Mexican state and local elections 2009

[jaichind]

A poll by El Universal 5/29-6/1 has

Has
PRI       43.3%
PAN      35.8%
PRD      14.4%
Others   6.4%

[jaichind]

Compos exit poll has PRI 40%, PAN 30%, and PRD 15%.  If so this will be a landslide for PRI.

[jaichind]

Won't last

Which Mexican party do you like, if any, ag?

PRI? just a wild guess

I am pretty sure he is pro-PAN.

[jaichind]

If it is PRI 40%, PAN 30%, and PRD 15% then PRI could be about get absolute majority on its own.  It would be the best PRI performance since 1994.

[jaichind]

Right.  I forgot about that rule.  Since the PRI and PREM has an alliance I wonder if the exit poll which has PRI at 40% was 40% for PRI or 40% for PRI/PREM alliance?  If the latter and that ends up being the result then it would be a result similar to 2003.

If it is PRI 40%, PAN 30%, and PRD 15% then PRI could be about get absolute majority on its own.  It would be the best PRI performance since 1994.

40% won't give them majority on its own. By law, no party can have the share of seats greater than their voting share +8%. But it would give them a majority w/ PVEM.

[jaichind]

Won't last

Which Mexican party do you like, if any, ag?

PRI? just a wild guess

I am pretty sure he is pro-PAN.

What makes you sure?

Just a hunch.  More like process of elimination.  I can tell you are not pro-PRI nor pro-PRD.  So process of elimination seems to indicate PAN.

[jaichind]

Are PT and Convergencia  running in an alliance with PRD this time?

[jaichind]

It appears another exit poll confirmed the PRI 40% versus PAN 29% lead for PRI.

[jaichind]

11% of votes counted.  PRI 35%, PAN 27%, PRD 12%

[jaichind]

IFE reports that 7% of votes are for NULL.  Turnout is around 43%

[jaichind]

IFE reports that 7% of votes are for NULL.  Turnout is around 43%

Yep. That was expected: there was a big campaign for it. But, of course, since those will be redistributed accross all parties, this means all of them are, actually, above their current totals.

I heard of this.  There is a movenment to put a big X across the ballot.  I guess up to 7% of the votes are like that.  This plus the low turnout seems to indicate disappointment with the PAN regime and distrust of the PRD and PRI as alternatives.

[jaichind]

With about 33% of votes counted it seems to be PRI 35%, PAN 26%, PRD 12%.  Of course one has to normalized them with the around 7% NULL vote.

[jaichind]

With about 33% of votes counted it seems to be PRI 35%, PAN 26%, PRD 12%.  Of course one has to normalized them with the around 7% NULL vote.

Keep in mind: PRI is hitting the 8% constraint here.

Yep.  This is why I think that given the PRI will win a majority of the 300 FPTP seats its main constraint will be the vote share it gets.  Ergo it is more important to track the relative vote shares of the main parties.  It would seem that 35+26+12=73.  Either PT C PVEM PANAL PASC all did very well or the NULL vote is actually greater than 7%.

[jaichind]

40% counted.  PRI 35, PAN 27, PRD 12.

[jaichind]

It would seem that 35+26+12=73.  Either PT C PVEM PANAL PASC all did very well or the NULL vote is actually greater than 7%.

Yes, they did do well

Here is the current distribution:

PRI alone 35.36%
PVEM alone 7.13%
PRI/PVEM joint 0.42%
(note, that together they have 42%, which will imply that they will be right around 50% of the seats, as PRI gets to the 8% bound)

PAN alone 26.84%

PRD alone 12.27%

PT alone 4.00%
Convergencia alone 2.44%
PT/Convergencia joint 0.27%

Panal alone 3.58%

PSD alone 1.20%

Null votes: 6.20%
write-ins 0.23%

Hmm, should we not normalize the PRI/PVEM 42% by the 6.2% Null vote which would be around 44.5% of the valid vote.  Or does the 8% rule apply before throwing out the null votes?

[jaichind]

It would seem that 35+26+12=73.  Either PT C PVEM PANAL PASC all did very well or the NULL vote is actually greater than 7%.

Yes, they did do well

Here is the current distribution:

PRI alone 35.36%
PVEM alone 7.13%
PRI/PVEM joint 0.42%
(note, that together they have 42%, which will imply that they will be right around 50% of the seats, as PRI gets to the 8% bound)

PAN alone 26.84%

PRD alone 12.27%

PT alone 4.00%
Convergencia alone 2.44%
PT/Convergencia joint 0.27%

Panal alone 3.58%

PSD alone 1.20%

Null votes: 6.20%
write-ins 0.23%

Hmm, should we not normalize the PRI/PVEM 42% by the 6.2% Null vote which would be around 44.5% of the valid vote.  Or does the 8% rule apply before throwing out the null votes?

