The Power and the Money

It is hard to get Manuel Bartlett to understand things when his paycheck depends on him not understanding it

Manuel Bartlett is the corrupt head of CFE, Mexico’s state electricity company. He hates renewables. So he’s trying to kill it with the support of Mexico’s president. From the man himself: “Wind and photovoltaic (plants) don’t pay the CFE for the backup. Do you think it’s fair for the CFE to subsidize these companies that don’t produce power all day? That’s not a free market, it’s theft!”

Do his arguments make sense?

Let’s unpack them. The first argument is that any firm feeding electricity into the grid should pay to transmit the electricity. The second argument is that renewables aren’t on all the time, so somebody has to have backup generators in place and that somebody should get paid. 

The first argument makes little sense. Cenace, the organization that runs the grid (owned by CFE) charges rates to feed in power. Maybe they should be higher for everyone, maybe not, but it isn’t a reason to discriminate.  

There is a technical argument that intermittent power sources raise transmission costs. That technical arguments have been solved. First, the more intermittents you have on the grid, the more predictable it is. Second, weather forecasts are pretty good these days. Third, getting “reactive power“ into the grid when needed is not particularly expensive or difficult. ERCOT, the Texas grid operator, found that you needed capacity about 0.1% of intermittent capacity to maintain grid stability. Internationally, there’s no relationship between the penetration of intermittent power sources and grid stability. Why? Well ... natural gas plants go offline all the time! California’s crisis in 2015 was caused by a leak at a gas storage facility in Aliso Canyon. 

So that’s a nothingburger. But what about the fact that somebody has to keep the lights on when the sun isn’t shining and the wind isn’t blowing? 

That could be a problem. Dispatchable power sources (i.e., ones you can turn on and off at will) suffer cycling costs, the same way that turning a lightbulb on and off every few minutes will raise power use and lower lightbulb life. It’s certainly a problem for CFE! The good news for CFE is that CCGT natural gas plants have relatively low cycling costs. The bad news is that you’ll run them less because solar and wind will eat their lunch during sunny and windy times. CFE will be able to charge a little more but that won’t make up for the loss of business and there will be some increase in cycling costs. 

In short, Bartlett’s argument comes down to the fact that cheap solar and wind isn’t just cheaper than burning hydrocarbons, it also raises the cost of burning hydrocarbons.

But so what? 

Imagine: A cheap airline comes in on a popular route. It steals away passengers from legacy airlines. The legacy airlines’ costs rise b/c they have stranded assets (unused planes) and high fixed costs. So prices on routes operated only by the legacy carriers rise a bit. Would you for that ban low-cost airlines? Or throw a tax on them? Or limit their landing slots? If your answer is no, then you think Bartlett is full of it. 

Yes, CFE has a mandate to keep the lights on. No, it’s not like they’d just turn off their power plants at night if that mandate didn’t exist. They want to make money. Prices will go up for baseload power and CFE will stay in business. It is remotely possible that Mexican electricity consumers will wind up paying so much for baseload power that net electricity prices will rise. Well, remotely possible other than the fact that it hasn’t happened anywhere else. (A decade ago Spain and Germany ran into problems with rising electricity costs because they oversubsidized solar power. But solar power was a lot more expensive back then.)  

The whole thing makes me sad, because I thought AMLO could have been a great president. Instead he’s bad for the Mexicans alive today and very bad for everyone who will be alive in the future. Future generations will revile him. The only reason he won’t be reviled around the world is because Mexico is small and George W. Bush and Donald Trump will outshine him as historical global warming villians.

Once upon a time, back in 2008, I sarcastically joked that (then citizen) AMLO’s terrible energy policy would be good for the world because it would lead Mexico to produce less hydrocarbons and that’s a good thing. What I never expected is that AMLO wouldn’t just want to hobble private hydrocarbon production, he’d also try to hobble anyone who wants to produce electricity without setting anything on fire. 

Anyway, there you have it. Bartlett’s argument has nothing to do with Mexican consumers, who will benefit from cheap solar and wind. Nor does it have to do with keeping the lights on, which is easy to do. It’s just about keeping out cheap competition to benefit CFE. 

The real question is why he feels so strongly about helping CFE and burning Mexican oil. Any answers to that would be greatly appreciated .... 

Covid-19 and American oil production, in gory institutional detail

Oil prices are crashing due to Covid-19. So what happens to American output?

Market reactions first, then political ones.

Well, first the oldest building in Dallas, circa 2019:

Yes, my shoes are bright blue.

Now market reactions to the Covid oil price crash.

Market

The “shut-in prices” is the minimum wellhead price operators need to continue producing from a production well. It’s equal to the amount needed to cover ongoing O&M, including where-necessary pumping and artificial lift, as well as water, gas and steam flooding and other secondary recovery costs.

The first wells to shut down are “stripper wells” nearing the end of their useful life. In the U.S. there are about 400,000 of these wells producing about 720,000 barrels per day. (I said they were small.) They usually rely on pumps to get the stuff out of the ground, meaning those famous mechanical donkeys you can still see in parts of California.

Next to go are wells that depend on secondary recovery. For example, in the Golden State companies inject about 400 million barrels of steam every year to get 161 million barrels of oil. (Governor Newsom has banned new wells using this technique.) These wells are high cost, but when shut down they tend to fill up fast with water. Companies that shut them down risk losing the remaining shut-in oil forever. In addition, companies that shut down wells risk losing their land leases. As a result, these wells don’t close immediately once they start losing money.

After that some wells will start to decline quickly in the absence of new drilling. Tight oil wells in North Dakota decline 65% in the first year and 77% over the first two years. That’s the next to go. In their early phase, however, tight oil wells can be easily shut-in and restarted. So tight oil production will decline in two waves, first by natural decline as new drilling collapses and then again as companies shut wells down. Since you can restart unconventional wells in about a week companies won’t be as reticent to stop production … if not for the fact that most of them are highly indebted and desperate for any sort of cash flow to satisfy demanding creditors.

This chart for tight oil plays in the Eagle Ford region of Texas gives you some idea of just how fast this decline can happen. Without new drilling, production would decline by about 700,000 bpd within a year from this play alone:

But production will decline faster than that, because unconventional wells can be shut-in fairly easily during their first year. 

In other words, the current oil price crash will cause U.S. production to drop, and drop fast. That is, it will if it is sustained. That’s not a given: countries have already begun to reopen their economies. Oil demand will recover as a result, possibly to prices that might keep tight oil production alive and stop the shuttering of stripper wells.

Political

Of course, state governments can act via a process called “proration,” whereby they can order well-by-well output cuts. Back in the 1930s, Congress passed the Connally Act, which made it illegal to ship oil across state lines in violation of state output quotas. It also created the Interstate Oil Compact Commission (here is the law), which brought representatives together from oil-producing states to hammer out overall quotas, America’s own little mini-OPEC before OPEC. (For whatever reason, Illinois and California never passed proration laws, although the California government sanctioned a quasi-private commission that functioned outside the law.)

The IOCC still exists, but its powers have withered. It no longer makes decisions by majority vote. And it now includes 31 states, five Canadian provinces, two Canadian territories, plus Egypt (W?) and the other Georgia (WT??) and the Bolivarian Republic of Venezuela (WTF???). Plus, while Arkansas, Colorado, Michigan, New Mexico, Oklahoma, and Texas all passed proration laws, only Oklahoma and Texas still exercise the muscles. (The Rail Road Commission of Texas and the Oklahoma Corporation Commission still publish quotas, set above most wells’ maximum production.) Executive agencies function like muscles: use ‘em or lose ‘em. Those states would need to design regulations from scratch to impose quotas, and they would certainly be challenged in court.

Now, Texas is big enough on its own to mandate a production cut. And the Texan authorities are talking about it! But it’s all talk. And there’s no sign of the Trump administration trying to resuscitate the Depression-era laws that let it coordinate oil production cuts, or even coordinating talks among the IOCC states. But production is falling fast, so Trump could claim that he jawboned producers into it.

Hell, he could do that even if production wasn’t falling fast. That’s his M.O.

Conclusion

Which means that the ultimate question is whether OPEC wants to be fooled. If Saudi Arabia thinks that letting Mexico de facto defect while pretending to cooperate will not hurt their cartel agreement, then they’ll accept the bullshit. If they don’t, then they’ll reject the Trump-AMLO bamboozle.  

Some crude nonsense: Trump, AMLO, and OPEC

OPEC has been trying to negotiate production cuts in order to prop up oil prices. Mexico resisted making any cuts. AMLO then declared that he had asked the United States to make part of the requested cuts on Mexico’s behalf. Trump, in a press conference, announced that the U.S. would make the cuts but request “reimbursement at a later date.”

What does it all mean?

1. AMLO really loses his senses when oil is concerned. Right now, Mexico loses money on every barrel it points. Inasmuch as it is geologically possible for Pemex to cut production, Pemex should cut production. But AMLO refuses to accept that. It may be because he believes that he’s linked his reputation to increasing oil production and that failing to do so will hurt him politically. But he believes that then he is a terrible politician … and we have evidence that he is not a terrible politician. Which leaves us with the conclusion that he really like to dig hydrocarbons out of the ground and set them on fire because he likes to dig hydrocarbons out of the ground and set them on fire.
2. President Trump does not have the authority to cut U.S. oil production. Individual state governments do have that authority—the Texas Railroad Commission is famous for controlling production in that state—but there is no sign that the President intends to call in state regulators and ask them to order private producers to shut wells.
3. S. production is crashing anyway. Unconventional oil wells in the United States decline hyperbolically. That means that companies need to keep drilling just to keep up production. When prices are low, they stop drilling, and production falls. The EIA projects that U.S. production will fall by at least 500,000 barrels per day this calendar year and another 700,000 barrels in 2021. This isn’t a hard projection to make, considering as production is already down 600,000 bpd from last week.
4. AMLO is putting on a show for OPEC. He’s trying to get OPEC to pretend that a U.S. production fall that would have happened anyway is actually a deliberate action on the part of the United States to help out Mexico. OPEC knows this is bullshit, but will likely give AMLO the fig leaf he’s looking for.
5. Trump’s demand for compensation is meaningless, but not for the reasons you think. When I was a kid back in New York, we’d talk about the “favor bank.” (Often called the “Irish favor bank” for reasons I don’t understand at all.) You’d pay into the favor bank, you’d make withdrawals from the favor bank. Donald Trump has never played by the favor bank. If he wants something from you, he will do whatever he can to get it from you. If Donald Trump wants something from Mexico in the future then he will ask for it regardless of whatever “favors” he granted now. And he will threaten Mexico in order to get it. In that sense, AMLO’s request was costless.
6. OPEC wanted Mexico on board because reasons. This is the only part of the story that I don’t understand. Why couldn’t OPEC have just ignored Mexico’s posturing.

