The Power and the Money

Mexico’s democracy is already in bad shape

I am on record as being relatively unworried about the AMLO administration. That is not because I do not think that President López Obrador will make capricious and counterproductive decisions. Rather, it is because I do not think that he will do major damage to Mexico’s institutions. And the reason I do not believe that he will do damage is because Mexico’s governing institutions are already in rather bad shape. It is hard to wreck a car that is already totaled. 

Consider Hungary and Poland. The Hungarian prime minister has been assaulting democracy and the rule of law in his country for almost a decade now. The country’s liberal democracy rating (measured by the Varieties of Democracy project) has declined from around 0.7 (on a scale of 0-1) in 2010 to 0.5 today.

The Polish ruling party has been weakening its country’s institutions for rather less time but with similar success: its rating has crashed from 0.8 in 2015 to 0.6 today. 

But both countries still rank above Mexico, which has hovered around 0.5 since 2000. That is to say, despite a multi-year and sustained assault, the quality of democracy in Hungary and Poland is still no worse than in Mexico.

To make another comparison, in South America only Ecuador and Venezuela rank below Mexico. (Bolivia falls inside the error bars.) Peru, Paraguay, all the same or better.  

Now, that is not to say that democracy and the rule-of-law cannot get worse. They certainly can. But that will not happen because AMLO built a political machine or engaged in populist politics. To get worse than Mexico’s current level AMLO will have to openly assault democratic institutions, in the style of Turkey or Venezuela.

That could happen, but there is no evidence of it yet, not even with the fraudulent electoral consultations or the idiotic cut in top bureaucratic salaries. When it happens, if it happens, you will be able to see it happening ... because Mexico is already on the bottom edge of what you see in normal countries.

Federal regress

To repeat the obvious, Mexico is a big federal country. A gigantic federal country, in fact. Which does not make it an advertisement for federalism. Quite the opposite, actually. For while there are a few success stories about states and cities that have creatively taken on challenges, the general picture of Mexican subnational governments is ugly.

A long time ago, in the last century, I dated someone whose job was to audit the accounts of Mexican county governments and help them spend wisely. She soon adopted an attitude not unlike those you might find among E.R. doctors spending their time to limit the damage caused by human stupidity. Two decades later, the causes of her cynicism have not improved. 

Now, that is bad for Mexico, but it also helps explain why I am so sanguine about the arrival of the AMLO administration and the creation of a new national political machine. (Hat tip to Viridiana Ríos for the data, if not the interpretation.)

Consider first that between 2013 and 2017 the federal authorities reported that state governments “lost” about 4.5% of all federal transfers, coming to about $8.8 billion. Poof, lost, or in the words of the federal government, “unable to be accounted for or recovered.” Between you and me, that means stolen. 

Of course, not all states are created equal. Veracruz, Michoacán, and Chihuahua between them managed to steal lose more than 10% of the federal money handed to them in 2013-17, accounting for more than half the total theft losses. But it is a rotating field: in 2017 you had Coahuila, Tamaulipas, and Oaxaca leading the list.

One problem is that Mexican electoral law makes machine politics easy: you only need a plurality to win the governorship (or county executive) so a machine only needs to gather a minority of the vote.

But a bigger problem is that the feds only come after subnational governments if they are egregiously corrupt or  get into bed with the wrong “cartel,” (e.g., criminal syndicate). The voters may hate you but find it tough to dislodge you; the feds may be relatively honest but lack the resources to police local governments.

Now, the Mexican electorate is in a really anti-incumbent mood. Voters have become so anti-incumbent that even long-lasting bastions flipped to Morena. In Mexico State, for example, the voters abandoned the party en masse, including such bastions as Jilotepec and How much of a PRI bastion was Atlacomulco? Well, you can read the link to see how much the town benefited from its association with the party. Consider also that the faction controlling the state PRI is called the Grupo Atlacomulco. The Grupo Atlacomulco is controlled by members of the Del Mazo clan (which includes President Peña) which originally hails from Atlacomulco. Hell, members of the Del Mazo clan ran Atlacomulco during the Porfiriato. (See page 258, bottom right.) So if there was a town that was going to stick to the PRI, it was Atlacomulco. But it did not.

But does that mean that local machines are finished? Not so fast. Consider that Atlacomulco flipped in the 2018 federal election. In last year’s gubernatorial election the PRI hung on with a hair under 34% of the vote, installing (of course) Alfredo del Mazo as governor. 