It applies separately by party - no constraint on PVEM, that's 1.  And, I believe, the null votes do not count for anything. So, yes, together they have 50% of the seats, most likely. But their alliance is not iron-clad, of course: PVEM will try to sell itself for a good price.

I would imagin it would not matter that much.  I would think that with an eye on the 2012 election, the PRI faction leaders will try to cooperate with the PAN administration to show they can be a constructive oppisition.  Hence I think the 2006-2009 mode of general PRI-PAN cooperation will continue but will wane as 2012 approaches when the PRI will look to differenciate it from the PAN government.

[jaichind]

So, here are the direct seats by state (started it before, never finished, of course still preliminary):

Aguascalientes: PAN 2, PRI 1
BC: PAN 8
BCS: PRD 2
Campeche: PRI 2
Coahuila: PRI 7
Colima: PRI 2
Chiapas: PAN 5, PRI/PVEM 4, PRD 3
Chihuahua: PRI 7, PAN 2
DF: PRD 16, PAN 7, PT/Convergencia 3, PRI/PVEM 1
Durango: PRI 4
Guanajuato: PAN 11, PRD 2 (!), PRI/PVEM 1
Guerrero: PRI 6, PRI/PVEM 2, PRD 1
Hidalgo: PRI 6, PRI/PVEM 1
Jalisco: PRI 8, PRI/PVEM 3, PAN 8
Mexico: PRI/PVEM 30 (!), PRI 8, PAN 2
Michoacan: PRD 7, PAN 4, PRI 1
Morelos: PRI 5
Nayarit: PRD 2, PAN 1
Nuevo Leon: PAN 7, PRI 5
Oaxaca: PRI 11
Puebla: PRI 13, PRI/PVEM 1, PAN 2
Queretaro: PAN 2, PRI 2
Quintana Roo: PRI/PVEM 2, PRI 1
San Luis Potosi: PAN 5, PRI 2
Sinaloa :PRI 6, PAN 2
Sonora: PAN 4, PRI 3
Tabasco: PRI 4, PRD 2
Tamaulipas: PRI 8
Tlaxcala: PAN 3
Veracruz: PRI 12, PAN 8, Panal 1
Yucatan: PRI/PVEM 5
Zacatecas: PRD 4

Wow, PAN ahead in 5 in Chiapas?  The Oaxaca PRI clean sweep looks fishy but who knows.

[jaichind]

59% of votes counted.  Still PRI 35 PAN 26 PRD 12.

[jaichind]

Prelim total seat count seems to be

PRI 209
PRI-PVEM 51
PVEM 15
PAN 133
PRD 67
Convergencia 5
PT 8
Convergencia-PT 4
Nueva Allianza 8

The PRI/PVEM alliance to get 275 out of 500
This is the worst result for PAN since 1997 and best result for PRI since 1994.
For the first time since 1994 an alliance (if you can call PRI/PVEM that) will have a majority in the Lower House.  Of course one can argue that in 1997 had the PRI and PVEM had an alliance they could have done just as well if not better than 2009, but back in 1997 PVEM was more likely to have an alliance with PAN, which it did in 2000, than PRI.
As for PAN, at least with 133 seats it can, barely, veto initiatives.

[jaichind]

Votes counted   96.32%
Turnout              44.7%

Vote share
PRI           36.59%
PVEM         6.55%
PRI-PVEM  0.42%

PAN         27.90%

PRD         12.22%

PT             3.58%
C               2.38%
PT-C          0.24%

PANAL       3.43%

PSD          1.03%  -> I guess it will be disbanded?

Write-Ins 0.18%

Null           5.42%

[jaichind]

Prelim total seat count seems to be

PRI 209
PRI-PVEM 51
PVEM 15
PAN 133
PRD 67
Convergencia 5
PT 8
Convergencia-PT 4
Nueva Allianza 8

The PRI/PVEM alliance to get 275 out of 500
This is the worst result for PAN since 1997 and best result for PRI since 1994.
For the first time since 1994 an alliance (if you can call PRI/PVEM that) will have a majority in the Lower House.  Of course one can argue that in 1997 had the PRI and PVEM had an alliance they could have done just as well if not better than 2009, but back in 1997 PVEM was more likely to have an alliance with PAN, which it did in 2000, than PRI.
As for PAN, at least with 133 seats it can, barely, veto initiatives.

