The gaming of Mexico’s PR disparity cap

One distinctive feature of Mexico’s 500-seat Chamber of Deputies’ mixed-member majoritarian system – under which 300 members are chosen by plurality voting in as many congressional districts, while the remaining 200 seats are filled by party-list proportional representation (initially determined on a nationwide basis) – is a provision which caps party representation at a maximum of eight percentage points above its national proportion of votes cast for parties entitled to participate in the distribution of PR mandates (in addition, no party may have a total number of Chamber seats in excess of 300 mandates). However, in recent years some parties running in electoral coalitions have successfully gamed the system to circumvent the limits imposed by this provision, effectively diluting its intended compensatory effects.

The eight percent disparity cap is implemented by determining the maximum number of Chamber seats each party may obtain; if a party exceeds this amount, its PR seat allocation is reduced so that its total number of plurality and list seats is equal to the capped seat total. For example, in the 1997 legislative election – generally regarded as Mexico’s first truly free and fair vote – the then-ruling Institutional Revolutionary Party (PRI) won 38% of the total vote and 39.98% of the “effective national vote” (“votación nacional efectiva” in Spanish), that is the total number of votes cast for parties entitled to take part in the allocation of PR seats. Therefore, PRI could obtain at most 47.98% (39.98% plus 8%) of the Chamber’s 500 seats, or 239 seats, disregarding fractions. Since PRI had won 165 of 300 plurality mandates in the election, and had been initially assigned 80 PR seats, for a total of 245 seats – six above the capped total – the party’s list mandates were reduced to 74 seats (239 minus 165); the remaining 126 PR mandates were then distributed among the other four qualifying parties.

Prior to 2008, Mexican voters could choose either a party running alone or a coalition, but not a specific party running within a coalition. Therefore, coalitions were treated as parties when it came to allocating Chamber seats and determining the eight percent cap. However, a 2008 electoral reform allowed voters to choose parties running as part of a coalition, and at the same time limited coalitions to plurality elections only. Since then, the distribution of Chamber PR seats has been carried out entirely on an individual party basis, even though coalitions have often continued to determine the outcome of plurality races; most importantly, the allocation of plurality seats among coalition partners has been determined on the basis of officially registered coalition agreements (“convenios de coalición” in Spanish). In due course, larger parties running in coalition with smaller allies discovered they could game the cited provision by assigning plurality seats in the coalition agreement to their junior partners, while these in turn nominated candidates who were actually members of the larger party. In this manner, the larger party artificially reduced its total number of seats, thus making it less likely that it would hit the eight percent cap during the allocation of PR list mandates.

Before 2018, PRI and its ally the Ecologist Green Party of Mexico (PVEM) gamed the system in the manner described, but the scope of their scheme was comparatively limited, not least because PVEM didn’t have a particularly large number of Chamber plurality seats to begin with. Moreover, in 2015 Mexico’s Federal Judiciary Electoral Tribunal (TEPJF) ruled that parties could nominate candidates from a different political party, provided the parties had a coalition agreement. On the other hand, the stratagem pursued in 2018 by the “Juntos Haremos Historia” (JHH; “Together We’ll Make History”) coalition of the National Regeneration Movement (MORENA), the Workers’ Party (PT) and the Social Encounter Party (PES) was anything but limited. For the Chamber of Deputies, the three parties ran together in 292 of Mexico’s 300 congressional districts, but in the coalition agreement 142 nominations were assigned to MORENA, and 75 each to PT and PES.

To be certain, it is not unusual for smaller parties like PT and PES to have a disproportionate share of single-member nominations in mixed-member majoritarian systems which allow party coalitions: in a nutshell, larger parties often bow to smaller parties’ demands, outrageous as they may be, if that is what it takes to avoid the risk of losing elections. That said, giving away half the Chamber’s single-member seat nominations to a couple of very small parties came across as both extreme and highly suspicious, all the more so considering that for the Mexican Senate plurality races, JHH nominated 49 candidates from MORENA, but only eight from PES and five from PT. However, unlike in the Chamber of Deputies, there is no eight percent cap in place for the distribution of Senate PR seats, and therefore MORENA gained nothing from giving away upper house plurality nominations to its much smaller allies.

JHH went on to score a landslide victory in the general election last July, securing an absolute majority for its presidential candidate Andrés Manuel López Obrador (better known by his initials AMLO), as well as a sizable plurality in the legislative election. In the Chamber of Deputies, final results show JHH won 220 of 300 plurality seats (including eight mandates won by MORENA running alone), with the following results for its constituent parties:

Party % (Votes) Seats % (Seats)
MORENA 37.2 106 35.3
PT 3.9 58 19.3
PES 2.4 56 18.7
JHH Total 43.5 220 73.3

The sole beneficiaries of the victorious coalition’s over-representation in the plurality component were PT and PES, as MORENA ended up slightly under-represented. Again, while it is not unusual for smaller parties within larger coalitions in mixed-member majoritarian systems to have a share of seats in excess of their percentage of the vote (even without considering the PR component), this was nonetheless a truly extreme outcome. By comparison, the Senate election had a more typical plurality outcome, as shown below:

Party % (Votes) Seats % (Seats)
MORENA 37.4 42 43.8
PT 3.8 5 5.2
PES 2.3 8 8.3
JHH Total 43.6 55 57.3

(JHH’s smaller majority of upper house plurality seats was due to the fact that unlike in the Chamber of Deputies, Senate plurality mandates are allocated on the basis of two seats for the winning ticket and one for the first runner-up in each one of Mexico’s thirty-two federal entities).