In short, Mexico is dominated by corrupt machines, even excluding the places where criminal syndicates freely assassinate local officials. These machines are very hard to dislodge and cost the country greatly. They have no been able to lock up federal elections ... but federal elections have done little to reform these local machines. 

The implication for AMLO is that Morena may be able to get the support of many of these local machines, peeling them away from the PRI. But that does not mean that Morena will get them to work any better or that controlling these local machines will help it remain in power at the federal level if things do not get better by 2024.

That is not to say that I am fully sanguine about Morena in power. I worry about populism: literally decisions driven by their short-term popularity. And oh boy does the new AMLO administration have a bunch of dumb-but-popular moves ready to go before Congress ...

¿Una maquina Morena o una maquina morena?

There is no doubt that AMLO intends to construct a new political machine. The question is whether doing so will degrade the quality of Mexican democracy or Mexican governance. (These are correlated but not the same.) In a couple of different posts, I am going to make the case that AMLO’s machine is unlikely to make either situation much worse — because both Mexico’s democracy and its governance are already highly degraded. Things can always get worse, but AMLO is unlikely to make it so.

For this post, I want to park questions about Mexican democracy under AMLO and consider what might happen to the quality of Mexico’s governance.

Let’s ask, what does a democratic political machine do? It doesn’t blatantly steal elections: that’s a dictatorship. But democratic machine politics is different in kind from regular interest group politics. Interest group politics works by promising veterans higher benefits or auto companies a higher tariff; conversely, it works by cutting veterans benefits or increasing taxes on gasoline. Machine politics works by offering Josefina the Welder a job or Ford Motor Company a special break; conversely, it works by stopping Josefina from getting a mortgage or launching audits against Ford executives.

Successful machine politics does not require stolen elections, but it makes opposition harder to organize, increases the autonomy of machine leaders and weakens democracy’s ability to hold officials accountable or translate public desires into public policies.

The PRI under President Peña used machine politics to implement energy reform in 2013. In order to pass and implement the reform, the Peña administration needed the support of the national oil workers’ union. At the time of the reform the oil industry provided 4% of all public revenue and supported much of the economy in the states Veracruz and Tabasco. A crippling oil strike could sink the party’s position in the midterms or next presidential election.

How to get union support? First, threaten the union leader. Carlos Romero Deschamps happened to be a sitting PRI senator and “enjoyed” second-place on the Forbes list of the ten most corrupt Mexicans. On a monthly salary of US$1,864, Deschamps managed to accumulate a $1.5m “cottage” in Cancún, a son who drove a $2m Ferrari, and a daughter who liked to post Facebook images of her jet-setting her three English bulldogs around the world. In other words, a serious corruption investigation would toss him and his children into jail.

President Peña made it clear to Deschamps that he had no choice but to go along with the reform in the most obvious way: in February 2013 he took the powerful leader of the teacher’s union, Elba Esther Gordillo — number 1 on the Forbes list and head of the New Alliance Party — and had her arrested on anticorruption charges. Gordillo was fortunate enough to be over 70 years old and therefore be eligible for house arrest instead of prison — she owned a rather nice apartment in Polanco, Mexico City’s equivalent of the Upper East Side — but the message was clear. Deschamps brought the union on board with reform.

Alternative facts are not an American monopoly: in a rather startling 2014 declaration, union leader Carlos Deschamps declared that energy reform would bring no lay­offs. Even more startlingly, Deschamps denied that mass firings were taking place even after they began. According to him, only oil workers employed in private companies lost their jobs.

Source: Abad and Maurer (2018).

Unfortunately, the reality was otherwise. Reform let Pemex lay off almost 30,000 employees in a three-year period. (See above.) Layoffs hit unionized and non-unionized workers alike; in fact, the share of union workers declined slightly during the mass layoffs of 2015. But the machine kept that from triggering strikes by (1) concentrating the layoffs in places that were already voting against the PRI—i.e., Tabasco and Veracruz—where workers did protest but the protests could be controlled; and (2) making sure that wages went up dramatically for those workers who kept their heads down, kept their jobs, and accepted the reforms. (See below.)

Source: Abad and Maurer (2018).

