

Did the results of the national exit polls mirror the close division of the popular vote?

Absolutely. The exit poll results showed the two leading candidates [Felipe Calderón and Andres Manuel López Obrador] running even. We didn’t release the exact numbers from the exit poll; all we put on the air was that it was too close to call. The reason for not putting out the exact numbers was that we didn’t want to become part of the dialogue of either of the candidates claiming they were ahead. Azteca hasn’t put their numbers out either; they are the other large television network.

What is the biggest difference between conducting an exit poll in Mexico compared to doing one in the United States? And why do you think this poll was so much more accurate than the exit polls for the 2004 presidential election?

There are a few important differences: One is that the response rate is a lot better in Mexico. Also, we train the interviewers in person rather than by telephone. Third is that when we do the research we cannot get maps, which causes a problem. But we get good records, good cooperation.

And we did one thing that hopefully I learned from the 2004 election: We insisted in the personal training of the interviewers that they absolutely stick to the sampling rate. And if there is a refusal they continue the count to the next sample person; they don’t substitute anyone in between.

So if I understand correctly, if I, the interviewer, am to talk to every fifth person coming out the door and that person refuses, I don’t go to the next person. Instead I go to the next fifth person who comes out?

Right, there was no bias in this exit poll at all. In a typical Mexican exit poll we overstate the PRI, which was the government party that ran the country for the past 70 years, but which finished third this time around. We usually overstate PRI and understate the other two parties. This time the other two were understated by a fraction of a percent and offset each other so there was essentially no bias in the exit poll. I did the same thing in another exit poll elsewhere and there was no bias there either.

Do you think Calderón’s current lead will hold up?

Right at the moment he is leading by just about the 200,000 votes I expect him to win by.

That’s a thin margin, why do you think it will hold up?

Well look, there were two counts of the votes in Mexico. One is the preliminary count, which they put out on election night; the other is the real count, which has taken place over the past two days. In the real count, 99.5% of precincts have been counted and they show the same 200,000-vote lead they found in the preliminary count. And no one challenged the preliminary count — no one said it was wrong — so I expect the final count to mirror that.

Do you think the results of the exit poll will reduce chances of fraud in the recount?

This election commission is so squeaky clean, I don’t anticipate fraud. There were international observers there who said this election was as clean as anything they had seen. The way the election commission works, is that it represents all the parties — not just the major ones — equally. They all have an equal voice, and they all got to inspect every last count that took place, at every polling place, at the 300 deputy districts where the votes were collected, and again at the national level. I don’t see a whole lot of room for fraud.

So you think there is less chance for fraud in a Mexican election than in an American election?

I would think the Mexican system with its strong election commission that is uniform across the country would be better than anything we are doing in the U.S. One of the problems with the U.S. is we don’t have uniformity from county to county and state to state. Every county and state is making their own rules and they aren’t making them consistently.

Let me mention one other thing: In this election there were absolutely no leaks in the results from our exit polls by either Telavisa or Azteca. The results were bottled up completely; they did not get into the public dialogue. And the reason for that is we didn’t circulate the results to anyone at Telavisa other than two senior VPs of the place. We had no trouble keeping these results out of the public dialogue.

Do you think you can do that in the 2008 presidential election in the US?

Well if the networks want to let us control it of course we can do that. But not if someone at the networks leaks it. If they don’t, it will stay bottled up. They signal every intention that they will do that, and I expect them to, but we have to see how it works out.

So I take it from this – that your view is that not only do the Mexicans run better elections than the Americans, but that the Mexican political class is more responsible than the American political class?

I think they were very responsible in this election and certainly more than the people who had access to the information in 2004. I mean the leaking in this country is an abomination. They should all know better.

What issues or candidate qualities were most important in determining whether voters backed either Calderón or López Obrador?

It had to do with the constituencies, not so much the issues. López Obrador’s constituencies were the poor, the intellectuals and a lot of former supporters of the PRI. Calderón’s supporters were some PRI supporters, but mostly the business community, the middle class and the interests looking for more stability in the country.

Did Fox’s record help or hurt Calderón?

I don’t think it had much effect one way or the other.

If the PRI had not been in the election, if there wasn’t that third candidate, where would those votes have gone?

Remember the PRI is a centrist party and got a lot of its strength from the poor, the uneducated, the rural people. I would suspect that’s the group that diminished over the years and I think that’s why the PRI lost in the 2000 election. The PRI votes in the current election probably would have gone more to López Obrador, that would be my guess, but not by a lot. As a centrist party the PRI also had a lot of business interests supporting it as well, except most of the business interests left after the last election and went to Fox and they stayed with Calderón. But this time I think López Obrador would have picked up more if [PRI candidate] Robert Madrazo had not run.

Do you think the leftist sentiment we have seen in Bolivia and Venezuela was a component of the support for López Obrador? Or do you think that this Mexican election didn’t have much to do with broader trends in Latin America?

I don’t think this had to do with the broader trends in Latin America. Mexico has had a left-leaning party for some time now. López Obrador was a very strong mayor of Mexico City and he built on that. Between the collapse of the PRI which had a leftist element to it and the presence in past elections of the PRD [López Obrador’s party], that tradition had already been in Mexico.


About the Interview

Warren Mitofsky is President of Mitofsky International a survey research company he founded in 1993. Its primary business is conducting exit polls for major elections around the world. It does this work exclusively for news organizations. Mitofsky has directed exit polls and quick counts since 1967 for almost 3,000 electoral contests. He has the distinction of conducting the first national presidential exit polls in the United States, Russia, Mexico and the Philippines.

Mitofsky started and directed the first network election pool, Voter Research & Surveys, from 1990 to 1993, later to become known as Voter News Service (VNS). Mitofsky and Edison Media Research have recently conducted exit polls in D.C., New Jersey, New York and for the 2003 California recall election. With the dissolution of VNS in 2002, the election consortium has chosen Edison and Mitofsky International to be the sole provider of Exit Polls for all Primaries and General Elections.

Mitofsky created the Exit Poll research model and its execution in 1967 at CBS News; he was director of its election unit for the next 27 years and a founder of the CBS/New York Times Poll. Mitofsky is a member of the American Association for Public Opinion Research and a fellow of the American Statistical Association. He is currently working on a book about exit polls.