For the distribution of Chamber PR seats, MORENA received 41.34% of the effective vote, so it could have no more than 49.34% of the 500 Chamber seats, which translated to 246 seats. However, after the initial allocation of PR mandates MORENA had a total of 188 seats (106 plurality and 82 list), well below its corresponding maximum seat total. On the other hand, PT, with 4.36% of the effective vote, would have been entitled to nine list seats, but unexpectedly exceeded its 12.36% cap, or 61 seats due to the party’s large number of plurality victories; PES fell below the three percent valid vote minimum required to take part in the distribution of PR seats. Consequently, PT’s list seat total was reduced to three mandates, and the remaining 197 PR seats were distributed among the other six qualifying parties. In all, JHH won 308 of 500 Chamber seats, as detailed below:

Party FPTP PR Total % (Total)
MORENA 106 85 191 38.2
PT 58 3 61 12.2
PES 56 0 56 11.2
JHH Total 220 88 308 61.6

Mexico’s National Electoral Institute’s General Council Agreement INE/CG1181/2018 [PDF] details the official distribution of Chamber of Deputies PR list mandates in Mexico’s 2018 federal election, while Federal Elections in Mexico has nationwide- and federal entity-level results of elections in Mexico since 1997.

Although the eight percent cap deprived PT of just six seats, this left several of its leaders without a seat in the Chamber of Deputies, and the party went to court to have the result overturned. To that end, PT actually admitted some of its successful plurality candidates weren’t really party members after all, and sought a judicial review of JHH’s officially registered coalition agreement. However, the court rejected what it correctly perceived as a transparent attempt by PT to re-structure the previously agreed distribution of Chamber plurality nominations, in order to bypass the eight percent cap and increase its number of list mandates. In essence, TEPJF ruled JHH’s coalition agreement was a done deal which could no longer be altered after the election had been carried out with the agreement in place, and for good measure reaffirmed its earlier 2015 ruling that parties could nominate candidates from a different political party, provided the parties had a coalition agreement.

Just as important, TEPJF ruled that under the prevailing legal and constitutional arrangements, it was not possible to consider JHH’s coalition performance as if it were a single party for the purposes of determining over-representation, as it would run counter to the requirement such determination must be carried out on an individual party basis. Had Chamber seats been allocated among coalitions (and had all three major coalitions been in place in every congressional district, as was the case with the presidential election), JHH’s seat total would have been limited to 269 mandates – a significantly reduced yet nonetheless substantial majority.

TEPJF also confirmed that PT exceeded the eight percent cap by six seats (because 12.36% of 500 seats equals 61.8 seats, it was argued the party went above the cap by 5.2 seats (67 minus 61.8), which should have rounded down to five seats; however, this argument was rejected on the grounds that with 62 seats of 500 – 12.4% of the total – PT would have had a disparity gap of 8.04%, a figure just above the eight percent maximum).  The court ruled as well that JHH’s plurality victories could not be assigned to MORENA on account of the fact that the party contributed the vast majority of votes to that end.

(MORENA outpolled both PT and PES in every congressional district, and even if the party had run alone it would have secured a large majority of plurality seats, while PT and PES would have won none. However, in such hypothetical scenarios – which assume an identical distribution of votes among the parties contesting the 2018 federal election – MORENA’s Chamber of Deputies seat total would have been limited to 246 mandates by the eight percent cap, while PT would have received no more than 13 list mandates and PES would have had no seats, leaving MORENA and PT with a relatively slender joint majority in the Chamber.)

Nevertheless, TEPJF’s rulings on some of these controversies were not unanimous. Judge Felipe de la Mata Pizaña wrote a concurring opinion calling for the current implementation of the eight percent cap provision to be reconsidered for future elections, while Judge Reyes Rodríguez Mondragón issued an extensive dissenting opinion, in which he argued that the actual party affiliation of plurality winners should be taken into account for determining the eight percent cap during the distribution of PR seats. Notably, the dissent cites a number of political science scholarly sources, including both the Spanish-language edition of “Presidents and the Party System” (edited by S. Mainwaring and M.S. Shugart) and “Seats and Votes” by R. Taagepera and M.S. Shugart, even carrying out the calculations to determine the effective number of parties index for Mexico.

(For further information, please refer to TEPJF’s rulings on SUP-JDC-0429-2018; SUP-JDC-0438-2018SUP-JDC-0444-2018SUP-REC-0934-2018; and SUP-REC-0943-2018 [PDF].)

However, beyond the practical difficulties entailed by implementing this approach, it is not clear that it would have resulted in a significantly different distribution of seats. After Mexico’s newly elected legislators were sworn in just over three months ago, a large number of JHH’s 220 plurality deputies ostensibly elected as members of PT or PES joined MORENA’s parliamentary group, as shown below:

Party Elected Group Gain/Loss
MORENA 106 162 +56
PT 58 28 -30
PES 56 30 -26

Had the eight percent cap been calculated on the basis of parliamentary group affiliation figures, the initial distribution of PR list seats, which as previously noted granted nine mandates to PT and 82 to MORENA would have been definitive, as neither party would have reached its eight percent-capped seat total. Moreover – and quite ironically – JHH would have obtained an even larger Chamber majority of 311 seats (of which 244 seats would have gone to MORENA, 37 to PT and 30 to PES).

In conclusion, while PT and PES contributed very little in terms of votes to JHH’s sweeping victory in the 2018 federal election, the gaming of the eight percent disparity cap through the assignment of a disproportionately large number of plurality nominations to these parties allowed the coalition to obtain a much larger majority in the Chamber of Deputies than it could have otherwise attained. Even so, the triumphant coalition parties overplayed their hand, and the stratagem backfired on PT; this in turn led to a public disclosure of the scheme’s existence (already highly suspected, not least because it was largely absent in the Senate election, which has no disparity cap provisions), as part of an unsuccessful effort to have the unexpected adverse outcome reversed in court.