Of course, you would be right that this particular exercise of machine politics removed most of the putative fiscal gains from allowing the layoffs. The chart to the right shows payroll costs as a percent of Pemex revenue: they did not fall. 

But that’s how machine politics works. Rents are generated and distributed on one end; individuals and individual group leaders are punished on the other.

Absent the subversion of democracy, it’s hard to see how AMLO could build a machine more effective than the PRI, circa 2012-18. (To reiterate, the PRI circa 1929-1997 was an institutionalized dictatorship masquerading as a political machine, not a genuine democratic political machine.) There is no obvious source of rents that the PRI did not already tap. There are places that have improved — say, competition policy — but where it would be hard for Morena to loosen the rules without pushback. Elsewhere, things are not good, as this OECD report on electricity procurement shows. In that sense, things cannot really get worse.

And a democratic machine, no matter how efficient, can’t stay in power if it really lets things go to hell. The PRI discovered that this year, losing even its heartland in the town of Atlacomulco, Mexico State. In other words, straight-up machine politics means more of the same in Mexico and is self-limiting if it delivers truly bad results — albeit less self-limiting than in democracies free of machine politics.

Morena will try to build a machine. But building a machine is hard and machines themselves do not guarantee that a bad government can remain in power.

The real worry, I think, is democratic backsliding. That is, machine politics plus institutional changes to make the vote less effective and hamstring the opposition. Without democratic backsliding, you might get a maquina Morena, but you won’t get a dark machine.

Thoughts?

The danger from AMLO

In comments, Andrei Gomberg writes that the biggest danger from the new AMLO government is the construction of an effective political machine and the slow subjugation of the independent parts of the Mexican state:

I think you miss the real source of additional congressional votes for the new government: the PRI. Admittedly, it will be a much smaller source this time — but it will be more important than the PRD or MC. The rump PRD (with well-known exceptions) already largely consists of those who hate him — why else were they not on board already?

On the other hand, big chunks of the PRI can be easily accommodated in the AMLO coalition. He is, really, an old priísta, and he has long shown to appeal to the PRI traditionalists (let´s call them “Bartletts”). Now that he is in power, he will have no problem attracting many of the rest of them.

One also does not need to look for any idées fixes to see what his primary objectives will be. We are dealing with the classic machine politician, who knows everything there is to know about building up power. His first and foremost task will be to rebuild the PRI machine of the olden days - but under himself. He is in position to rearm the old (and by now quite decayed) clientelistic networks — why else was he bringing old union leaders on board?

Of course, this time the machine will be rebuilt under the unitary leadership. There are few local caudillos within Morena, and AMLO has shown that he is very much willing to slap down the few there are (greetings, Sr. Monreal). It is hard to see where any independent force may grow within the governing coalition. Unlike old priísta presidents the new leader will be much less constrained from within the party. I mean, if anybody were to tell me 15 years ago that I might miss good old Fidel V. ....

Again, thinking along the lines of what you taught me about the way Mexican power functions. I would expect him to start as a good friend of “national” business. Any machinery needs oiling to run smoothly — he needs rents. Competition regulators will be an interesting area to look at.

The other area, of course, will be the electoral machinery. He does not trust it — and for a good reason: it is still independent. Unlike you I fully expect him to go for the 2021 “plebiscite”: it is his pre-electoral promise, and it is very much in character. What I expect him to do is to run it largely or fully outside the Instituto Nacional Electoral system (if he needs the excuse, it could be the lack of constitutional authority). The objective, of course, will be, precisely, to get the country used to the idea that the Instituto Nacional Electoral is not necessary for his version of “democracy.” Chances are, he would still be popular enough to get a resounding confirmation of that popularity in any properly organized vote — but he will not want that.

Once again, where I expect to see the most important changes is in gutting the strength of all sorts of independent institutions. That is where the things that will really change Mexico will happen. Everything else, at least in the first couple of years will be more of a show.

More thoughts to follow.

How to swear in Argentine 1: Irse a las piñas

I should and will write more about the Mexican elections and the future of Pemex. Especially given the food for thought Andrei Gomberg presented here! 

Instead, though, I want to write about swearing in Argentina. In my travels to and around that country I have often found myself utterly dumbfounded by what people are trying to tell me. I understand the words, but the sentences just don’t make any sense. So therefore we begin a new series: “How to swear in Argentine.”