In other words, this is much ado about nothing.

In fact, if this is right, then it really was about nothing for nothing ...

 

Saudi Arabia plans to issue a statement blaming Mexico for not agreeing to cut its production (by the allocated amount of 400k) #OOTT

— Amena Bakr (@Amena__Bakr) April 10, 2020

Stating the obvious about the tariffs on Mexico

I am union-bound to say this, I suppose. The Mexican tariffs are stupid.

First, they will not achieve their stated goal, which is to get Mexico to stop Central Americans from arriving at the U.S. border. Consider what the Mexican government is currently doing:

1. Inviting Central Americans to stay in Mexico. The problem there has been that the Trump administration does not believe that Central Americans will stay in Mexico. Rather, it believes that legalizing their presence in Mexico only makes it easier to get to the border. Under pressure, the Mexican government suspended its visa program after 13,000 applied. But Mexico has allowed Central Americans to apply for asylum from Mexico, as the Trump administration wanted. 
2. Arresting and repatriating Central Americans: 29,660 in the first two months of 2019, about the same as the 29,239 average for those months in 2016, 2017, and 2018. President Trump actually thanked AMLO for this just last month. Through May, Mexico has detained 72,410 migrants and deported 53,371. That is less than the 340,944 Central Americans the CBP detained so far this year, but it is a substantial number. It is far from clear that Mexico could do better given the limited resources available to the Mexican government. 

Second, if the administration sticks to the plan, it will be an own-goal for President Trump. The tariffs on China are bad enough, but companies can adapt and the economy is strong enough to withstand them. Moreover, they are politically popular. Tariffs on Mexico are a different story. These are tariffs that will mess with supply chains. Alonso de Gortari (Princeton) recently calculated that 30% of the headline value of Mexican manufacturing exports to the U.S. consists American exports to Mexico. So they not only hits American consumers, they punch American producers as well. Plus, President Trump can kiss goodbye any chance of ratifying the USMCA. There are cheaper ways to keep immigration on the front burner for 2020, which I presume to be the political goal. 

Third, the tariffs benefit China by taking Mexico off the table as a potential site. Manufacturing has been moving to Mexico from China for years now. The trade war would have accelerated that, at least a little. It is a good thing for America to have industry move next door. But now, well, not so much.

Finally, here is something that is not a problem for this administration: the loss of any credibility as a negotiating partner. Yes, China and Europe will take notice of how the U.S. told Mexico that all it had to do was renegotiate NAFTA but then turned around and screwed it anyway. But that is a feature, not a bug. The current administration intends to impose tariffs forever. Sure, the President will tack back and forth according to political pressures, temporary advantages, and his own strange ego. But the goal is clear, as we reported back in October 2018. Blowing up your credibility just makes it all that much easier to insure that you cannot win the trade wars that you start. 

I have to admit, however, that I think the administration is going to back down on this one. The logic of point (2) above is pretty clear. So put that down as a prediction ... albeit not one that I am particularly confident in. 

Fear and loathing in Mexico

What would I fear if I were an investor in Mexico? (Oh, wait, I am!)

It is plausible to say, “Not much.” For all the ups-and-downs of AMLO’s first six months, country risk remains low. In fact, it remains as a low as Brazil. Now, that may be a low bar, but wealthy investors love President Bolsonaro. I heard one hedge fund manager describe him as “my guy.” They would not describe AMLO as “their guy.” But they are not running away.

And it makes sense that they would not run away! The country is in great fiscal shape; it can afford the refinery stupidity (which won’t happen) and the Tren Maya nonsense (which might). It has been picking up manufacturing fleeing China for years (as we pointed out back in 2013); now it is poised to benefit even more from the great trade war. And if you are afraid of taxes, relax. AMLO’s party has called for tax hikes, but AMLO himself has been repeatedly clear that they will are not on the table: “Income taxes will not go up. We are going to keep the same tax rates in real terms, the same payments. New taxes will not be created.”

AMLO knows that he benefits from the credit that international markets give his country; if foreign investors treated Mexico the way that they treat Argentina, his plans would be ruined before they started. 

So what is not to like?

Well, I would worry about one thing: becoming a sacrificial lamb should AMLO need one. 

If (when) things go bad, even temporarily, he will need a monster to destroy. What better than a rich company? As long as his proxies can credibly claim in elite circle that it was a assault on an organization that specifically done did wrong and not an attack on the investor class, then he can have the best of both worlds. 

This is now happening.

Two days ago, the AMLO administration has moved against Altos Hornos de México (Ahmsa), freezing its accounts. The Mexican government has evidence that Ahmsa paid a $3.7 million bribe to a shell company owned by the infamous Odebrecht. Yesterday, Spanish authorities detailed Ahmsa’s CEO. Today, the Mexican government opened a corruption prosecution against Emilio Lozoya, the former Pemex CEO. 

To be clear, the prosecution is long overdue. It is astonishing that the Odebrecht scandal, which has toppled governments elsewhere, has been basically invisible in Mexico, despite Odebrecht testimony about bribes. There is fire behind the smoke. 

And while Ahmsa and Pemex are clearly in the wrong, the timing of the announcement and the decision to prosecute are clearly political. 

Meaning an investor’s primary fear should not be macroeconomic. Nor should it be contract enforcement. Rather, it should be that the government might choose to single them out in order to satisfy short-term political needs.

Fortunately, AMLO has more than enough strong domestic monsters to slay that he is unlikely to need to kill weak foreign ones. But for large investors, the danger is there ... especially if they did business with the previous PRI government with its spectacular levels of corruption. 

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.

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 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 ...

Observatorio argentino 38: Who needs checks and balances?

It is astounding to realize just how much power the President of the Argentine Republic possesses. 

This came home to me while I was checking over how President Macri recently tightened the immigration laws. Context: Argentina has a rather liberal immigration regime. (I am surprised even more Venezuelans have not shown up. I guess the limiting factor is that they need a job offer to get a work visa, although there are no limits on such offers.)

Last year, the Macri administration wanted to tighten the law. (See Observatorio Argentino 32.)  The changes were small, but they involved specific clauses in the legislation itself. You can find the original 2004 text here. So they could not be made via simple executive order.

In other presidential republics, the only option would be to go through Congress. Get the law changed. How hard that is depends on the country. 

An American president needs to be a hella politician. They need to convince a legislator to introduce the legislation; they then need to cajole the Speaker and Majority Leader to bring it to a vote. Even then, under the new 21st Century norms, they need at least 60 senators to vote yes. 

A Mexican president is a little more powerful, but only a little. Under the original wording of Article 71 of the constitution, they could introduce bills straight to the relevant committee. But by itself that gave them only a little more power than an American president. In 2012, however, a constitutional amendment changed Article 71 to give the president the ability to introduce two bills every session that go straight to the floor. Each house then gets 30 days to vote on them; if they don’t hold a vote, then the bill becomes the first item on the agenda in the next session. So congressional leaders can still delay things; and note that if they delay both initiatives the President cannot introduce any in the next session. So even with the additional post-2012 powers, they need to be good legislative politicians if they want to get anything done. 

A Colombian president is more powerful still. She or he can introduce bills and demand that Congress vote on them within 30 days, with no limits. But Colombian presidents are not legislators. They have no direct control over the amendment process. And, of course, Congress can just say no. 

But Argentina? Sweet Mary. Under a 1994 amendment to Article 99 of the constitution, the President can decree pretty much anything. The article is nuts. Read section 3 for yourself:

The Executive Power shall in no event issue provisions of legislative nature, in which case they shall be absolutely and irreparably null and void. Only when due to exceptional circumstances the ordinary procedures foreseen by this Constitution for the enactment of laws are impossible to be followed, and when rules do not refer to criminal issues, taxation, electoral matters, or the system of political parties, he shall issue decrees on grounds of necessity and urgency, which shall be decided by a general agreement of ministers who shall countersign them together with the Chief of Cabinet.

Within the term of ten days, the Chief of Cabinet shall personally submit the decision to the consideration of the Joint Standing Committee of Congress, which shall be composed according to the proportion of the political representation of the parties in each House. Within the term of ten days, this committee shall submit its report to the plenary meeting of each House for its specific consideration and it shall be immediately discussed by both Houses. A special law enacted with the absolute majority of all the members of each House shall regulate the procedure and scope of Congress’s participation.

Yes, that is as crazy as it sounds. Exceptional circumstances are not defined, and every president since 1994 has interpreted the clause to mean that they can make legislative changes to anything they want whenever they want. In theory, the joint standing committee has ten days to vote and then Congress can reject the decree. In practice, the committee often stonewalls or deadlocks and then the decree just goes ahead and becomes law, poof, just like that. 

How the hell did that happen? The 1853 drafters of the constitution did not set out to create an elected dictatorship. But they got an unelected dictatorship in 1930, when the military seized control. The Supreme Court remained in session, though, and it validated the decrees handed down by the junta. In 1947, the Court tried to block a presidential decree ... at which point President Juan Perón impeached all but one of the justices. You won’t be shocked to find that the restaffed Court ruled in Ziella v. Smiriglio:

The decree-laws issued by the de facto government are valued by reason of their origin and since they have the force of law they continue in force even though they have not been ratified by Congress as long as they are not repealed in the only manner in which they may be so, that it is to say, by other laws.