Even though a majority of Mexico’s Federal Judiciary Electoral Tribunal justices have reaffirmed the validity of electoral provisions which allow outcomes like these, it seems an obvious flaw in the design of the system that parties can run in alliances, but the caps on over-representation are calculated at the individual party level, as pointed out by Dr. Matthew Shugart himself on this blog shortly after the 2018 federal election took place. Moreover, the gaming of the eight percent disparity cap in the 2018 federal election effectively altered the nature of the Chamber of Deputies electoral system, causing it to function almost like a parallel system, in which the outcome of plurality races had very little impact in the distribution of party list seats. That may not appear to be cause for excessive concern in the here and now, given that JHH would have in all likelihood secured a majority of Chamber seats even if the coalition hadn’t gamed the eight percent disparity cap, but the issue could have wider ramifications with unpredictable consequences in the years to come. 

Mexico, 2018

Mexico has its elections for President, Chamber of Deputies, and Senate on 1 July. It has been clear for a while that, barring a big surprise, Andrés Manuel López Obrador (popularly known as AMLO) will win.

AMLO’s support has risen steadily out of what looked like a tight three-way contest some months ago into a strong lead. When voters responding “no preference” are removed, it even looks likely that AMLO could win a clear majority of votes. Mexico elects its presidency via nationwide plurality, and no Mexican president has earned half the votes since 1994 (at a time when most experts still considered the regime authoritarian, albeit increasingly competitive).

Assuming AMLO wins, it will highlight the competitive three-party nature of the system. When the center-right National Action Party (PAN) won the presidency in 2000, it broke decades of continuous control by the Institutional Revolutionary Party (PRI). The PAN won again in 2006, on less than 37% of the votes in a very tight race, with AMLO close behind (and refusing to acknowledge defeat). The PRI returned to the presidency in 2012, and now AMLO will give the left its chance. (AMLO was with the Party of the Democratic Revolution, PRD, but in recent years has set up a new party, MORENA, while the remnant PRD is backing the PAN candidate this time.)

I would be very interested in seeing an analysis of AMLO’s own manifesto (and his party’s, if separate). There is much hand-wringing over his leftist “populism”. However, when he ran in 2006, he staked out a centrist economic platform well to the right of his own party–a clear case of what “presidentialization” does to parties. (See the discussion of the general point, and also the 2006 Mexican campaign, in my book with David Samuels, Presidents, Parties, and Prime Ministers). Is he doing so this time? I can’t claim to have followed closely enough to know.

As for the Chamber of Deputies, if the pattern of recent Mexican elections holds, the party winning the presidency will win fewer votes for its congressional candidates. That could mean MORENA (and pre-election allies) will not have a majority of seats. On the other hand, as noted above, these previous presidents have not themselves won majorities. Moreover, the electoral system is mixed-member (with the voter having a single vote). It is sometimes erroneously categorized as mixed-member proportional (MMP), but it is actually leans much more to the majoritarian category (MMM). Seats won based on nationwide votes for party are added to single-seat districts won (by plurality).

The allocation is not compensatory, but it is also not strictly parallel. There are caps on allowable over-representation (unlike in a “pure” MMM system). The most important cap is that no party can have a final seat percentage that is more than eight percentage points above its vote percentage. Thus if a party wins under 42% of the votes, it is unable to have a majority of seats. If it gets over 42% it is not guaranteed a majority, but a majority becomes likely, due to the non-compensatory nature of the allocation. This cap kept the PRI from retaining its majority in the midterm election of 1997, and I believe it has been hit in several subsequent elections, as well. This is what I will be watching most closely: Will MORENA (and allies) get a Deputies majority?

The Senate is also elected in a mix of regional and nationwide seats. Each state has three senators, elected by closed list, limited-nominations plurality. The largest list gets two seats and the runner up gets one. Then there are 32 seats elected by nationwide proportional representation (allocated in parallel, not compensatory manner).

These provisions, combined with the regionalization of party support in Mexico, make it difficult for a party (or alliance) to win a majority of the Senate’s 128 seats. AMLO is unlikely to have majorities in both houses, but it is worth noting that the federal budget must clear only the Chamber. There is no Senate veto on the spending side of the budget, although both houses must pass all other types of bills. Thus the left will be in a strong, but not unchecked, position to implement its program for the first time in Mexican democratic history.

Mexican elections 2015

In addition to Turkey, about which we already have an active thread, there also are elections today in Mexico.

The entire Chamber of Deputies is being elected today, along with governors in nine states. Those nine states, plus some others, have elections for their state assemblies.

This is the first midterm election under a president from the Institutional Revolutionary Party since 1997. Each of the preceding has seen the familiar “midterm decline” phenomenon, whereby the president’s party sees its seat share of the legislature reduced in an election near the mid-point of the president’s term. Will this one defy the trend?

An average of polls that I have seen* put the PRI on around a third of the vote, which is approximately what it won in 2012. The National Action Party (PAN) also was polling around the same range as it won in the last election (around 27%). The difference maker could be the left, where the Party of the Democratic Revolution suffered a split: former leader Andres Manuel Lopez Obrador set up a new party, Morena, after the 2012 election. This party will compete with the PRD (and allies) for votes from the left.

The Mexican electoral system is a mixed-member system. Its PR-list component is neither fully compensatory nor fully parallel, but it is somewhat more the latter. That is, a party that wins a lot of the 300 single-seat districts will tend to be over-represented, even after the 200 list seats are allocated.** For instance, in 2012, the alliances of the PRI (33.6% of the vote) and Greens (PVEM, with 6.4%) took 241 seats (48.2%).

Given the disproportionality of the electoral system, it is not out of the question that the PRI (and PVEM ally) could gain seats even if it loses some votes from 2012. In much of the center and south of the country, the PRI’s main competitor is the PRD. With the PRD split, the PRI could pick up districts (which are decided by plurality) that the PRD won last time, even if its votes did not increase.