We begin with “irse a las piñas,” which does not mean,  terrible Guardian translation notwithstanding, “go to the pineapples.” I don’t  mean that the translation is terrible because the expression is figurative; I mean simply that “piña” does not mean “pineapple” in Argentine. For some reason, the word for pineapple in Argentine is “ananá”; “piña” is “pinecone.”

Which makes more sense when you consider that “irse a las piñas” means “come to blows” or “have it out” or “step” or “throw down.” And since pinecones are not something that Argentines talk about particularly much (save Christmastime) other than in the context of “going to,” there is never going to be any confusion.

So if two people in a Pilar bar start whaling on each other, now you know how to describe it. 

Seven initial thoughts on the Mexican election

Wow. AMLO cleared an outright majority. That was impressive and a sign of just how fed-up Mexicans were with the status quo. I saw some signs of that last summer; now it is confirmed. 

Seven initial thoughts:

1. AMLO is not a left-wing version of Donald Trump. He is his own thing. His platform is frustratingly vague on most things, but he is a professional politician and a rather good one at that. And his idées fixes have to do with helping the downtrodden, not personal vendettas, foreign trade, or ethnic resentments.
2. AMLO is not a huge fan of the rule of law and he does not respect Mexico’s governing institutions. But neither did the outgoing PRI, which buried prosecutions and spied on its own population. Creeping dictatorship is unlikely and corruption will not get worse. That is sadly not saying very much. 
3. AMLO is not in a position to become Hugo Chávez, even if he wanted to. Chávez began by rewriting the constitution, faced down an old-fashioned military coup and rode a wave of oil money. That is not going to happen in Mexico.
4. AMLO is in a position to shepherd through substantial legislative changes. As we have said, he can pull off PRD and M.C. legislators. But he has another advantage, which is that federal legislators can now be re-elected. That makes them less beholden to their party leadership. Don’t expect radical change, but do expect him to pass what he wants to pass. 
5. The challenge is going to what it has been: can AMLO get a grip on violence and accelerate growth? I am skeptical but hope for the best. The real questions will come with events. AMLO will be tempted to respond to adversity with populist rage. That is my biggest worry. 
6. I do not expect AMLO to hold a referendum on his continuation in office in 2021. If he does, that would be a bad thing.  
7. The PRI is the cockroach of Mexican politics; I do not expect it to die.

I would be delighted to hear other reactions and predictions. 

In Mexico, like America, a Congressional majority is not enough to govern

Earlier we pointed out that Mexican presidents need an effective legislative majority of two-thirds to get much of anything done, since so much legislation requires constitutional amendments. Did AMLO’s coalition pass that threshold?

We won’t know for sure for a while, but the answer appears to be no. Preliminary data is here. Mexico elects its Senate by giving the plurality party two senators per state, the runner-up one senator, and allocating the rest by proportional representation in a national 32-senator district using the largest remainder method. That method gives AMLO’s coalition roughly 53% of the seats. 

In the lower house, the results look similar. Mexico elects 300 deputies in geographic winner-take-all (aka, FPTP) districts and allocates the remaining 200 via proportional representation. The calculation is more complicated than for the Senate, because the country is divided up into five 40-deputy districts for the purposes of allocating seats. There, my initial estimate is that AMLO’s coalition got about 63% of the seats.

AMLO’s coalition, of course, includes the crazy Social Encounter Party; I have no idea how many seats they will wind up with. It depends on the affliation of the senators and deputies who won under the coalition banner in the geographic districts. They may prove somewhat unreliable for the new President. So his actual majority will be a little less than the stated number.

That does not mean that President López won’t be able to govern! As mentioned, he can pick up votes from the two social democratic parties. The PRD and Citizen’s Movement won two senate seats each and picked up about 4% of the House; both parties will also get some seats from their weird left-right coalition with the PAN, up to a ceiling of 18 senate seats and another 8% of the House. If they meet the high-end of the above projection (which they won’t) then AMLO will have a governing majority. 

But he will still need their cooperation. And if the social democrats get less than 18 senate seats, then he will need the cooperation from other parties as well. (Like the Greens. Shudder.) 

Right now, attention should be moving to the state legislatures. In the Nuevo León congress, pictured below, the PRI-PAN coalition managed to hold on to about half the seats, down from 74% before the election. The result is that AMLO cannot count on N.L. to ratify constitutional changes despite his gains in the state. (In Nuevo León’s 42-seat congress, the PRI collapsed from 16 seats to three; the PAN seems to have gained, from 15 to 18 at last count. The AMLO coalition blew up, from two lone Labor Party state deputies before the election to 15.)