Still, only 15 decrees were issued between 1930 and 1982. When democracy returned for good in 1982, the new civilian presidents were loth to give up the power. Raúl Alfonsín used the power ten more times, mostly for small-bore stuff; his big decree replaced the peso with the austral. 

But Alfonsín was succeeded by Carlos Menem, a Trump before Trump, a Berlusconi before Berlusconi. Menem liked executive power. He started by increasing the size of the Supreme Court from five to nine. He then issued a decree that seized most bank accounts and replaced them with government bonds. People sued. 

But the packed court ruled that President Menem was entirely within his rights as long as Congress did not explicitly repeal the decree with another law. American legal scholars will be amused to know that the Argentine justices cited Home Bldg. & Loan Association v. Blaisdell (1934) as justification for a decision ratifying exactly the sort of thing that Hughes, Cardozo and Brandeis were trying to prevent.

The 1994 constitutional amendment creating the new Article 99 was an attempt to limit the decree power. As such it failed. Menem would go on to issue over 300 decrees. (Note again that before Menem there had been only 25 presidential decrees in the history of the Republic.) Congress would not even bother to create the joint standing committee until 2006. And it remains toothless. 

Consider the immigration reform that started this discussion. What happened after President Macri issued it? Well, the joint standing committee has only 16 members but it could not get everyone to attend. It voted 6-5 to support the decree on February 22, 2017 ... but because there was no majority its recommendation was never reported out. The decree continued as law. In October, a federal judge ruled that it was in fact law even though the whole Congress had never voted. (In this he had ample precedent, although Argentine judges are not formally bound by precedent the way they are in the Anglosphere.) He also ruled that the deportation of criminals was somehow not a criminal matter and thus the reform was kosher under Article 99. 

No matter how you slice it, it is unbelievable that a democratic president has the power to make law unless Congress actively says no. Giving Donald Trump the power of a Mexican or Colombian president would change little; GOP initiatives are hard to pass because they are unpopular. Allowing the President to force votes on them would not help. But giving Donald Trump the power of an Argentine president? Well, that is scary.

And that may be one of the prosaic reasons why Argentina is not particularly well governed. They really ought to do something about that Article 99, I think. 

Raising barriers to pay for a wall: we anticipate Trumpian logic!

Back in March of 2016, we suggested that a Trump administration could repeal NAFTA and claim that Mexico was paying for the wall via higher tariffs. A static back-of-envelope calculation implied that the U.S. would receive about $2 billion a year in additional revenue. Cutting that in half to account for falling trade volumes would give you about a billion. Thirty-year treasuries are currently yielding about 2.875%, so that billion-dollar revenue stream could plausibly be said to allow for $19.9 billion in borrowing to finance the wall, currently estimated to cost about $18 billion.

In reality, repealing NAFTA would not pay for the wall for two reasons. First, the economic dislocations would almost certainly reduce other tax revenues. Second, much (most?) of those tariffs would fall on American consumers more than Mexican producers. Third, inasmuch as Mexican producers will pay, many of those producers are American owned. Add that up, and repealing NAFTA will cost rather more than the additional tariff revenue.

But you could, in a static sense, claim that repealing NAFTA would pay for the wall. 

And now, in a fairly incoherent interview for the Wall Street Journal, the President has now jumped on the “logic” we laid out last March. “[Mexico] can pay for [the wall] indirectly through NAFTA. We make a good deal on NAFTA, and, say, I’m going to take a small percentage of that money and it’s going toward the wall. Guess what? Mexico’s paying.” 

The obvious read is that he intends to bring back tariffs on U.S.-Mexico trade (the small percentage) and will use that revenue to pay for the wall. If there was any consistency to this White House at all, I would take that as the death knell for the trade agreement.

But who knows what the President meant? Does he even know what he meant?

Bad news in tragedy avoided: building codes and the Mexican earthquake

Four days ago, I worried that the smaller force of the recent quake meant that we could not be sure that better building standards were the reason fewer buildings collapsed. There was some pushback in comments, noting that the epicenter was much closer to Mexico City in 2017 than in 1985. There was also some pushback on Twitter making a similar point. The Twitter exchange ended with Leticia Arroyo-Abad presciently saying that we could tell nothing from the size the quake because “it would depend on how different frequencies affect different structures.”

And she is absolutely right. The New York Times now reports that the closer distance of the 2017 quake meant that the shaking was far less damaging to buildings above six stories in height. Moreover, a 2016 study showed that many new buildings did not meet the minimum requirements of the building code. 

This photograph was taken in the same neighborhood as the new high rise mentioned in the NYT article, the one which collapsed. (More details here.) We know people who live in new buildings in that neighborhood. They dodged a bullet. There should have been no bullet to dodge, even in a 7.1 quake. That makes me angry. 

 

The jury is still out, but evidence is piling up that repeat of the 1985 quake would have killed far more people than the one Mexico City just experienced. The heads of bureaucrats and politicians in CDMX need to roll. The Line 12 saga shows that Mexico can have such a reckoning ... and once we are over counting our blessings and celebrating the heroism of the citizenry, this reckoning should be big. 

Good news in tragedy

Mexico City seems to have come through this earthquake much better than it did in 1985. Earlier, I worried that the reason might be that the quake was far smaller on the Richter scale; Dwight Dyer, in comments, pointed out that the epicenter was also much closer to Mexico City and the shaking was likely comparable. 

Is there any way to know if the tightening of building codes helped make Mexico more resilient? After all, it is not like the country stopped suffering from corruption in ‘85. 

One good sign is that it seems that no public schools collapsed during the quake; only 16 experienced major damage. The horror in the southern suburb of Villa Coapa happened when a private school fell in; subject to the new codes but not under the direct control of the federal government. The reason that the performance of the public school matters is that they were all supposed to be retrofitted after the 1985 quake. Their relatively good performance does not mean that the new codes were properly implemented ... but any public school failures would have been a damning condemnation.

Mexico has a lot to be proud of in how it has responded to the disaster. It also may have much to be proud of in how it prepared.

You take hope where you find it. 

El terremoto

I do not have much to say about the quake, save that it is horrifying, and that fortunately it seems (we hope) that no one we know has been hurt, although several have been terrified. There are some places where you can help. One place is below, managed by the sister of a good friend:

Ayuda para el terremoto en México https://t.co/yTwMHsew6t via @youcaring

— Marusia Musacchio (@tintachina88) September 20, 2017

I suppose I have a few more things to say. The first, for those of you who don’t know Mexico City’s geography, the worst damage has been in the center of the city. The reason is that the city was built on a lakebed: when the earth shakes, the ground liquifies. The most horrifying damage in ‘85 came about when high rises around the Alameda or in Tlatelolco came crashing down. The Metro stopped working. 412 buildings collapsed. 

The geographic scope of the damage seems similar, but so far only 47 buildings appear to collapsed, far less than in 1985. That said, this quake was an order of magnitude less on the Richter scale. So while I would like to believe that the better results are due to better building codes, I am sure that this is true. 

There has not been much looting or rioting, but two forms of criminal activity have been reported. The first seems to have been a spate of carjackings aimed at cars stuck in traffic, particularly out around Santa Fe. There have also been scattered reports of people entering homes disguised as rescue workers and robbing the inhabitants. I will predict that these reports turn out to scattered exceptions, both because the authorities seem to be responding to them with alacrity and because social disorder generally tends to decrease in disasters, as long as the authorities avoid general food and water shortages. 

Godspeed, Mexico City. You are loved. Please give above if you can.

More evidence that Trump caused the peso to fall

First, pray for Miami. I have a love-hate relationship with that city, even deeper than my love-hate with Texas. I would not want to live in that entire metro, not in Miami-Dade, not in Fort Lauderdale, not in Palm Beach. But my family is there along with seven million other people. They have not evacuated. But my brother is sheltering the families of two of his daughters at his house outside the flood zone; the third is an R.N. and she is on call, along with her police officer husband. 

Second, allow me to present some evidence that Donald Trump was indeed responsible for most of the weakness in the Mexican peso since he arrived on the political scene in mid-2015. The Mexican peso has strengthened over the past year. When we were there this summer, the peso rose above 5½¢ and stayed there. It had plunged below 5¢ after the election. This was good for Mexicans who wanted to buy imported goods, and I have to admit that Mexico City is still very inexpensive, but still. 

The rise means that the peso is all the way back to its level of January 2016, before Donald Trump began winning any primaries. It is not quite back up to the 6¢ level of June 2015, when Trump rode down that escalator. That is probably due to residual uncertainty regarding the NAFTA negotiations: markets are partially pricing in the chance that the negotiations will fail, thereby exposing Mexican producers to WTO tariffs when exporting to the United States. 

And so, a prediction. If NAFTA comes out intact from the negotiations, then the real exchange rate of the peso will return back to the level it had back in 2015. 

Do note that candidate Trump could move the market with his words; President Trump much less so. Bets are being hedged, but nobody believes that White House words carry any information anymore. 

Kiddie fútbol and the Trump administration

My son had a bit of culture shock in Mexico City, where he learned that 4-to-6 year-olds there take fútbol seriously. None of this everybody-wins bullshit we have back here in D.C. Nope, kids in Mexico play fút the way kids here play basketball. My son overreacted to that realization: he started to play dirty, like Luis Suárez dirty. The professors (that is what the coaches are called!) could deal with it, but the other kids would complain to my wife and me after practice.

Fortunately, nobody ever attributed the boy’s dirty pool to his nationality. He scored a lot and made saves, albeit diosmio the red cards. Whatever image the other kids had of American children was not affected by this crazy American fond of fouling, handballs and other Suárez-like program activities. He was the youngest kiddo out there and represented nobody but himself. (And maybe his parents, but the other kids’ parents came across as amused and sympathetic, not judgmental. Much like the reaction of soccer fans when they talk to Uruguayans about Suárez. Which is fortunate for Uruguay’s international reputation.)