The caveat is that I have not looked at district-level patterns from 2012, nor have I seen state-level (let alone district-level) polling for this election. So I do not know how plausible this scenario is. But it is something to watch as the results come in later. This is not an mixed-member proportional (MMP) system, so coordination matters. And the left does not appear well coordinated in this election, a factor that should benefit the PRI, the president’s party, more than the PAN.

_________
* Sent by a contact in Mexico.

** The basic feature is that seats are allocated in parallel (i.e. based on nationwide proportional share of the 200 list seats) up to the point at which a party has an over-representation that is more than eight percentage points. Once a party has hit that cap, any remaining list seats are allocated proportionally among parties that have not hit the cap. There is a 2% threshold. Voters have just one vote, so parties must have candidates (their own or an alliance partner’s) in the single-seat districts to collect “list” votes, which are simply aggregations of parties’s candidates’ votes. (It is a bit more complex than this, but these are the main points.)

The PRI’s stability

Aside from 2006, not much change over time. What has changed is the relative ability of one of the other parties to consolidate enough of the anti-PRI vote to win a plurality.

Year, Deps, Pres
1997, 39.1, —
2000, 36.9, 36.1
2003, 40.8, —
2006, 28.2, 22.3
2009, 43.7, —
2012, 38.0, 38.2
(PRI-PVEM from 2003)

That’s a mean of 37.8% (for Deputies).

So the answer to my question before the election–how big will the PRI’s comeback be?–is, not that big. Just a regression to its 15-year mean.

_______
Data are from the IFE website, following the links for respective years and institutions.

How big will the PRI’s comeback win be?

If you are reading this, there is a pretty high probability that you know Mexico has elections today for President, Deputies, and Senators. There seems to be no doubt that the formerly hegemonic PRI, which last ruled Mexico in 2000, will win. The question is, how big?

The presidency is elected by plurality, and so a voting result in the low/mid-40s will be good enough–as it was for Vicente Fox of the PAN in 2000. (The incumbent, Felipe Calderon, also of the PAN, won in 2006 with only around 36%.)

The Chamber of Deputies is elected by MMM. There are 300 single-seat districts elected by plurality, and 200 list seats determined by nationwide proportionality. ((Some sources mention regional districts in the PR tier. That is true, but only on the intra-party dimension: these districts, and the separate list each party presents in each district, matter only in determining which candidates take the list seats a party has won. They do not affect the total balance of seats among parties, which depends on the parties’ nationwide total votes.)) These seats are elected in parallel. That means that a party’s proportional share of the 200 is added to however many district seats it has won, with a partial exception to be addressed below. The votes cast for single-seat candidates are summed up to determine shares for list allocation, as there is no separate list vote.

In the Senate, there are three members elected from each state, and another 32 elected nationwide; again, the nationwide seats are in parallel and without a separate vote. In each state a party presents a (closed) list of two candidates. The plurality earns two, and the second party wins one.

The President and Senators are elected for six-year terms, the Deputies for three years. No one at any level can serve consecutive terms.

Will the incoming PRI president–assuming no huge surprise when the results come in–have a majority in either or both houses?

In the most recent Deputies election, in 2009, the PRI won just short of a majority: 237 seats (47.4%) on only 36.9% of the votes. The PRI was in pre-election (pri-election?) coalition with the Green Party (PVEM), and this combine actually did win a majority: 258 seats on 43.6% of the votes.

The 2009 result shows clearly that the system is MMM (parallel), not not MMP (compensatory) as some sources claim. A near-majority of seats for one party that wins less than 37% of the vote is pretty non-proportional! The one way in which the system is not purely parallel is that it includes caps on over-representation. No party may win more than 60% of the seats, or more consequentially, a seat percentage that is more than eight percentage points greater than its vote percentage.

It seems (though I am not sure) from the 2009 result that this provision is applied to a pre-election coalition, and not to such an alliance’s parties individually (–UPDATE–see Manuel’s comment): the PRI’s own over-representation was more than 10 points. What matters, apparently, is the alliance, which was almost precisely at eight percentage points over-representation. Together in the nominal (plurality) tier the parties won 188 seats (184 PRI, 4 PVEM), which is 62.67% of the districts. (Oh, doesn’t plurality produce big distortions when there are three major national parties!) The parties’ combined votes, as noted, were 43.6%. However, there is a 2% threshold, so what really matters for the proportional-list seats is the “effective vote”: when below-threshold parties’ votes are removed, the two parties had 46.73%. That would earn them around 93 of the 200 list seats. This would get them to 281 seats. So, if my calculations are correct, the cap was triggered in 2009. ((I have said previously that it was not triggered at an election after 1997. Perhaps that was incorrect. The rest of this footnote is geared even more for anoraks than the main entry.

Maybe I am wrong here, rather than in my previous comments about the cap. That is, maybe the cap on over-representation is applied to individual parties, but is triggered only by comparing seat percentages with effective votes. If that is the case, the PRI, alone, did not trigger it, as it had 39.5% of the effective vote, and its total seat percentage won was then less than 8 points greater.

Maybe someone can clarify this arcane point!))

Will the cap be triggered in 2012? If so, will it affect the PRI’s chances of winning a majority? I would think it is likely to get a majority, as unless there is a great deal of ticket-splitting, the PRI should win over 42% of the vote. Winning 42% does not guarantee a majority of seats, but makes it likely. The party would have to win a sufficient number of district seats–about 167–to ensure the majority. However, even a small increase in the votes for the largest party should result in an even greater swing of seats in the PRI’s favor than the very large swing we saw in 2009 (when the PRI alone won 184). On the other hand, winning less than 42% of the vote makes a majority impossible–not counting votes and seats of alliance partners. ((The PRI is again in alliance with the Greens.))

Conclusion on Chamber of Deputies: A majority for the PRI looks likely.

As for the Senate, a majority depends on winning the votes plurality in many states, as well as a large enough share of the total nationwide votes. Given that the PRI currently governs about two thirds of the states, it obviously has the regional spread to pull this off.

