Here's a link to my U.S. News column on the Mexican election, in which I try to put it in a Latin American and in a North American perspective. Bottom line: Not all of Latin America is turning left, and all three countries covered by the North American Free Trade Agreement now have and will have for at least the next couple of years center-right governments. That last conclusion assumes that Stephen Harper's Conservative government in Canada, which currently does better in polls than it did when it won a plurality of parliamentary seats in the January election, will stay in office for some good period of time.
Let's start off by comparing Mexico's vote in 2006 with its vote in 2000, when PAN's Vicente Fox won with a 43-to-36 percent victory over PRI's Francisco Labastida, with 17 percent for PRD's Cuauhtémoc Cárdenas. Obviously, this represents a gain for PRD and a loss for PAN and PRI. PRD rose from 17 to 35 percent from 2000 to 2006, PRI fell from 36 to 22 percent, and PAN fell from 43 to 36 percent. But Mexico's politics, like that in the United States, can be candidate-centric. Cárdenas, mayor of Mexico City from 1994 to 1997, was generally felt to have done a bad job; Andrés Manuel López Obrador, mayor from 2000 to 2005, was felt to have done a good job. So López Obrador was a much stronger candidate. In contrast, Felipe Calderón was little known as late as early 2006, while Fox, as the governor of Guanajuato, was a leading national figure going into the election of 2000. Moreover, Fox wasn't associated with the heavily pro-Roman Catholic Church, pro-business founders of PAN; Calderón's father was one of the fathers of PAN. Fox and López Obrador both ran well ahead of their parties' congressional candidates. Calderón ran only slightly ahead of his party's candidates, while 2006 PRI candidate Roberto Madrazo ran well behind his.
Remember that the dynamic in the two elections was different. In 2000, the dynamic was PRI/anti-PRI. Fox was clearly the strongest alternative to PRI, and many votes evidently moved toward him near the end. In 2006, the dynamic was AMLO/anti-AMLO, with Calderón clearly the strongest alternative to AMLO, and at least a few votes seemed to have moved to anti-AMLO in the end.
When you look at state results, you get an interesting picture. (I'm using the 2006 numbers that were on the PREP website Sunday night, which for these purposes will not differ substantially from the final figures.) Calderón runs ahead of Fox (all these comparisons are in percentage terms) in four northern states (Aguascalientes, Durango, San Luis Potosí, and Labastida's home state of Sinaloa) and less than 3 percentage points behind in others (Nuevo León, Sonora, Guanajuato Fox's home state Baja California Sur, Baja California, Chihuahua, the traditionally leftist state of Zacatecas, and Querétaro. He also ran ahead of Fox in his own home state, usually heavily pro-PRD Michoacán. Calderón ran far behind Fox in Distrito Federal and Mexicomore evidence of López Obrador's perceived strong performance as mayorand in Tabasco (the home state of both López Obrador and Madrazo), Chiapas (where the Zapatista movement is strongest), and Quintana Roo (Cancún).
Now let's take a look at the regions that I used in an earlier blog posting. For both elections, I'll list the parties in order of number of votes, which means PAN-PRD-PRI for 2006 and PAN-PRI-PRD for 2000. I think this reflects the different dynamics of the two elections.
Region PAN '06 PRD '06 PRD '06 PAN '00 PRI '00 PRD '00 PAN PRD PRI
MEXICO 36 35 22 43 36 17 -7 +18 -14
Border 46 22 25 49 39 9 -3 +13 -14
North 48 22 23 48 39 9 0 +14 -17
Center 36 34 24 41 39 16 -5 +18 -15
Mexico City 30 49 15 43 30 22 -13 +27 -15
South 20 46 28 26 39 31 -6 +15 -11
Yucatán 39 25 30 45 42 9 -6 +16 -12
Collapsing these into three regions, we get:
North 47 22 24 48 39 9 -1 +13 -15
Mexico City 30 49 15 43 30 22 -13 +27 -15
South 29 39 26 34 39 22 -5 +17 -13
So the biggest movement toward PRD and away from PAN was in Mexico City. López Obrador turned out to be a vote maximizer for PRD because he ran ahead of party in a region that cast about one third of the entire nation's votes. The PAN vote declined in the South but remained steady in the North, and PAN's margin in the North grew larger, because PRI as the traditional chief opponent of PAN there (as it is also in contemporary governor elections) did not disappear.