And that brings us to the impact of another individual at about the same maturity level as my 4-year-old son: President Donald J. Trump. Earlier this summer, I noted that there seemed to be very little anti-Americanism among middle-class chilangos, certainly when compared to the wave of hate-America that passed over the same group in 2001-02. Mexicans realize that the current American president is unpopular and does not represent the American people as a whole any more than my son represents all American children.

A former Mexican ambassador to China initially shared my views but now fears that the Trump administration is doing permanent damage to Mexicans’ view of the American people:

I believed Mex people can tell diff btwn Trump & Amer people & would revert to normal once he's gone. Now afraid damage will be long lasting

— Jorge Guajardo (@jorge_guajardo) August 27, 2017

How can we know who is right? Polls are not helpful. The standard question asks about Mexican views of the United States, and those views, obviously, depend greatly on the current government. Consider the below results from the Economista poll. They are obviously influenced by the policies of the American government-of-the-day. The U.S. was highly unpopular in early 2004, only a year after Gulf War 2 and while the Iraq insurgency dominated the headlines. By late 2005, that had passed from Mexican headlines (albeit not American ones) and people’s view of the United States recovered.

I would like pollsters to ask Mexican respondents a different question: “What is your view of the American people?” or “What is your view of American society?” Asking them about the “United States” invites conflation with the current administration. What we want to know is the damage that Trump has done to the relationship between the two peoples, not to the image of his administration.

Does anyone have any data on this? I would also like to know qualitative impressions. Please, comment below.

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.

Departures

With four of us, we will be calling a carshare to take us to the airport. I know, right now “Uber” has become a generic word, the way “Fridgidaire,” “Hoover,” and “Xerox” used to be. Since I dislike the company with a passion, however, I will do my part to insure that it remains a proper noun. “Carshare!”

It is slightly easier than calling a radiocab, I suppose, and it does insure working seat belts in the back. I am of mixed minds, to be frank. In Mexico City there is an upwards redistribution of income, since carsharing drivers tend to be richer.

And I miss taking the subway to and from the airport. It’s an experience that most middle and upper-middle class chilangos avoid, for reason that I do not understand. You do have to traverse the entire terminal from the international area to get to the station, but it isn’t that big a terminal: call it twelve Manhattan blocks, a little over a half-mile. You will also have to change trains at least once, since Line 5 runs a peripheral route down the middle of the Interior Circuit expressway. But unless you’re encumbered with bags or headed beyond the reach of the metro, it’s cheaper and faster than taking a taxi.

So why don’t more people do it? Well, practically, the station interchanges are a major PITA. They are long labyrinths. Not unpleasant, but not efficient. The reason the connections between lines are so ridiculous is that Mexico City, wisely, decided to build its metro on the cheap. Sure, you could have had lines dive under each other with platforms carefully laid out to make transfers quick and easy — the Washington Metro is pretty good in this respect — but building cathedrals to urban transit is not how Mexico City rolls. So other than the first three lines (1, 2, and 3, easily enough) everything is jury-rigged.

The most relevant connection for this post is La Raza, which connects Line 5 (the airport line) to Line 3 (a north-south line that runs through a lot of nice middle class neighborhoods). The damn thing is over a half-mile long (like the airport) but rather less pleasant. The first time you do it you will notice the science museum laid out along the walls, covering the history of the universe. The 78th time you do it, you will start to feel like the transfer has taken you the entire age of the universe. I can see this being a deterrent, but come on. By the time you have memorized the panels on the evolution of the cockroach, you can start people watching instead.

The wildest is Chabacano, which connects lines 2, 8, and 9 in wild mess of tunnels, bridges, and escalators. Up-down up-down up-down, great for kids, less so for commuters. But the brutalist architecture is so cheap that it’s charming. In fact, you might recognize some it: they filmed the subway shootouts in Total Recall there. Click the link. The station really is laid out just like the chase! No directorial tricks; they just had Arnie make the transfer. With people shooting at him.

The weirdest is Ermita connecting Line 2 and the newest Line 12. The corridors zig and zag like a protractor. I have no idea why, but if I had to guess, it was much cheaper than just barreling a straight connection through. Kinda triples the distance.

The best is Zapata. Just take it. Much more interesting stuff on the walls than the science museum at La Raza.

And the longest? Oh sweet Mary mother of God, Atlatilco. It connects Line 8 and Line 12. It goes on forever. I did it precisely once, just to see it. I will not voluntarily do it again.

I could go on.

Anyway, the long La Raza connection is not why few people take the airport line. Nor is it crowding. (Want to go to Condesa? Line 5 to Line 9. Annoying transfer; you gotta climb a lot of escalators. But it is still crazy fast and there is very little crowding, even by D.C. standards.)

No, it is just that too many upper middle class chilangos still feel weird on the metro. You know, people selling you candy and phone chargers and music CDs. (I do not understand that market.) Which is too bad.

But which will soon change. Mexico City is building a new airport right outside the city. The talk is of extending Line B to it ... but the real project will be an express train running from Observatorio, where the new commuter rail from Toluca will come in, out to the new airport. It will not be part of the Metro system (although it will connect to it) and it will cost a lot more than the current 29¢. So the middle class will take it, but the poor will not, and so transit will remain class segregated.

It also looks like the airport might be open before the airport connections, which is going to be a Godawful mess, but that is another story.

What is going on in Tamaulipas?

I may have a hypothesis. But that is for later.

For now, consider this private report from June. Roughly 400 businesses in the three rural counties (Güémez, Padilla, and Hidalgo) right north of Ciudad Victoria, the state capital, went on strike for 72 hours to protest rising extortion. The area mostly grows citrus fruits, but nowadays that isn’t too humble. Three separate organizations: the Gulf Cartel, the Northeast Cartel, and the Zetas are trying to muscle into the business. 

Actually, that would be the “Old School” Zetas, the Northeast Cartel being the new school Zetas. They split from the organization after the federal government took out Omar Treviño. It might be sort of ironic, given that the Zetas arose from a split with the Gulf Cartel. 

Extortion and kidnapping (which is what happens if you do not pay the extortion) has caused the value of production in those counties to collapse by 60% ... at least according to scattered reports from producers. Reputable statistics are hard to come by. Abraham González, Tamaulipas state director of the National Federation of Commerce Chambers (Fecanaco) on 21 June said that organised criminal groups were aware of the harvesting season and were attempting to extort businesses all along the agricultural value chain. He also said that they have warned producers to halt the collection and transport of agricultural goods “to reduce their exposure to incidental gunfire.” Which you sort have to love, as a phrasing.

Governor Francisco García (PAN) seems helpless in front of the chaos. The closest he has come to a plan has been to announce that the police will be stepping up highway patrols and surveillance in the three affected counties. Which may come in handy, considering as the main road from Monterrey passes right through Hidalgo and Güémez. But which is also a little sobering, considering as I have got to drive said road in January. I know, I drove through worse areas in 2010. And I was working for Uncle Sam in Afghanistan in 2006-07. But there were some differences:

1. Being in a dangerous area both times was the point of the exercise. I was there because it was dangerous. I had a job to do that maybe just-possibly if-lucky could have resulted in both areas becoming less dangerous;
2. Despite (1) being entirely valid, I also had an excess of machismo back in 2010 that has  recently burned off, perhaps because;
3. I have kids now.

Since I am not going there because I want to study or combat the violence,  the idea of travelling into a violent zone on business is somewhat less attractive than it was seven years ago. And travelling there on business seems to carry a modicum of risk these days. Consider the following bottom-line recommendation to foreign businesses desirous of operating in the state: 

Chronic insecurity is likely to continue posing challenges to the operations of foreign companies in Tamaulipas. Enhanced threats of extortion, kidnapping, vehicle and cargo theft, robbery and incidental exposure to gunfire will persist. Additionally, roadblocks erected by organised criminal groups or security forces can occur on occasion, affecting logistical operations. However, this does not represent an insurmountable obstacle to doing business in the state, providing the requisite security measures have been put in place.

Well, if you are a scholar wanting to drive down from Monterrey to work in some archives, what exactly would be the requisite security measures? 

Painting a scary picture for President Trump: civil asset forfeiture edition

In comments, NYCMT wrote the following: “You know what might survive a presidential pardon? A civil asset forfeiture claim.” I have now talked to him in depth about what this meant.

The federal government has a lot of power to seize property involved in the commission of a crime. 18 USC 981 (a)(1)(A) states “The following property is subject to forfeiture to the United States: any property, real or personal, involved in a transaction or attempted transaction in violation of section 1956, 1957 or 1960 of this title, or any property traceable to such property.” Sections 1956 and 1957 deal with money-laundering, tax fraud, and dealing in assets that are the result of crimes. 

The key is that civil forfeiture is undertaken against property, not an individual! You can pardon a person from a criminal act but you cannot pardon their property against a civil judgment. Moreover, forfeitures are undertaken under a preponderance-of-the-evidence standard. Since a pardon would essentially admit as true all of the facts of the criminal charges involving the property, a civil asset forfeiture proceeding would be a slam dunk for a federal prosecutor.

NYMCT laid out the following hypothetical. On the last day of the Trump presidency‎, right before the Senate takes its vote on the articles of impeachment reported out of the House, Donald Trump signs several hundred pardons, not least for himself, but also for his associates in the Trump Organization, his counterparties in a number of different transactions, and the principles of all financial firms whom may have been involved in any of his financial transactions. The pardons cover not only any financial crimes but all violations of the various sedition, conspiracy, and obstruction of justice statutes.