Bottom line: a two-house majority for the new PRI president, Enrique Pena Nieto, looks likely, but not a sure thing.

Now, does this mean a restoration of the “old” PRI? Probably not, as the internal lines of authority in the party have changed, probably irrevocably. But that is a topic for another day…

Mexico’s ballot format

Update: it is clear now that I misinterpreted the rule. (In my defense, the linked story presents the matter less than clearly.) See the comments for clarification.

Mexico votes Sunday for president and all members of both chambers of the federal congress.

The Chamber of Deputies election has an interesting ballot format. The Deputies are elected by a Mixed-Member Majoritarian system (with caps–more below–but it is not MMP). Unlike most mixed-member systems, the voter has only one vote. The vote for a candidate in any given single-seat district also counts for the party list; that is, there is no separate list vote.

Candidates are sometimes nominated by pre-electoral coalitions. However, the parties keep their separate ballot identities. A vote is valid even if the voter marks the ballot for two or more parties in coalition. However, such a vote would count only for the candidate, and not for any of the parties’ list. This is an unusual provision, and I am not aware offhand of anything similar elsewhere. (See earlier thread, and comment by Manuel, in which this feature was mentioned.)

Mexico: Calderon calls for legislative reelection

Mexico is one of few* countries to prohibit legislators from serving consecutive terms. This past week, President Felipe Calderon announced that he will propose that legislators be permitted to seek reelection.

Quick reaction #1: good idea, as it would give the 300 members elected from single-seat districts (200 others are elected via closed-list PR) the incentive to actually represent the electorate of their districts, rather than immediately upon election seek to curry favor with whoever may offer them their next job.

Quick reaction #2: good luck passing it. The PRI, which is currently just short of a majority in the Chamber of Deputies, is unlikely to agree to a reform that would restrict the authority of party leaders (who tend to offer members that next job referred to in quick reaction #1). Even if the party wins back the presidency,** we are unlikely to see the degree of centralization and presidentialization of the halcyon days of PRI hegemony. However, in opposition, the PRI has become a “gubernatorialized” party, and the governors presumably would stand to lose much influence if legislators could seek longer tenure in congress.

______
* One of only two? (Costa Rica is the other one I know of.)

** Conventional wisdom seems to be that the party is a shoo-in for 2012. I am not so sure. That the party can do so well in midterm congressional elections when it is in opposition (such as in July 2009), and in gubernatorial elections (it governs almost two thirds of the states), says less than meets the eye about its prospects of finding a single candidate who can unite the party and appeal broadly enough win the presidency. Much will depend on whether the PAN finds a popular enough candidate to appeal beyond its narrow base and whether the PRD can pull itself together enough to appeal to the more leftist elements of the PRI constituency. (Mexican presidents are elected by nationwide plurality, and Calderon himself won about 36% in 2006 and defeated then-PRD candidate Lopez Obrador by the narrowest of margins.)

Mexico may experience ‘divided government’ after all

Beneath this morning’s Mexico planting, Ed, and later Manuel, have provided preliminary seat counts. If the results hold, the PRI and Greens together may have a majority.

They ran in a partial pre-electoral coalition: joint candidates in 63 districts, but separate lists (and candidates elsewhere).

The coalition apparently has won in 50 districts, which is impressive.

Also noteworthy is that the Mexican Green Party, at 6.5% of the votes in this election, would suddenly find itself among the largest Green parties in the world. I did not see that coming. (I will leave it to others to decide whether a Green alliance with a party like the PRI is anything to celebrate.)

What I am unsure of is the extent to which the PRI and Greens have been cooperating in congress and whether they have anything like a commitment to cooperate in the upcoming congress. If they do, then I would be inclined to put this election into the relatively rare category of those in (pure) presidential systems that produce divided government.

Mexico: PRI will be short of majority

See the comments for some updates on the preliminary seat totals.

According to preliminary results, the Institutional Revolutionary Party (PRI) has won the most votes and seats in the midterm Chamber of Deputies election.

Turnout was around 45%, which is higher than at the last (2003) midterm election, despite a “null vote” protest movement. (Null votes themselves were about 5% of total votes cast, which was barely any larger than in 2003.)

This is the firstsecond time in Mexico that a party other than that of the President will have the plurality of seats in the Chamber of Deputies. (Thanks to Manuel, below, for the correction.)

The PRI has not won enough votes to qualify for a majority of seats, as Mexican electoral law prohibits a party from having more than an 8-percentage point over-representation. The PRI has a preliminary total of 36.2% of the vote.

Thus it will not be divided government in the narrow sense of majority opposition, unless the PRI and other parties cooperate after the election to control the Chamber in opposition to the President. That is possible, but I would think unlikely.

The National Action Party (PAN) of current President Felipe Calderon is a distant second, at 28.0%. At about 5 percentage points, this is not much of a midterm decline for the incumbent party, and it is about 3 percentage points worse than its 2003 votes result. However, given the large boost the electoral system and three-way competition in the single-seat-district tier gave the party in 2006 (when it won about 41% of seats), it will likely have suffered a bigger decline in seats.

The Democratic Revolutionary Party (PRD) did quite badly, with only 12.2%, or barely more than a third of the votes that it won in 2006 when it narrowly lost the presidency.

The Green Party, which is an ally of the PRI, has 6.5%. These two parties had a partial pre-electoral coalition, running jointly nominated candidates in 63 of the 300 single-seat districts, but separate party lists for the “parallel” tier of 200 proportional list seats (as well as, of course, separate candidates in the other 237 single-seat districts). While seats totals are not yet reported (as far as I know), it is unlikely that these two parties would have a joint majority, despite 43% of the votes, as the Greens presumably have won no single-seat districts (outside of coalition districts, where they might have helped the PRI fend off another party’s candidate). I wonder if there will prove to have been districts where the PRI and Greens running separately might have “spoiled” a chance at a PRI win. Perhaps the parties’ pre-electoral strategy was sufficiently targeted to have avoided such an outcome.