It is unlikely that AMLO won a majority of the state legislatures, since many are elected on off-years, but somebody should check what happened out in the provinces. 

None of this will leave us much wiser about what AMLO will try to do. But it does mean that whatever it turns out to be, it will have to be done judiciously and after political negotiations. Of course, there is a lot that the chief executive can do to wreck a country, even one with stronger institutions than Mexico.

But I would venture that most of that sort of wrecking has already occurred under President Peña. I am not particularly optimistic that President López  will do much to help that little girl in Iztapalapa that I mentioned back in 2012, but I also do not think that he can or will do much to weaken the country’s democratic and legal institutions, at least inasmuch as they have survived the last six years. 

Call it an extremely cautious optimism, if you must. 

Mexico’s real election is for Congress

It has been an exciting week in European politics. I’m in Spain, where a corruption scandal just brought down Mariano Rajoy. Spain is near Italy, where the President just refused to ratify a ministerial appointment and then agreed to a face-saving compromise that moved the offender to a less-prominent post. (Ask European steelmakers how that is working out them; Peter Navarro is not Commerce Secretary but seems to be winning the day anyway.) 

So I am going to talk about a place where politics looks pretty boring at the moment: Mexico. AMLO is running away with the presidential race. Miracles never cease, but right now absent divine intervention the presidential race is sown up.

What is left is control of Congress. Moreover, what matters is the ability to cobble together a two-thirds majority in Congress, since so much of Mexican legislation requires constitutional changes. How close is AMLO to clearing that hurdle?

TL,DR: Close! But probably not close enough.

A bit more detail: Mitofsky has polled congressional voting intentions. According to their models, AMLO’s coalition has a 50% chance of winning between 236 and 298 seats in the lower house, with 251 needed for a majority and 334 needed for a supermajority. They are similarly tipped to win between 51 and 73 senators, with 65 needed for a majority and 86 for a supermajority. 

Here you can see the estimated probability distributions for number of seats for AMLO’s coalition. The dotted line shows a majority. 

The AMLO three-party alliance will almost certainly win the house and has a one-third chance of winning the Senate. But it is exceedingly unlikely to win a supermajority.

Does that mean that we can forget about left-wing control of Congress?

No, because AMLO can win support from outside his formal coalition. 

Mexican party alliances are purely transactional. Consider AMLO’s alliance. It includes the Social Encounter Party, a reactionary beast opposed to abortion and marriage equality. These are not AMLO’s positions, although he can be squishy. But Social Encounter is happy to turn left on economics because it will win more seats as part of the coalition. AMLO, in turn, will win low-information voters who might have otherwise voted PAN. In other words, the coalition is transactional, not ideological. 

The main opposition alliance is equally transactional, marrying the conservative PAN to the social democratic PRD and Citizen’s Movement. Once AMLO wins, there is nothing stopping PRD and Citizen’s Movement legislators from voting for AMLO’s legislation should they so choose.

Will AMLO be able to cobble together a supermajority with support from the other left parties?

The modal projection for the PRD is 31 deputies and seven senators. For the Citizen’s Movement it is 29 deputies and four senators. These projections are not independent and should not be combined, because a better PRD result implies a worse result for AMLO’s coalition. Moreover, the methodology is supremely inaccurate; it does not rely on a large enough sample to make true constituency projections.

That said, there will certainly be a working left-wing majority in Congress but likely not a supermajority. The modal result is that AMLO gets 50% of the Senate, plus 8% from the two social democratic parties. He also gets 54% in the House, plus 12% from the social democrats. Still not quite over the top unless there are some big vote swings before the election.

So can we stop now? Not quite.  

The Green Party, you see, is made up of a bunch of mercenary SOB’s who would sell their parents for lunch money. Personally, I do not think that AMLO will offer sufficient lunch money. But even if he did, Green support would not put him over the top in the Senate. 

And now we can stop. It is remotely possible that AMLO will be able to put together a working supermajority, but that is not the way to bet. He will need the cooperation of the PRI or PAN to change the constitution ... and that is before we consider the state legislatures.