Well, if the pardons specify all illegal conduct against the United States then they become open ended gateways to factual stipulations in later court proceedings. If one single transaction involves the Trump Organization, even as an intermediary, then you have trouble. I dunno, imagine a purchaser of first resort for a condominium in Trump Tower funded through a secured credit facility that relies on underlying assets like Trump Tower or Trump Village in Coney Island. If that credit facility was repaid with money traceable directly to the proceeds of crime, then it satisfies the materiality requirements under 18 USC 981(a)(1) and the feds can seize the securing assets.

This does not mean that President Trump is going to lose his stuff! I have no idea what crimes, if any, have been committed. What it does mean is that if Mueller uncovers financial chicanery, then President Trump cannot pardon himself out of trouble. (Assuming, that is, that the courts interpret the pardon power as allowing the president to pardon himself.) In addition to potential state prosecutions, he would run the risk of losing many of the things that really matter to him.

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. 

¿No hay racismo en México?

I moved to Mexico in 1996 for my doctoral research. I lived in Colonia Alamos, a declining middle-class area, and went to a mediocre gym in Narvarte. I have a vivid memory of one of my workout friends there telling me, quite emphatically, the title of this blog post. The context was a discussion of what he did not like about the United States.

Mexican families, like American families, and border-crossing families, show a wide variety of skin colors.

My wife has noted that she gets far less unwanted attention for her skin color than she received 13 years ago. No one asks to touch her hair. Stares and catcalls have also disappeared.

The only strange incident occurred in a Comercial Mexicana on M.A. Quevedo Boulevard, where an elderly gentleman asked my wife where she was from. When she said, “Estados Unidos,” he insisted that she wasn’t telling the truth. “No, where are you really from?” That one was odd. Mexicans know full well that one out of eight Americans are black, and the former President of the United States was not exactly an obscure figure down here. But really, that was it.

Well, that was it unless you count some odd moments where people, always female, and usually young, want to touch my daughter’s hair. Now, in Mexico City strangers feel free to rub my kids’ heads all the time. Any kids’ heads, in fact. What makes it a little strange is that when it is just me and the girl, people just rub her head; when it is me and my wife and our girl, well, then they ask permission. You could interpret this in several different ways, I think.

And anyway, these are anecdotes! Since when is this blog about anecdotes? This blog is about science™! So let’s do some science. 

Eva Arceo (CIDE) wanted to see how skin color affected life outcomes in Mexico. So she persuaded the Mexican statistical agency (INEGI) to add a question to the 2016 Modulo de Movilidad Social Intergeneracional survey. INEGI provided each respondent with a color palette and asked them to match it to their skin tone.

The palette (at right) as Dr. Arceo notes, is far from perfect and self-identification is not the best way to go about this, but the palette was developed by the Proyecto sobre Etnicidad y Raza en América Latina in conjunction with scholars at Princeton University, so it is what they used. (I think my skin color is an H? Maybe I? Like I said, it is not perfect.) 

Well, here is the histogram of the results. 

 

And here is how skin color lined up with years of schooling:

 

Here is how it lined up with relative income:

 

The results held when she accounted for the educational level of the respondents’ parents and their family income when the respondent was 14 years old. 

There were a few weird results that I would not take too seriously. (The author did not.) For example, when you account for parental characteristics, the educational level of people with skin color G or H spikes up compared to everyone else, including paler people. (I suppose that as an overeducated person of skin tone H, I should be more predisposed to believe this.) More plausibly, the rise in income level by skin color becomes steeper and more monotonic when you adjust for education and parental characteristics.

This result jibes with experimental studies. Rosario Aguilar (CIDE) found that voters were more likely to support white candidates with the same platform; Eva Arceo (CIDE) and Raymundo Campos (Colmex) found that employers were more likely to call in white applicants for interviews. And as Prof. Arceo points out, it should not be a surprising finding in a country where advertisements still commonly call for “tez blanca.”

But it is a little depressing. 

The unbearable lightness of Mexican law-and-order spending, seven years on

Seven years ago, I drove around Mexico’s most dangerous towns. In January of 2018, I may need to do so again, albeit for different reasons. So the topic of violence is on my mind, even if Coyoacán is still as safe as ever. 

After that trip, I noted that Mexico spent shockingly little on law and order, defined broadly. Total spending on police, courts and defense came to only 1.1% of GDP, less than a third of the level in the United States. Seven years later, as violence once more swings upwards, Mexico spends ... all of 1.4% of GDP on law and order.

Spending data come from page 20 of this new report; figures divided by GDP from here. 

The report digs in a little further with international comparisons. Consider how much Mexico spends on police (seguridad interior, as opposed to justicia or seguridad nacional.) Well, 1.4% of all federal spending goes for that ... compared 4.7% for most OECD nations and 6.2% in Italy ... and Italy, unlike Mexico, is a country which has managed to contain its serious problem with organized crime. 

Since I was not sure how Ethos made its calculations, I dug up the OECD numbers on general government expenditure by function. (General government includes subnational governments.) For comparability, I converted everything into a percentage of GDP. 

What has the money been spent on? The number of federal police has almost quadrupled between 2006 and 2015, to 43,724 from 11,663. Real average salaries only rose 9% between 2010 and 2015. (Page 34, deflated by the INPC.) You might wonder how such a mighty expansion in payroll was sustained on such a small increase in spending, but note that the federal police payroll takes up only 5% of all security spending.

So more cops, but becoming a cop did not become more attractive (quite the reverse, given the rising danger) and the quality of police officers did not rise. Now, to be fair, federal federal police officers are well-paid by Mexican standards; approximately US$15,400 per year. But state and local police are paid much less: US$7,800 on average. The highest paid are in Sinaloa, reaching barely US$10,900. 

And by other measures, the Mexican criminal justice system is a mess. The U.N. Office on Drugs and Crime (UNODC) collects figures on prosecutors and judges. Per 100,000 people, Mexico had 6.8 prosecutors and 4.2 judges in 2013. Call it 11 per 100,000. That same year, crime-ridden Scotland mustered 9.9 prosecutors and 4.9 judges ... 15 per 100,000. And Italy enjoyed only 3.1 prosecutors but 17.1 judges ... 20 per 100,000. (In the Italian system prosecutors are considered judges; it is not clear how UNODC made the distinction.) 

Combine that with badly-trained cops, and you have a mess. 85% of all criminal cases in Mexico were commited en flagrancia; meaning that the cops caught the perp while he or she was committing the crime. And in the rest, most cases have been resolved by beating confessions out of whomever the police pick up. In theory, new reforms have made confessions inadmissible save in the presence of a defense attorney (page 20). In practice, it is not clear that anything has changed, especially given as the ability of the police to investigate crimes is practically nil and bribery still reigns supreme.

I still tend to think that Mexico can solve its crime problem. I also tend to think that while throwing money at it is not a sufficient condition for doing so, it is a necessary one. 2.6% of GDP, like Colombia? Five percent, like nowhere? I do not know. But more seems clearly necessary. Am I wrong?

Crime in Mexico is rising

Violence has been ticking back up in Mexico since 2014. We are now back to the levels of the bad old days.

Below you can get an idea of the changes in the national homicide rate over the past quarter-century. For comparison: the U.S. is now around five, nationally. During our bad-old-days in the 1970s and 1980s, it was around 10. 

The distribution of violence in Mexico, however, is very different today than it was during the 2007-12 explosion.

From the indispensible Diego Valle-Jones comes this map. In the first half of this decade, the violence explosion centered on the northeast: Chihuahua, Coahuila, Nuevo León, and Tamaulipas. Violence has declined markedly in the first three of those, although it remains much higher than its pre-2007 level.

In Baja California, however, the much-ballyhooed violence decline has entirely reversed itself. B.C. became a paragon of the mano duro: the idea that the simple use of violence by the state could work to reduce violence by criminals. (This link is to a New Yorker story from 2010 that was very skeptical of the success. Philippine readers, take note.) Colima, a small coastal state, has spiraled up to levels seen only in wartime. The Bajío region is seeing a steady and rapid rise in violence. Elsewhere violence is also generally up. Slowly up, but up.

So what is happening? The best hypothesis is that this is the result of the “kingpin” strategy. Dismantling criminal organizations root-and-branch is hard. Much easier is to go after the leadership. Decapitate the organization. Kill the head and the body will die.

Well, not quite. Sending Shorty to the United States created splits in the Sinaloa Federation organization. That has embolded Jalisco New Generation — made up of former enforcers for Sinaloa — to attack its old bosses in Tijuana, B.C., and Manzanillo, Colima. (Manzanillo is Mexico’s biggest port.) Meanwhile, Sinaloa itself is starting to fracture. Shorty Guzmán’s sons were briefly kidnapped in Puerto Vallarta; outside observers suspect it was a signal from Dámaso “The Graduate” López and his eponymous son. The Graduate was Shorty’s former right-hand man. 

In the northeast, the Zetas have been weakened and (mostly) pushed out of Tamaulipas, but they’re still there and fighting. Including fighting each other as the organization has fractured. The Gulf cartel is still there, but also weakened. In Michoacán, the Knights Templar are gone, having failed in their attempt to become Mexico’s version of the Cosa Nostra, but leaving a chaotic landscape of armed gangs in their wake. 

Recent events in Mexico City have shown just how counterproductive the kingpin strategy is turning out to be, but that is the subject for another post. 

Greater or Lesser Mexico is hella more likely than the Mexico what is

People tend to think of what is as what had to be. But that is not true.

The U.S.-Mexico border as-is, the place the President wants to build a wall, is a very low probability event. 

The Mexican-American war is a veritable cornucopia of counterfactuals. The United States could very easily have avoided the war. A lot had to go wrong for Clay to lose. Unchallenged vote stealing in Louisiana, dumb flip-flops on Texas annexation (see below), an idiot flirtation with nativism. The Clay campaign made more errors than the 1901 Detroit Tigers. Had 2,553 New Yorkers switched their vote in 1844, Henry Clay would have become president. 