The electoral law, and changes made in 2008 (see explanation by Manuel in the previous thread), permit parties to have their separate ballot symbols in districts where they have a joint candidacy. Because there is no separate vote for the list tier, this is important to parties in a coalition: they can still collect their separate PR seats despite the common candidacies. However, any pre-electoral coalition is bound to lose some votes for one of the parties from voters who dislike the partner (or its candidate in the district). Presumably this is why the parties ran separate candidates in the vast majority of districts.

The smaller Workers Party (PT), however, ran a coalition in all 300 districts with the Convergencia. And here I need some more help on how the new electoral law works. These two parties are shown having 3.6% (PT) and 2.4% (Convergencia) for their separate ballot symbols and 0.2% for the coalition, per se. So do voters have the choice of endorsing the coalition as a whole or choosing one of the component parties? Obviously any of these options would be a vote for the candidate in the given district, but how are list votes calculated in the case of voting for the coalition? (For the record, the PRI-Green coalition obtained 0.4% in addition to the separate votes by party reported above). (Again, Manuel to the rescue in the comments!)

Despite all the talk of a “comeback” by the PRI, this result is roughly the same as its performance in the last midterm election, 2003. Maybe 35-38% is just the party’s “natural” level of support in Mexico’s new democratic era. Of course, it does look like a comeback from its dismal third-place showing in 2006, but even then its congressional vote was only about eight percentage points worse than its showing in the current election. That is well within normal election-to-election variation in democracies–especially in presidential systems, where every legislative election takes place in the context of voter evaluations of the parties’ presidential candidates, or of the incumbent president. In the concurrent election of 2006, the PRI suffered defections due to an unusually strong PRD presidential candidate who was the main challenger to the PAN. The PRI also had an extremely weak candidate of its own in 2006: Roberto Madrazo, one of the ultimate “PRIosaurs.”

The PRI thus remains about where it has been since 2000: the pivot party in congress, with a strong regional base (having won most of the governor races in recent years including apparently four of the six up yesterday). Yet it remains a party that will need a very good candidate in 2012 if it is to reach the 40% range that will likely be needed to win back the presidency if the PRD does not make it another strong showing three-way race.

Divided government in Mexico?

Will Mexico’s midterm elections on 5 July result in divided government? If so, it would be a first in many decades–probably first ever–for Mexico.

For the record, it is important to note that I define “divided government” as a situation in which the legislative majority (at least one chamber) is controlled by a single party or alliance of parties that is opposed to the president. This is a stricter and narrower definition than many others use: it is common to see reference to “divided government” whenever the president’s own party does not control a majority. However, simple cases of no-majority situations are a mixed bag: in some, the president may have effective control over the legislative agenda through alliances (formalized or not) with other parties. Or the president’s party may be sufficiently dominant that the president enjoys something very nearly approaching “unified” government, the absence of a majority notwithstanding. In still other cases, there may be bill-by-bill negotiations between the president’s party and others parties (or individual legislators) to achieve majorities. None of these quite matches the image conjured up by the idea of “divided government,” which implies that the institutions of governance are divided from one another in their partisan preferences. While some cases of no-majority may qualify, it is likely that many (probably most) do not.

In addition, a situation in which no party has a majority is actually the norm in most presidential systems–including Mexico in the past 12 years. If the concept of “divided government” it is to be useful for setting off specific situations from a presumed norm, then we might not find very useful a conception in which presidentialism is “divided” most of the time.

So, will Mexico have divided government after these elections? That is, will one party, other than that of the president, win a majority of seats? It is possible, but hard (for me, anyway) to say just now likely.

In Mexico’s bicameral congress, only the Chamber of Deputies is up for election Sunday. The Chamber is fully renewed every three years, concurrent with the presidential election every six years, and again three years later at each president’s midterm. (The Senate elections are entirely concurrent with the presidential; no party currently has a majority in that body, either.)

The former ruling party, the Institutional Revolutionary Party (PRI), is expected to do well. No party has had a majority in Mexico’s Chamber of Deputies since the the PRI lost its decades-long control in 1997, the same year in which the current electoral system was used for the first time.

Mexico uses a quite majoritarian version of a mixed-member system (not MMP, as sometimes erroneously said in some sources–see my previous discussion for why). There is a cap on just how majoritarian an outcome can be: no party may obtain more than 60% of the seats (which would, in any case be unlikely) and, more importantly, no party may have a percentage of seats that is more than eight percentage points greater than its votes percentage. Thus the PRI would have to win just over 42% of the vote to have full control of the Chamber. Winning this percentage would not guarantee a majority; rather, winning less than 42% would guarantee no party would have a majority.

Could the PRI win 42% of the vote? I do not have time to look at any actual polling. But who needs polls? Let’s look at trends and see what would be necessary for the PRI to pull this off!

To win 42% of the vote, it would need a 14-percentage-point improvement on its 28.4% showing in 2006. A tall order, but not impossible. In that election, current President Felipe Calderon of the National Action Party (PAN) won only 36.7% of the vote, and his party ran slightly behind that, at 34.6% (which was good for 41.2% of the seats, given the mixed-member majoritarian system). The Revolutionary Democratic Party (PRD), whose candidate, Andres Manuel Lopez Obrador, narrowly lost to Calderon with 36.1%, won 29% of the Deputies vote in 2006.

The most relevant comparison for 2009 would be the 2000-03 cycle, the last time a PAN president faced a midterm election. Here are the votes for the three parties in 2003, compared to 2000:

    PAN, 31.9, 38.2 (-6.5)
    PRI, 38.2, 36.9 (+1.3)
    PRD, 18.3, 18.7 (-0.4)

That is, the PAN lost 6.5 percentage points of the Deputies vote in the 2003 midterm election, compared to its share at the 2000 concurrent election.