I leave whether that is good or bad for Mexico as an exercise for the reader. Thoughts?

Mexico’s amendable constitution

That’s “amendable,” not “amenable.”  

Too many observers have recently declared that Mexico’s energy reform is safe, because it was enshrined in the Mexican constitution as an amendment. And why would they be wrong? Amendments certainly look difficult to change. It takes a 2/3rds majority of both houses of Congress and a majority of the state legislatures to pass one. That certainly looks like a high bar!

Except that if the bar were really high, the Constitution of 1917 wouldn’t have been amended 232 times in 101 years. Or 64 times in the past dozen. That’s right, Mexico has been amending its constitution once every ten weeks.

For whatever reason, an amendment process that is on paper equivalent to the United States is in practice much much easier. How did that happen?

Let’s start by saying what did not happen. Many analysts, yours truly included, thought the high amendment rate of the Mexican constitution was a epiphenomenon. The phenomenon was that Mexico was a one-party state in which the written Constitution did not matter. Since the constitution did not matter and the ruling party controlled all branches of government, amendments could be passed for propaganda reasons and without consideration of their effects.

Sounds good, right? The problem with the above hypothesis is its converse corollary: If the Mexican constitution was amended all the time because the country was a one-party dictatorship, then the amendment rate should fall when it becomes a multiparty democracy.

Well, sometime around 1997 the country became a multiparty democracy. But the rate of amendment rose from 1.6 amendments per year to more than five. So much for that idea, then.

So what is the reason?

The first thing to keep in mind is that Mexican political parties are much more centralized than their American equivalents. When the party bosses say “jump,” (most of) the legislators ask “How high?” For example, once energy reform made it out of the Congress, it took only 83 hours for a majority of the state legislatures to ratify. Imagine that happening in the United States.

Second, the President can introduce amendments under his or her power to introduce legislation. Until 2012, they could introduce them straight to the relevant committee. In practice, that means that a Mexican president can almost always get their preferred amendments to the floor for a vote.

Third, Mexican parties are more transactional than ideological. The PRI can turn its positions on a dime for electoral (or other) advantage. The Green Party, despite its, name, is not particularly concerned about the environment. Rather, it is a family enterprise, winning seats in order to collect federal subsidies which it uses to maintain its seats, parceling out votes to the highest bidder.  The PAN and PRD are a little more ideological, but the PAN in particular has a long and successful history of logrolling and the PRD has split. 

Finally, Mexican politicians have no choice but to amend the constitution if they want to govern. Too many policies are enshrined in the Mexican constitution. The bloody thing is now 61,765 words long, up from 27,122 back in 1917. If Mexican politicians suddenly developed stronger norms about amendments, then the country would become ungovernable.

To put this in context, Mexico isn’t the only Latin American country with a flexible and overly-detailed constitution. The difference is that the other countries have amendment process that look easy. Colombia’s 1991 constitution runs 47,383 words and has been amended at almost as high a rate. But with a few exceptions, Colombia’s constitution can be amended by two consecutive majority votes of both houses. That is barely a bar at all.

The Brazilian  constitution is only a little bit harder to amend: you need 60% of both houses to vote twice. The Peruvian constitution (which is a paragon of brevity at 18,341 words) allows for amendments with either a majority vote plus referendum or a two-thirds vote of Congress twice. It’s no surprise that those countries alter their constitutions with regularity. What makes Mexico unique is that it has an amendment process that looks fairly onerous … but isn’t.

In other words, Mexico suffers from many many ills, but legislative gridlock is not one of them. That is a good thing, but it means that supporters of the energy reform shouldn’t get too complacent. It wouldn’t take the second coming of Lyndon Baines Johnson to undo the amendments. It would just take some political skill. Of which there is no short supply in Mexico right now.

Now, that is not to say that I think reform is about to be completely undone ...

The mystery of tranquility in Mexico City

We are back in D.C. Time to think about the time in Mexico City, which we intend to repeat every summer that we can.

Crime has gotten worse in the statistics, but not in the feel. The center of the city is as safe (or not) as ever.

But there are some signs of worse to come. Right after our arrival, there was a large-scale shootout between the Marines and a local gang called the Tláhuac Cartel, Tláhuac being a poor borough out on the edges of the city. (The new Line 12 runs there.) 