Now, Clay was not a die-hard opponent of annexation. In fact, he wobbled on the issue, writing letters against annexation but flip-flopping during the campaign. (This cost northern votes without gaining southern ones.) That said, he insisted that annexation could only happen if there was an agreement with Mexico, which seems, well, unlikely. It is true that the annexation resolution passed in February 1845, before Clay would have taken office, but the Senate would not have taken it up had Clay been the president-elect.

But if that does not convince you, consider the legislative history of the annexation resolution. The Senate Foreign Relations Committee initially voted down the resolution. What revived it was an amendment from Senator Robert Walker (D-Mississippi) that gave the president the option of admitting Texas under the resolution or negotiating a new treaty. The amended resolution made it out of the Senate on a 27-25 vote with the expectation that President Polk would negotiate a new treaty. In a shocker, outgoing President Tyler on his last day in office invited Texas to join without a treaty. Polk kept the offer open and the rest is history. But in a history where Henry Clay becomes president, none of this happens. And as we all know, the United States of 1844 was about to be consumed by its own internal divisions: if the Clay administration survives to January 1853, then the window for annexation would close. (Clay died in 1852, but the point stands.)

Think on that one a moment. A close easily-swung election (kinda like the one in 2016) resulted in the end of the Republic of Texas and the transfer of the whole southwest into the United States. Maybe possibly the U.S. would have eventually gotten northern California, but the history would have been different and Texas would be outside the United States to this day. 

What about the other direction? As recounted, the current U.S.-Mexico border is unlikely in the extreme. Had Nicholas Trist not disobeyed orders, it would have been further south. But how much further south? This post discusses President Polk’s musings and discusses a counterfactual border. But Polk was just spitballing to the cabinet. What do we know about how much territory he would have claimed?

Well, we know two things. First, the point of expansion was to get more slave states. That means places where cotton could grow, labor was scarce, and the elite open to slavery. Second, we know that southern Senators tried to amend the Treaty of Guadalupe Hidalgo to include more Mexican territory. Their proposal can be found here, and a map is below:

It includes all of what became the Gadsen Purchase and a big chunk of northern Mexico. Not coincidentally, it is the chunk of the northern Mexico that was in fact good cotton land. (Chihuahua eventually became a good place for cotton, but it was too dry and isolated for that in the 1840s.) It is also the chunk of Mexico that tried to join the Confederacy in 1861. (1)

Here is a close up of the northeastern part. I have no idea if it would have been admitted as one or two states:

In short, once you get a Mexican-American War, it is more likely than not that northeastern Mexico winds up inside the United States: Monclova, Saltillo, Monterrey, all American cities. 

Today the border feels inevitable. But it was not. As late as 1844, the current line would have been a very unlikely outcome.

The beginning of the end of parking?

Mexico City is creaking under the number of cars on its streets. For details, see this post. So what to do about it?

The metro is already a modern miracle and is being expanded. Notably more middle-class people appear to be riding it. It should be expanded faster, but there are practical limits to how fast that can be done. 

Road-pricing is a distinct possibility, but would be politically, administratively, and technically extremely difficult in this town. Mexico City is much nicer than it was, but it still ain’t London. Although they have managed to close off major streets in the city center on Sunday.

Which leaves you with ... limit parking! In a brilliant move, the Ciudad de México just moved to limit parking spaces in new residential construction. Current regulations mandate minimums, on the assumption that people will drive no matter what. (For offices, it was one spot for more-or-less every 300 square feet.) The newest office tower downtown, for example, has more than 3,000 parking spaces.

Now there are no minimums. The maximum is three per apartment (which is rather high) or one per 300 square feet of office space. (See pages 7-8 of the new regulation here.) You can build more, but you need to pay a tax. 

The literature says that this should discourage driving. European cities have de facto parking limits, and they do indeed have fewer auto trips than American ones. The question for Mexico is whether this is enough to keep all of the future middle class from commuting by automobile. The question for the rest of North America is whether this policy can or should be emulated.

Thoughts?

A new chapter in the NAFTA saga? Likely not

NAFTA renegotiation is upon us. The newspapers here have plastered their front pages with stories about the united front Canada and Mexico are putting up against the elimination of Chapter 19.

What’s Chapter 19? It’s the part of NAFTA that mandates binational panels review any anti-dumping actions imposed by one NAFTA country on another. Without it, NAFTA countries would be at the mercy of national courts if they contested anti-dumping actions aimed at their exports. You can find the details of all 118 NAFTA trade cases here; roughly 69% of all cases have been against the United States. (More stats at this link.) 

It turns out to be remarkably hard to calculate success rates. Who wins more in these panels? Someone would need to go through all 118 decisions and figure that out. (The headline decisions often split the difference, so it would be a judgment call.) That said, the data here let me do a quick-and-dirty estimate: in cases brought by Canada against the United States, the panel either made no decision or found in favor of the U.S. around 65% of the time. Winning!

There is a case to replace the ad hoc panels with a genuine international court, European Union style. But that is not on the cards. Another possibility would be to trust domestic courts: let American exporters put their faith in Mexican courts and Canadians trust American ones. But national court systems are slow and expensive, so Mexico and Canada have made it clear that they would rather junk NAFTA than rely on domestic courts to make decisions.

Now one can imagine all three countries setting up special domestic trade courts to review agency decisions on behalf of NAFTA plaintiffs in a timely manner. But the Trump administration has shown no sign of wanting to reform American courts along those lines and it unlikely that Canada and Mexico would accept such reforms regardless. It really looks like Chapter 19 or nothing.

Moreover, Mexico is putting its opposition on autopilot. The Third Commission of the Permanent Committee of the Mexican Congress (don’t ask, it gets complicated) has gone on record rejecting changes to Chapter 19. A full congressional resolution could tie the hands of the Mexican negotiators ... which might be, in fact, exactly what the Mexican government wants Congress to do. Consider two idiots playing chicken on Francis Louis Boulevard at 3am in Queens. One of them throws their steering wheel out the window. That idiot would certainly have an advantage in the resulting game ...

I would bet that the Trump administration quietly backs down, but with this man in office, you never really know. Still, FDI into Mexico is keeping pace, around $25 billion so far this year, similar to the $27.4 billion in 2016, $33.3 billion in 2015, and $27.7 billion in 2014. Similarly, the peso has recovered to around 17.8 per dollar, pre-Trump levels.

So, yeah, our terrible negotiator-in-chief is not likely to win this one, which is good.  

More subway geekery, Mexico City edition

I know, up north the Senate is debating ... uh ... something that might impoverish and sicken millions of people, while the President burns through every norm we have left. But I am south of the border, so let’s talk about subways.

Back in 2014, we blogged about the Line 12 mess. (The above station is on Line 12 where my boy goes for soccer practice.) The problem was that the city decided to use steel rails and overhead catenary power lines instead of rubber tires and third rails. The trains proved too heavy for the tracks, 15 kilometers of which had to be replaced. In addition, curves on the elevated structure had to be widened and the camber reduced from 160 millimeters to 100. It was a design mess, but as I argued in the post, the Mexican government actually comported itself fairly well in the aftermath.

The question is, why did they decide to switch designs? This wasn’t the first steel-wheeled line constructed in Mexico City, but Line A runs out the eastern suburbs mostly on a straight ground-level right-of-way in the middle of a freeway. Moreover, Line A was originally designed to be a light rail that would link into the end of the Metro but not really be part of it. Its cars are a full 30 centimeters narrower than regular metro cars and weigh 88 tons less. (The history of Line A can be found here.) In other words, the decision to build Line 12 with overhead wires and steel rails had nothing do to with the earlier experience with Line A.

The decision seems to have been based on a 2000 study for the World Bank, which compared Madrid’s successful subway roll-out to experiences in Mexico, Caracas, and Santiago. Madrid used steel and overhead and got it done cheaper. The reason was simply that most of its lines had historically used overhead wires and its maintenance operations were designed around them. (You don’t have to turn off the power to protect maintenance crews when you have overhead lines.) So it could use the same sort of railcar that it always used. In that sense, Madrid went with overhead for pretty much the same reason that Mexico City should have stuck with third rail. 

As for steel versus rubber, the study provided evidence that steel wheels should be cheaper. Steel wheels last longer, so there will be some savings over the long-run. And you need three sets of rails for rubber-wheeled trains, because you need guide tracks to keep them on course. So steel should be cheaper. But is it? As the study itself noted (page 77), Mexico’s rubber-wheeled trains can operate on 7% slopes and make turns with 105 meter radii, compared to 4% and 250 meters in Madrid. If Line 12 had used that technology, it would not have needed to be rebuilt.

In short, it looks like the Line 12 decision was the triumph of hope over experience. There were plenty of reasons to believe that steel wheels and overhead wires would be a better design if you were starting a subway from scratch. But Mexico City was not starting a subway from scratch. It was extending a well-established system. And its only experience with steel-wheeled cars was on what was a suburban light rail with much smaller cars and trains than the rest of the system. 

Still, the ending so far has not been unhappy. As we noted in 2014, the cost of the new line is still reasonable. Moreover, in May 2017 the government obtained a further discount of 2.1 billion pesos from the consortium built the line.  And ...

... the new trains, they are very nice!

President Trump cannot pardon himself out of trouble

The Trump presidency is different, as I note when I am outside the United States. Back during the Bush administration, foreigners would often (and without prompting) verbally criticize traveling Americans for the actions and policies of our President. And when that happened to me, which was often, I often felt an emotional need to defend President Bush, no matter how terrible a president I actually thought he was.

With the Donald, though, not so much on either tip. Oh, he certainly comes up all the time, far more than President Bush ever did. But nobody — not in the Netherlands, not in Uruguay, not in Argentina, and not in Mexico — has ever put on an accusatory tone. It is more pity, sometimes shaded with incredulity. That may be why I have never felt any urge to defend the Trump administration ... but I suspect that even should I hear a vicious hostile angry attack on his presidency my reaction would be, “Well, yeah.”