We can also go back to the last midterm under a PRI president, 1997, and compare to 1994:

    PRI, 39.1, 50.2 (-11.1)
    PAN, 26.6, 25.8 (+0.8)
    PRD, 25.7, 16.7 (+9.0)

(Data from 1994-2003 are from the Nohlen, et al, handbook.)

It has been a long time since the PRI has seen 40% of the vote–in fact, not since 1994. So the answer to the question about divided government depends largely on where we think the PRI’s “natural” level of support lies. If its natural level in legislative elections, when it lacks any presidential-candidate coattails (which can be negative or positive) is reflected in its 39% share in 2003, it has a good base on which to build. And then it needs only a few percentage points more to make it to majority-possibility territory.

A lot also depends on how much the PRD has fallen since what was likely the highest share it will see again for many years in 2006. It is unlikely that many of its voters in 2006 have subsequently moved in the direction of the PAN.

The PAN actually may not lose many votes. It certainly will not lose 6.5 points, as it did in 2003, when it was coming off an “unnaturally” high share at the election of the first non-PRI president in 2000. It managed to retain the presidency with barely over a third of the vote in 2006, or not much beyond its “core” electorate. That is, smaller surge, smaller decline. So if we figure 30-33 percent for the PAN, the main question is how much of the remainder goes to the PRI, rather than is split between the PRI and PRD.

I think the implication is that 42% of the vote for the PRI is within reach, but far from assured. And then there is the question of whether the PRI could get the majoritarian boost it would need to get more than half the seats, even if it cleared 42% of the vote. Here, the prospect depends on how many districts there are in which an increase in its votes is sufficient to place it ahead where a PAN (or PRD) candidate won the plurality in 2006.

There is also a “null vote” protest movement. I have no idea how successful that could be, or which party would gain if many voters heeded it. Turnout is also a factor: in 2000, 63.2% voted and in 2003 only 41.1%. In 2006, turnout was back to nearly 2000 levels (58.5%), so we might expect the turnout to be again near 40%, or perhaps lower. But, as with the null-vote movement, I can’t say how that would help (or hurt) the PRI. (The turnout decline in 2003 obviously did not hurt.)

Divided government is most certainly possible. But the PRI will have to have everything go its way.

Elections in Michoacan, Mexico

On 11 November, the Mexican state of Michoacan, in the central region of the country, had elections for governor, state assembly, and municipal offices. The state is the birthplace of Cuauhtemoc Cárdenas, the founder of the center-left Democratic Revolution Party (PRD). The PRD’s presidential candidate, Andrés Manuel López Obrador (“AMLO”), narrowly lost the 2006 national presidential election.

In the state elections, the PRD held on to the state’s governorship, but with not even one third of the votes. Leonel Godoy Rangel had 33.1%, beating the candidate of the National Action Party (PAN, the party of President Felipe Calderón). The PAN candidate, Salvador Lopez Orduña, had 30.5%. ((Orduña has been the mayor of the capital, Morelia, for the past three years. (The PRD also won that city’s mayoralty on 11 November.) )) The candidate of the PRI won 24%. In addition to the PRD, Godoy was backed by the smaller PT, Convergencia, and Alternativa parties.

From a preliminary count of the elections for state deputies (for the unicameral legislative assembly), it appears the PRD-PT alliance won about 31.9% of the vote to 29.2% for the PRI and 27.5% for the PAN. ((I do not know the electoral system of Michoacan. All Mexican states have variants of the national single-vote MMM system, though some lean more towards MMP. There are 24 single-seat districts, and it appears that 12 of them were won by the PRD-PT, while the PAN won 8 and the PRI 4. There would also be some number of PR-list seats, but I do not know how many or how they are allocated. As long as they are not highly compensatory–i.e. that the system is not MMP–the PRD-PT will be substantially over-represented in the legislature–perhaps around 40% of the seats. The legislature’s website was not working when I tried to check on its size or electoral system. On the Google search page, there was a page within the legislature’s site indicated as being about an Acuerdo de Reforma Electoral.)) Assuming those results are correct, note that the order of finish for the second and third parties was reversed between the two elections. The obvious conclusion would be that some PRI voters favored the PAN gubernatorial candidate in an effort to block Godoy. Similar tactical voting (on a much larger scale) by PRI voters probably prevented AMLO from winning the presidency in 2006.

Despite the “juxtaposed government” of PAN at the center and PRD in the state, ((I owe the term, juxtaposed government, to Alain De Remes.)) and despite AMLO’s continuing refusal to accept the PAN national victory, Governor-elect Godoy promises that his relations with the President will be “cordial.” He further says:

Nosotros no podemos adoptar actitudes suicidas, de no tener una relación de plena colaboración ante tal dependencia del Gobierno federal.

Indeed, it would be “suicidal” to adopt a confrontational attitude, given that 96% of the state’s revenues come from federal transfers. ((According to an article in the 12 November edition of Reforma by Adán García, Denis Rodríguez y Daniel Pensamiento, which was also the source of the quote from Godoy. (Via Lexis Nexis.) ))

Normally, like the federal executive, a state governor in Mexico serves a six-year term. However, Godoy’s term will be four years, following a state constitutional change. Reforma says the change is meant to synchronize state and federal elections in the future. How far in the future? The next federal elections will be in 2009 (lower house of congress) and after that, 2012 (presidency and both federal chambers). So, only if this governor and his successor are elected for four-year terms will elections be synchronized–in the federal midterm election of 2015 (presumably again for a six-year term). It seems if synchronization is the goal, a clever and mathematically inclined political engineer might have come up with another way (e.g. elect this governor for five years, and then have state and federal elections in 2012).
________

Incumbent’s party leads narrow, disputed race in Mexico

This time, the partisan tables are turned. Unlike the narrow, disputed presidential contest, the PRD is the incumbent party in the state of Chiapas. The PRD’s candidate, Juan Sabines Guerrero, is holding a razor-thin lead over Jose Antonio Aguilar Bodegas, the candidate of the PRI, with PAN tactical support, in Sunday’s gubernatorial election.