You seem to be miscalculating smthg. I believe, you double-count the PRI seats that it got as part of the PRI/PVEM alliance or misapply the 8% bound

Reforma has the following seat count:

PRI 233 (+127)
PAN 146 (-60)
PRD 72 (-58)
PVEM 22 (+5)
Others 27 (-17)

PAN's total is (slightly) worse than in 2003 - but only slightly, by 3 seats. However, its voting share now is actually 4% higher than back then. You could eve argue, it's PAN's third highest vote total ever (I know: third out of 5 "honest" elections; and 4th, if you look at the number of seats rather than the vote share). 2003 was, arguably, not that much better for PAN, and, of course, the came back just 3 years later. The real disaster is on the left - even compared to the previous midterms in 2003 PRD lost about 25 seats. PRI, of course, are the clear winner.

Must be some mistake of my.  But I got my numbers from El Universal
http://www.eluniversal.com.mx/elecciones/congreso.html

[jaichind]

According to the research of Alejandro Schtulmann, head of EMPRA research, PRI will get as many as 235, PAN as many as 153, PRD as many as 75 and PVEM as many as 25.  Of course these numbers represents a cap.  But 235+25=260 does seem to indicate that the El Univesal numbers are most likely wrong. 

[jaichind]

According to the research of Alejandro Schtulmann, head of EMPRA research, PRI will get as many as 235, PAN as many as 153, PRD as many as 75 and PVEM as many as 25.  Of course these numbers represents a cap.  But 235+25=260 does seem to indicate that the El Univesal numbers are most likely wrong. 

They are, definitely, wrong. PRI/PVEM candidates, overwhelmingly, are priistas (and recorded as such). The total vote share of PRI (even w/ its share of the truly joint votes) is under 37%. It's capped at 45% of the total. There arer some variations on the rule (it may not not get directly won seats, and there are some apportionment rules accros circumscriptions), but there is no way it is allowed a simple majority on its own.

To be fair, as I pointed out before 37% translates to 39.1% once we normalized for the 5.4% null vote.  But I agree the numbers even then will not add up to get the results that El Universal came up with.

[jaichind]

The PR seat results won't be there for quite a while. First, they have to do the official count - that is on Wednesday. Then IFE will do some majic and produce the numbers. Then parties will go to court Tha PR laws aren't very straightforward, and things depend on who belongs to what. But by tomorrow the papers should have decent estimates.

Wow.  This is messed up.  I do have a couple of questions on how this works.  It seems that there is a vote in the FPTP in the 300 discticts.  It also seems that parties can have alliances where someone can vote for a joint party line.  If so is it where the two parties (assuming two parties in alliance) jointly present a candidate and voters can vote for them under 3 different party lines (party A, Party B and Party A/B alliance.)  If so it seems similar to how New York State elections are run.  It also seems the way they figure out who and which party wins in cases like this is to split the votes for joint party line 50/50 between the two allied parties.  Also is there a seperate vote that the voter votes for a party list Circunscripciones?  If so I could not seem to be able to find those numbers.  I assume that the 8% rule is based on the Circunscripciones vote share and not the FPTP vote share (if there are two seperate ballots.)

[jaichind]

The PR seat results won't be there for quite a while. First, they have to do the official count - that is on Wednesday. Then IFE will do some majic and produce the numbers. Then parties will go to court Tha PR laws aren't very straightforward, and things depend on who belongs to what. But by tomorrow the papers should have decent estimates.

Wow.  This is messed up.  I do have a couple of questions on how this works.  It seems that there is a vote in the FPTP in the 300 discticts.  It also seems that parties can have alliances where someone can vote for a joint party line.  If so is it where the two parties (assuming two parties in alliance) jointly present a candidate and voters can vote for them under 3 different party lines (party A, Party B and Party A/B alliance.)  If so it seems similar to how New York State elections are run.  It also seems the way they figure out who and which party wins in cases like this is to split the votes for joint party line 50/50 between the two allied parties.  Also is there a seperate vote that the voter votes for a party list Circunscripciones?  If so I could not seem to be able to find those numbers.  I assume that the 8% rule is based on the Circunscripciones vote share and not the FPTP vote share (if there are two seperate ballots.)

No. That's the point: we are not in Germany There is just one federal ballot, and the votes are counted twice. If your favorite candidate has been nominated by multiple parties you can either vote for just one of the parties, or for all the parties at once (in which case your vote is split between parties for the PR purposes). But you cannot vote for a PAN candidate in the district and for a PRI candidate on PR.

Actually this is similar to the way Republic of China ran its legislative elections 1992-2004.  There is one vote which you vote for one candiate in a disctrict which can have multiple winners.  The party of the candidate you voted for also gets incremented by one toward a party list vote.  But they did not have these complicated rules of having alliances and the 8% cap.  This system got abolished for something similar to Germany in the last (2008) legislative elections.  Main reason was to take into account people that supported the party candidate in a district but wanted to back another party in the party list.