The worst that has happened to anyone I know is the American daughter of a good friend, who got a nasty note in her Mexico City school calling her a “Trump lover.” (Both the friend and his daughter are also Mexican citizens, but kids don’t care about that.) Considering what has been happening in U.S. schools, that is small beans. And compared to the Bush administration, it is nothing.

It is something to have Mexicans pity you for living under a corrupt president. In Mexico, one of the biggest problems that officials seem to be immune from prosecution. They cannot be indicted while in office and it is damnably hard to prove anything once out.

Which brings me to the recent news that the Trump administration is considering issuing pardons for the president and his family. Such pardons would be unprecedented in American history, but it is not clear that they would be unconstitutional. I certainly think that it would have been ridiculous for the framers to have given the president the power to pardon themself, but the question is open.

Except it might not matter! Donald Trump is a resident of New York State and the owner-proprietor of 500 different partnerships, LLC’s, S-corps, and pass-through proprietorships forming the Trump Organization business collectivity domiciled in New York. Trump’s business activity occurs in New York State as a consequence of his residency and offices located in New York, and his financial transactions, insofar as the financial entities within which those transactions take place are under New York State supervision and regulation, are subject to New York State law.

Imagine that the Mueller investigation uncovers some sort of financial chicanery. Well, I talked to Dave Greenbaum, a lawyer from Trump’s home borough (not pictured above!), and as you might expect any such evidence would be admissible in New York courts and give Eric Schneiderman all the ammunition he needs to go after the President. (At least once he is out of office; beforehand is less clear.) 

The easy stuff: misstated income and fraudulent income tax returns. If Trump has been using his labyrinthine corporate structure to hide income, he is liable for New York income tax fraud under Tax Laws 1805 and 1806, which would be felonies. Even if he pardoned himself for the federal income tax evasion, he would still be liable for state tax evasion.

Getting harder: securities and business entity fraud under GBL 23-A-352 and 353, aka the Martin Act. If Trump has been using his business entities to shuttle debt and conceal liabilities from lenders, limited partners, and other persons, he is potentially liable for felony misrepresentation under identical elements as SEC rule 10b-5. (Misrepresentation or Omission, and Materiality). A federal SEC violation would be chargeable under New York law, as long as New York had jurisdiction, and since Trump is a New York citizen and most of his business happens in New York, despite choice of law clauses, New York would have jurisdiction.

Harder still: New York no longer has a treason statute, but any foreign payments structured so as to defraud co-owners or partners or secured parties would serve for bank fraud or Martin Act charges. And New York has a conspiracy law just as strong as 18 USC 371. Plus, in New York you can be charged for attempting financial crimes, even if you fail to pull it off.

Barely possible: New York has two relevant laws on obstruction of justice: NYPL 195.05 and  NYPL 205.50 et seq. 195.05. “Obstructing govt. admin in the second degree” consists of affirmative action to prevent a public servant from carrying out their official duties and it is a Class A misdemeanor. 205.50 is “hindering prosecution,” which can range from a Class A misdemeanor to a Class E felony. That law could conceivably be used to prosecute the Bharara firing, given that Bharara was working with the New York attorney general to trawl Trump payments through New York banks, but it would be hard to argue that it really obstructed government administration or hindered prosecution in New York State.

Basically impossible: payoffs that take the form of capital investments in Trump businesses unless there is an obvious quid pro quo or you can use them to charge conspiracy to commit tax fraud. It is barely possible that if Trump uses a vehicle as a conduit for a payment that he only partially owns, he could be charged with a Class B misdemeanor for ripping off a co-owner.

In short, if the Mueller investigation uncovers untoward things in the President’s business dealings,it is highly likely that the State of New York will come after him. It will not be as effective as a federal prosecution, but even if he pardons himself and Congress fails to remove him, he will still be in no small legal peril.

¿Viva Nueva York?

The fall in the Mexican peso was good until Trump came along

The Mexican peso began falling well before Donald Trump came on the scene. Opinion varies on Trump’s responsibility for the fall between April and November 2016 (Kevin Drum and Boz make good cases that he did not) but his administration is pretty clearly responsible for the gyrations after Election Day.

Let’s grant that the peso’s fall from 7.4¢ in November 2014 to 5.4¢ in November 2016 had nothing to do with the U.S. election. Was it a bad thing for Mexico?

Well, let’s start with inflation. Did the fall ignite inflation?

Wait, what? Let’s close in on that:

Inflation fell during most of the depreciation. It only started to rebound around the time of the U.S. election. I suppose it is possible that there is some model here where inflation is somehow stored up and then rebounds, but absent a big rise in inventories I do not see the mechanism.

Maybe the reason inflation fell was that the central bank tightened in the face of the depreciation? Let’s then look at interest rates:

The story is similar. Interest rates stay low through most of the depreciation. They only start rising in 2016. The central bank seems remarkably unworried by the fall until the latter half of 2016 ... which (perhaps coincidentally) is around the time that Donald Trump’s role in the depreciation becomes a point of contention. 

What about consumer confidence? Did the peso fall sap the animal spirits of Mexican consumers? Maybe the central bank did not need to raise interest rates because Mexican consumers refused to spend. 

Huh. Not a whole lot there until Trump’s election.

Finally, we’ll take a look at real wages, as a proxy for the labor market:

In short, 2015 was not a bad year at all, and 2016 was not much worse.

Some people have reported big price hikes in Mexico in 2016, but I am skeptical. Imaginary inflation has been seen all over the world — not least around the time of the adoption of the euro. And I have been travelling to Mexico pretty extensively over the period without noting big nominal hikes, save public transportation ... which is a pretty salient price and could easily cause people to feel that there is inflation going on when there is not. 

Either way, I submit to you: the depreciation of the Mexican peso has been a good thing for Mexico. It helped exports and cushioned the fall in oil prices to no discernible economic harm. 

Has Trumpian enforcement reduced illegal immigration in 2017? Depends on what you think happened in 2015

Border apprehensions in March 2017 fell a lot compared to March 2016. This is taken as evidence that Trumpian enforcement is scaring off potential migrants.

But as Kevin Drum pointed out, March apprehensions have been falling for a very long time. The red line shows March apprehensions for the corresponding year; the blue line shows February.

Then again, that long fall in March apprehensions stopped in 2011. They were stable until 2015, when they fell again, followed by stability in 2016, and then another big fall this year.

So is it Trump? Well, that depends on what you think happened in 2015. If something happened to start migration falling, then 2016 is the blip and this year is just the resumption of the trend. But February apprehensions (before Trump unleashed the CBP) fell in 2015 but did not rise year.

That makes 2015 seem like a one-off. If 2015 was a one-time drop, then 2017 might also be an independent one-time drop.

So what could have caused a sharp drop in apprehensions in 2015? Take a look at the purple and green lines. Those represent Mexican apprehensions of Central Americans headed north in March and February respectively. Those apprehensions jumped in 2015, when Mexico began to crack down on Central American migrants.

Now, the rise in Mexican apprehensions is less than the fall in American apprehensions, but the Mexican crackdown was brutal and likely dissuaded many people from heading north in the first place. 

Which means that we do have a one-off policy to explain 2015: a one-time upward shift in Mexican enforcement efforts. Mexican apprehensions have since fallen, but that appears to be more likely due to dissuation (as Central Americans realize how much more unfriendly Mexico has become) than a slackening of Mexican efforts. 

To sum: apprehensions fell steadily from 2007 to 2011 due to the Great Recession. They then stabilized until 2015. In 2015, apprehensions shifted downward when Mexico stepped up enforcement, stopping people before they reached the U.S. In 2017 apprehensions shifted downward for a second time, possibly when Trumpian enforcement started to scare people off. 

We won’t know for sure until the numbers for April (and May and June and July) are in. But it seems quite plausible that the current angry policies have reduced migration. Simply drawing a trend line from 2000 to date and declaring no effect can be quite misleading. 

The Trump wall keeps getting shorter, another guide for New Yorkers

Man, at first the Trump Wall was 60-feet high. That was a wall! A tenement-high wall. You could be impressed by that.

But now the specs are calling for prototypes between 18 and 30 feet tall. Dude. That is not tall. Thirty feet would make it shorter than the row of taxpayers behind me in this photo:

 

Sad! Come on. If we are going to go stupid, we should go stupid bigly. I say 80 feet!

By the way, for those of you who are curious, those taxpayers are where they are because the zoning will not allow anything higher on that corner. That is ironic, since they were tossed up in the first two decades of the 20th century as placeholders awaiting a skyscraper boom that took ten decades to arrive. Now that it is here, you cannot build there. Also sad!

NAFTA and agriculture

There was some interest in this topic but not a whole lot. So I will put up a short post, which will be pretty straightforward. I will discuss U.S. exports today.

The key thing to keep in mind is that agriculture is a global market. So if third-countries displace American agricultural products in the Mexican market, then American producers will be able to export their products to the places the third-country used to sell to. But if Mexico just reduces its consumption of all agricultural imports, then U.S. producers will lose unambiguously.

Mexico takes 8% ($1.8 bn) of American soybean exports. It takes 21% ($2.4 bn) of U.S. corn exports. It takes 11% ($0.9 bn) of our wheat exports.  It turns out that 36% of U.S. corn travelled to Mexico by sea in 2007-10. See page 10. The share is even higher for other products: 55% for wheat and 38% for soybeans; see pages 20 and 23. I used USDA data to calculate costs per ton-mile for soybeans: subtracting out $230,000 per ship Panama Canal tolls for cargoes from the U.S. costs gives you an average cost per ton-mile of 49₵ from the U.S., 43₵ from Argentina, and 39₵ from Brazil; fixed costs come to $5.70 from the U.S., $7.20 from Argentina, and, inexplicably, only $2.70 from Brazil.

With that data in hand, let’s look at those three markets and imagine that Mexico slaps MFN tariffs on the U.S. while opening to the South Americans. 