According to the Instituto Electoral Estatal (you must go to their website and see the silly graphics!), as of the time of this planting, the PREP shows the PRD candidate ahead by 0.22 percentage points, or 2,405 votes.

bar_gobernador_chiapas.png

The bar on the left (albeit not ideologically) represents the PAN candidate, registered before the party opted to back the PRI candidate in attempt to oust the PRD from the governorship. The PRD candidate, Sabines, defected from the PRI only in recent months and is the son of a former governor. As George Grayson notes in the just-linked El Universal English story from 18 August, Chiapas has been the scene of some interesting shifting party alliances:

[Current governor Pablo] Salazar narrowly beat the PRI candidate in 2000 thanks to a PRD-PAN-PT-PVEM-Convergencia coalition, and could not muster support for his preferred successor Rubén Velásquez.

In the spring, he was eager to support a PRI nominee, and he and party envoy Tomás Yarrington agreed that Sabines would be the ideal choice. [National PRI leader and presidential candidate Roberto] Madrazo initially gave the thumbs up to Sabines on ly to infuriate local leaders by shifting to Aguilar.

That’s ancient history. Sunday’s gubernatorial election represents a continuation of the July 2 presidential conflict. Last weekend, López Obrador left the encampments of Mexico City to barnstorm on behalf of Sabines.

The need for the PAN’s Felipe Calderón–the apparent narrow victor in the national presidential election–to develop a good relationship with the PRI for “governability” reasons is behind the PAN’s throwing its support to Aguilar in Chiapas. Grayson again:

To propitiate PRI legislators, Calderón’s PAN — in a move spearheaded by teachers union boss and Chiapanecan Elba Esther Gordillo — has extended a helping hand to the PRI. Specifically, they have convinced the gubernatorial aspirants of the PAN (Francisco Rojas) and the Gordillo-controlled Panal (Emilio Zebadúa — Gov. Salazar’s former government secretary and a strong PRD pre-candidate) to step aside in favor of Aguilar Bodegas, who already enjoyed the endorsement of the PVEM [Green Party].

Ah, the confluence of nested games and Duverger!

Three-party dispersion in Mexico

The graph below was prepared by frequent F&V propagator Rici Lake. It shows the dispersion of votes among the three main presidential candidates in Mexico’s 2006 election, at the level of the polling place.

Mex_2006_dispersion.jpg

You may view a larger version in a new window by clicking the image. Better yet, go to Rici’s website and download the PDF, which also has graphs of the House and Senate elections, and for each state. Fascinating stuff, and thanks to Rici for the permission to post the graph and link here at F&V.

From Rici’s correspondence with me, here is a quick quide to interpretation:

I only used the votes for the three main parties, so all points are normalized to add up to 1; they can then be placed on an equilateral triangle where each component is the distance from the point to one side of the triangle. I plotted every acta (taking the data from the official IFE results, which still have a few errors in them, but not enough to alter the results significantly), weighting the acta by the size of the corresponding casilla. Each plotted point represents the sum of the weights of the corresponding actas, normalised so that the 99 percentile point is solid black.

As you can see, the main plot essentially has four modal points (or clouds), one at about <55, 27, 27> (PAN, APM, PBT); one at about <30,17, 50>; one at around <20, 7, 73> and a faint one (at the bottom) at around <5, 37, 58>. Arguably, the first and second clouds show three-party contention, although it is clearer in the first cloud. The third cloud is PRD vs. PAN with PRD in the strong majority (possibly showing PRI->PAN strategic voting?) and the fourth cloud is mostly Tabasco.

This is clearer if you look at the dispersion maps per state. For example, you can see very clear two-party races (everything clusters to one side of the triangle) in Chihuahua, DF, Mexico state, Nuevo León and Tabasco.

For the images mentioned in the last paragraph, go to the PDF, linked above.

UPDATE: There was an interesting follow-up to the above in the comments at Andrew Gelman’s blog.

Income and the vote in Mexico

I just came across a really interesting analysis by Jeronimo Cortina and Andrew Gelman of income and voting choice in Mexico. In a paper (which I skimmed, but will read later), the key graph of which is posted at the blog, Statistical Modelling, Causal Inference, and Social Science, Cortina and Gelman analyze the relationship between income and the vote in each state and the DF.

The key finding, from the 2000 presidential election, is that:

The slopes are higher–that is, income is a stronger predictor of the vote–in poor states.

This is similar to a relationship found in the USA by Gelman, et al.:

One difference between the two countries is that in the U.S., the conservative party does better in the poor states, but in Mexico, the conservative party does better in the rich states. But at the level of individual voting, the patterns in the two countries seems similar.

In carrying out their analysis, the authors use a left-right scale that places the PRD at the left, PRI in what they call the “blurry center,” and PAN at the right. Nothing at all wrong with that ideological description, but is this the expected relationship of income level to the vote? I would think not: the poorest Mexicans probably have the strongest tendency to vote PRI, with the PRD strongest in the lower-middle classes (at least in 2006). In fact, the paper contains a set of graphs of the percent vote per state, arrayed by state income. The PRD shows hardly any relationship, while there is a negative relationship for the PRI (and the unsurprising rise for the PAN). I should emphasize that the paper (and the graph at the blog) analyzes individual data in each state (in a multilevel model); whether the ordering of the parties under the assumption of poor–>PRD/ middle–>PRI/ wealthy–>PAN matters to their results is not clear to me (at least until such time as I have actually read the entire paper).

Cortina and Gelman note that they hope to replicate their study when (if!) 2006 exit-poll data are made public. Believe me, they are not the only ones waiting to get their hands on those data!