UPDATE 5 July: Calderon’s lead is now down to about 0.6%, as the PRD turned out to be right in its claim that three million (actually about 2.6 million) votes had been improperly excluded from the preliminary results. Remember, it is all preliminary!!

UPDATE: Rici has some information in a comment below about the preliminary seat breakdown. My own information is very close to his: PAN 209 (of 500) Chamber seats, 52 (of 128) Senate seats. PRI-PVEM 119 and 38, PRD+PT+Conv 159 and 36. So the PAN alone will have over 1/3 of the seats in each house–in fact, it will have over 40%, despite under 35% of the vote. I suspect it has won the presidency, too, but that’s not yet certain.

Please note, this is all preliminary. A final count will not be known till at least Wednesday. But the IFE’s “PREP” (via the mirror site at Universal) shows the following as of around 8:00 a.m., Pacific Time today.

President, with over 96% of actas processed:

Calderon (PAN), 36.4%
López Obrador (PRD), 35.4
Madrazo (PRI), 21.5

There appears to have been a lot of ticket-splitting, and the current PAN vote shares for each of the congressional chambers rests at under 34%, while the PRI is around 28% and the PRD around 29%.

Nonetheless, a correspondent who has looked at the district-level results tells me that the PAN has won substantial pluralities in both chambers. The PAN has emerged as the most national of the parties, and the electoral systems of both chambers, while having a proportional component, are quite favorable to any party that has good regional spread.

The “parallel” feature of the lower-chamber mixed-member system means that a party that can win a lot of the single-seat districts (SSDs) will be substantially over-represented. The PAN may have won close to half the 300 SSDs, and will be close to the cap (whereby no party is allowed to add so many seats off the party lists that its seat share of the whole chamber is greater than its nationwide votes share, plus eight points).

In the Senate, the “limited seats” feature, whereby the plurality party in each state gets two seats and the runner-up gets one (plus a parallel nationwide PR allocation of 32 seats) also greatly favors the party that has the greatest regional distribution of strength.

These features of the system used to benefit the PRI. Now, with that party in third place, it is the PAN that is benefited by the plurality features of the system for congress. The PRD is somewhat over-concentrated (dominating the capital), and while this concentration has no negative impact on it for presidency (single nationwide plurality contest), it means that even if the PRD wins the presidency (still very possible), it will face a congress with up to 40% held by its main opponent and perhaps well under a third in its own hands.

The PREP results will be taken down Monday afternoon, and then we will have to wait for the full final count. This a real test for IFE, as the PRI may have had some opportunities in rural areas to pad the count for its congressional candidates. Not all polling places are monitored by the opposition. Can the PRI pressure poll-workers? I do not know. But it is not out of the question. Presumably the PRI, if it could pull it off, would prefer the more divided government under AMLO to Calderón and a strong PAN plurality. [UPDATE: Apparently the number not monitored is very small, so this concern expressed in this paragraph may not be so valid.]

I think Mexico’s electoral institutions are up to the challenge. But this is a big test.