The short version of the shootout isn’t ominous at all. The gang’s leader, Felipe “The Eyes” de Jesús, was dealing drugs on the UNAM campus. University officials complained to the feds, so the Marines started an investigation. (Yes, it is a problem that the Marines needed to handle the investigation instead of the federal police.) They found him, moved in, and everything went wrong. Tragic and possibly incompetent, but these things sadly can happen anywhere.

No, what’s worrisome are what else the gang was doing and what happened after the shootout. The gang wasn’t just selling drugs on a big college campus. Rather, it was engaged in extensive extortion of Tláhuac’s retail markets and taxi services. That is worrisome. Now, if taking down The Eyes ends that, then we do not have a problem. But right after the shootout that killed him, mototaxis controlled by the Tláhuac cartel blocked off roads and gang members set a large truck and three buses on fire. That does not sound like the organization’s grip on the borough has been broken.

What makes it worse is the mayor’s insistence that everything is fine, cartels don’t operate in Mexico City. Homicides leaped 25% this year. Just a blip? Possible, but unlikely in a city of this size. Now, I understand why Mancera insists on this, for reasons beyond the political. (He wants to run for president in 2018.) First, homicide was not that high (through September 2016) in the affected areas. Consider this map, from the indispensable Diego Valle. In 2015-16 there were only 19 reported killings in the sector where the shootout happened (outlined). That is a rate of 10.3 per 100,000, which is neither low nor high by American standards. The area is far from a no-go zone. And while it is clear from the overall time series (see below) that something happened in CDMX around August 2016, things do not look to be out-of-control. 

So even though a federal report states that cartels do indeed operate in greater Mexico City (you can see a map of their territories here) the situation is not yet scary, at least not to a jaded child of the 1970s and 80s like myself.

I am, however, I think, French at heart. Mexico City may be under control in practice, but I do not fully understand how it stays under control in theory.

It suffers from the same corrupt and ineffectual police as the rest of the country. And it has had the same trouble making the new legal system work as a the rest of the country. So why does violence remain relatively low? After all, CDMX is safer than Washington, D.C., but D.C. has a decent police force and a functioning justice system. (For example, D.C. clears up about 70% of homicides. Mexico City, around half.) What then keeps a lid on the violence?

Until I can explain that, even if only in theory, I will remain nervous about the future.

En busqueda de Macron

I heard the most terrifying thing from a friend of mine in Mexico City, a Mexican citizen and voter. He is an upper-middle-class homeowner, two wonderful daughters, owns an apartment with an amazing view on the far outskirts of Polanco. We were in Parque Chapultepec with the families on a sunny summer Saturday. What did he say?

“Things are so bad that I’m willing to give López Obrador a chance, just to see what happens.”

This is scary because when I was in Colorado earlier that week, on a trip to a conference, an otherwise sensible upper-middle-class Republican had said the same thing to me about why he voted for Donald Trump. And before him, I had heard similar things from other GOP friends and a few of the people we canvassed in Cheltenham. 

That kind of thinking leads nowhere good.

That said, Mexicans have better reason to be fed up: President Peña really has been terrible. You can point to energy reform, but that would be like the Republicans I met in Colorado who consoled themselves by hoping that tax reform would happen and pointing at Justice Gorsuch. Whatever gains energy reform might give in the future have not been worth the setbacks the Peña administration has inflicted up Mexico today.

So why is López Obrador (aka AMLO) so worrisome? Well, read this. AMLO does not believe in Mexico’s democratic institutions. I doubt he would go full Maduro, but the country’s institutions would not grow stronger under his administration. And Mexico desperately needs stronger institutions.

And so, the search is on for the person who can stop AMLO, the Macron of Mexico. The PRI just took a step in the direction of trying to find its Macron by changing the rules to allow the party to nominate non-members for various offices and reduce the time that the presidential nominee needs to be a member to ten years. They also passed rules to try to make sure that they don’t nominate someone who is completely corrupt.

The beneficiary of all this will be Finance Minister José Antonio Meade. He seems like a nice man. He celebrates his Irish heritage. He has held practically every important office in the executive branch. His grandpa helped found the opposition National Action Party. 

But at the end of the day he is a member of the corrupt administration sweeping Mexico back to the bad old days while failing to protect its people from extortion and violence. In a world where well-off homeowners are willing to give AMLO a chance, I don’t see a white-skinned soft-spoken Priísta as the man to stop him.