Let's compare PAN's percentage margin by region over its chief opposition parties (PRI in 2000 and PRD in 2006).
MEXICO +1 +7 -6
Border +24 +10 +14
North +26 +9 +17
Center +2 +2 0
Mexico City -19 +13 -32
South -26 -13 -13
Yucatán +14 +3 +11
Clearly, the partial collapse of PRI helped Calderón increase his margins in the North and in the Yucatán, enough to survive López Obrador's impressive margin in Mexico City and the gains in margin made by López Obrador by the partial collapse of PRI in the South.
From the national returns, you can link to returns from each state, their electoral divisions, and finally to every individual precinct. (It's not so easy in the United States; here's a website I use for the same purpose, but in most states it doesn't go down to the precinct level.) Here are some of the things I found from looking through the pages.
First, there are relatively few differences within each of the states. Perhaps this is because Mexico's division of states reflects ethnic origin: The small Tlaxcala area not far from Mexico City is a separate state, and those who remember accounts of Cortés will remember that the Tlaxcalans fought on his side against Montezuma and the Aztecs (more accurately, the Mexica). So all three divisions of Aguascalientes, for instance, voted for PAN, and all nine in Guerrero for PRD. I found a few splits in Campeche (the city narrowly for PAN, the countryside a little bit more for PRD), Coahuila (one of eight divisions for PRD), Chiapas (two divisions for PRI in a state where it got 95 percent of the recorded vote in 1988), Distrito Federal (PAN carried the Benito Juárez division, which includes the rich areas of Polanco and Lomas de Chapultepec, which from inspection of actual precincts seem to have voted 70 percent or more for Calderón), Michoacán (in his home state, Calderón carried four of 12 divisions), Morelos (Cuernavaca narrowly for PAN, the rest solidly PRD), Nuevo León (one of 12 divisions, part of Monterrey, for PRD), and Sinaloa (one of eight divisions for PRD). The only states with large numbers of different results were the state of Mexico, the largest in the country, and the closely divided states of Puebla and Veracruz east of Mexico City.
To me this undermines at least a little the picture most of us have of this as being a contest between the rich and the poor. That's underlined by the Reforma exit poll, which I referenced in an earlier post. The lowest-income voters were actually about evenly split, the highest-income definitely for Calderón but by a less than unanimous 50-to-30 percent margin. Only when you get to the top of the scale (Reforma's highest income category starts at just over $10,000 a year, not enough to get you anywhere close to affording an apartment in Polanco or a house in Lomas) does Calderón's margin increase more. Are there no large poor neighborhoods in Jalisco or its largest city, Guadalajara? I should think so, but every division there voted PAN. Are there no well-off people in Guerrero? I should think so, but the Acapulco division voted PRD.
I would look at ethnic factors as well. Aficionados of Mexican cuisine know that it actually consists of many regional cuisines: Yucatán, Veracruzan, Oaxacan, etc. You can find restaurants specializing in each of these regional cuisines in Mexico City. The regional differences in Mexico are persistent and at least as distinctive as in the United States, Britain, Italy, Spain, and Brazil. I have already mentioned my suspicion that Mexico's often convoluted state boundaries were drawn to put together people of similar ethnic varieties. People still speak Maya in the state of Yucatán (which voted heavily PAN), Zapotec in Oaxaca (which voted heavily PRD), and Mixtec in parts of Puebla (which was about evenly split). (For the persistence of Indian languages in Mexico, see this book; interestingly, it was encouraged by the Roman Catholic Church.) These differences were not reflected in election outcomes when the process was controlled by PRI and results were well nigh unanimous, from 1929 to 1988. Now we're seeing them in the results of 2000 and 2006. We norteamericanos tend to think that all Mexicans are pretty much the same, that they all eat Tex-Mex food, and that they all have the left-leaning politics. Well, they aren't, they don't, and they don't.