Soybeans: Mexico could easily supply itself from Argentina and Brazil. Landed costs to Mexico are already cheaper for the Latin American nations, even with higher transport costs; only South American export taxes keep us competitive.

Slap a 4% tariff on American exports (while removing them on the South Americans) and Argentina and Brazil would drive us out of the market. But remember that soybeans are sold in a world market. So more South American exports to Mexico means less South American exports to somewhere else. American production will fill up the gap. When it all shakes out, you would have a small loss for the Americans, a small gain for the Argentines and Brazilians, and about a wash for the Mexicans. 

Corn: The U.S. and Argentina have more or less the same corn production costs. But transport costs from Argentina are quite a bit higher. If Mexico imposed MFN tariffs on corn and suspended them on South American exports, the landed cost of U.S. maize would still be about 13% less than Argentine corn. Americans would lose sales via a general cutback in Mexican purchases, not a switch to South America. 

Lose-lose, although more for Mexican consumers, since U.S. producers would presumably find alternative markets. That said, Mexican corn farmers would gain, and rural voters are a major constituency, about 30% of the electorate. (See page 10. Rural voters punch slightly above their weight in turnout terms: see page 9.)

Wheat: See soybeans. 

Upshot: ending NAFTA would not be catastrophic for American soy and wheat exporters, but both would have to discover new markets right quick or suffer some unpleasant transition costs. It would be a clear boon for Argentine and Brazilian farmers, who could capture a whole new market. Mexican corn farmers will gain; Mexican consumers will lose.

Still, NAFTA probably won’t end.

Hechos alternativos in the Financial Times?

Today the Financial Times published a mysterious map purporting to show Mexican-American border disputes, according to Senator Patricio Martínez. I am more than slightly surprised that the newspaper is giving such credence to the Senator’s bizarro-world claims with no further fact-checking. 

It is a very nice map, except that we do not have the foggiest idea where it comes from, save perhaps the febrile imagination of Senator Martínez.

Article I of the Gadsden Purchase Treaty lays out the border pretty specifically. Martínez seems to be suggesting that nobody noticed that Nogales lies entirely inside Mexico. He also seems to be suggesting that the Gadsden Purchase Treaty said that the line west from El Paso should have run 115 miles instead of 100. Which would be fine, if the treaty didn’t, you know, say “one hundred miles.” 

The story in the FT piece about the Mexican government discovering that the stone cairns marking the border had been destroyed is true. The problem with the story is that the U.S. and Mexican governments spent six years resurveying and delineating the border in 1891-96. The Barlow-Blanco resurvey went over the whole damn thing and replaced all the missing border markers. You can read the whole report here. 

As for the 1897 note from the Mexican foreign minister, I have no idea. I do know that Matías Romero signed a border convention in October, so his objections, whatever they were, did not seem to be major. 

To conclude: the newspaper should report that a Mexican senator is claiming that the border is in the wrong place! But I would think that maybe, you know, they might also at least hint that there is pretty much nothing behind those claims. 

How high is the Trump Wall? A guide for New Yorkers

OK, the President claims that the wall will be 60 feet high. If I had to bet, I would bet that it is not going to be built, but who knows? If I were in Congress, I would vote for the boondoggle in exchange for leaving Obamacare alone. 

Still, how tall is sixty feet? It is not a height that you would normally reference. So I am here to tell you! A “standard” six-story NYC new law tenement rises about 66 feet. (I lived in one for a few years as a teenager, a building completed in 1910.) The below picture shows my boy (in the blue jacket) playing jump rope in front of a row of new law tenements in Manhattan. 

The proposed wall would be about six feet shorter than the row of buildings in the middle of the picture. Here is another view of similarly-sized structures:

Just imagine that this was South Texas and those buildings were the wall! Easy.

Of course, the President has also said that he wants the wall to be eighty feet tall. That would make it the height of the building with the watertower in the background (77 feet) of this picture from last May:

And now you know! I will admit that if we are going to blow billions on a pointless monument across the southern border, I kind of hope we go all out and build it eighty feet tall. Our grandchildren in the 2060s will zip over it in their flying cars and marvel at our folly. If you are gonna be stupid, be impressively stupid! 

ADDENDUM: And expensive! For those of you who want an idea of the cost, a good calculator is here. An 80-foot high 18-inch thick barrier built over a little more than half the border (1,000 miles) clocks in around $28 billion. Another calculator, which uses higher cost assumptions, gives $45 billion.

Hechos alternativos

A Mexican senator has called for Mexico to reopen all its border disputes with the United States in order to oppose the Trump administration. (Hat tip: Mexfiles.) According to Senator Patricio Martínez — who formerly served as governor of Chihuahua, a border state — the Mexican government should insist on the transfer of about 400 square kilometers that Porfirio Díaz asked for back in 1890. “Díaz was unsuccessful, but Mexico should send a new diplomatic note to Washington to rectify the border limits,as defined in the two 19th century treaties. If Trump refuses, Mexico should seek the intervention of the International Court of Justice, in the Hague.” 

Senator Martínez is planning to fund a commission “to document what has been known since the time of Díaz, that the treaties of Guadalupe Hidalgo and La Mesilla, which set the border between the two countries are not applied, and there are differences and errors in favor of the United States which must be corrected.”

That would be a great plan if there were any evidence that the United States and Mexico had any outstanding border disputes. Sadly, there does not seem to be evidence of any such conflicts, not since the Boundary Treaty of 1970. The idea that the U.S. is sitting on 400 square kilometers of Mexican territory would be an alternative fact.

No, actually Trump really did drive down the Mexican peso!

Kevin Drum announces with much certainty that Donald Trump has nothing to do with the peso’s fall! Is he right?

No. He’s not.

It is true that the Mexican peso weakened in November 2014, when oil prices dropped from around $80 to $66 and analysts concluded that prices were heading further down. It is also true that the peso kept weakening for much of the remainder of 2015, for reasons which are unlikely to have anything to do with Donald Trump, even if the above graph might have you think otherwise.

But after that? Let’s start by eyeballing it. First you have the weakness in January and February, when Trump polls well and then, on February 1st, starts winning. After that the markets breathe a sigh of relief as Trump starts to attract opposition … until the day Trump cleans up in Pennsylvania (and DE, RI, CT, and MD) and makes it clear that neither Cruz nor Kasich can stop him. After that, some stability … until Election Day, 2016.

Sure, you can draw a nice line through the trend and say, “Nope, Trump had nothing to do with it.” But that makes the ups-and-downs look might coincidental. 

But it gets worse for Drum’s argument. The peso moved on individual debate performances. It jumped right to the minute when the election result became clear. It tracked events with entirely understandable precision. And academics took notice: for fancy econometrics making the point, see here and here.

Finally, theory. For Drum to be right, he needs some sort of causal theory, something that would predict a long slow decline with the rest being random noise. (Highly coincidental noise, but still noise.) In other words, he needs a hypothesis. Oil prices won’t do it; they were priced in early. There might be another just-so story out there. I’d be curious to hear him propose it. But it seems silly to look for one, when there is an obvious explanation that fits the data staring us all in the face.

NAFTA, trade costs, and the difficulty of economic divorce

Market access can have a big impact even when the numbers are small. For example, extending normal trading relations to China did not dramatically increase access to the U.S. market. But it did increase certainty: previously, China’s access was subject to annual Congressional votes. Remove that uncertainty, and poof, suddenly firms were willing to make long-term investments aimed at the American market. (The paper on this topic is here.) 

The importance of permanence is why giant fluctuations in exchange rates should not be expected to have the same effects as small changes in tariffs. Exchange rates rise and fall subject to unpredictable economic forces. Tariff rates stay the same subject only to politics. Gabriel Mathy made that point here. 

So what about ending NAFTA? Would it undo economic integration between the U.S. and Mexico? 

Quite likely not.

Wait, what? Why would an end to NAFTA not benefit China and other low-wage manufacturers in Asia?

Well, simply put, they are too far away from America. In 2014, according to the Bureau of Transportation Statistics (see Table 10 and Table 11), the average container imported from Mexico carried goods worth $52,921. According to J.P. Morgan, the average cost of moving that container from Mexico was about $1,800. (The World Bank says $1,499, but that includes maritime and non-NAFTA shipments.)

So that is a cost  of roughly 3.4% ... compared to 15% from China. (The number of containers from China comes from the Maritime Administration; the value is from the Bureau of Transportation Statistics.)

This is back-of-the-envelope, but very little of the stuff that Mexico exports to the U.S. is set to move to Asia, Europe, or South America if NAFTA falls apart. Mexico is too close and its labor costs are too low.

Still, a NAFTA collapse could move production back to the USA, no? After all, it is not like U.S.-made goods don’t need to be transported to market. They do. Domestic truck and rail shipments came to $8.8 trillion in 2007. (To prevent link rot: BTS, National Transportation Statistics, Table 1-58: Freight Activity in the United States: 1993, 1997, 2002, 2007 and 2012. I used 2007 because 2012 trucking freight rate data is not available.) The total cost of moving all that stuff came to $262 billion. That comes to an average freight cost of about 3.0%. In other words, and not unexpectedly, the cost of moving goods around inside the United States is roughly equal to the cost of moving them from northern Mexico to the United States. So one would expect a rise in tariffs from zero back up to 3% to have an impact. 

But tariffs and transport costs are not the only trade costs! Total trade costs come to the difference in the markup between the factory gate and the final consumer between domestic goods and imported goods. When Dennis Novy (Warwick) used that approach (albeit indirectly) he found that NAFTA reduced the total trade cost from about 70% to around 33%. (Page 10)

How did that happen? The intuition is as follows. Signing NAFTA cut U.S. tariff rates from 3% to zero. Companies then established cross-border value chains. Those chains were not easy to set up, but once they were in place, it became far cheaper to export. Think of it this way: the first autoparts factory to open up in San Luis Potosí needed to figure out everything about selling to the United States. The fifteenth had a much easier time. Thus, NAFTA caused trade costs to drop by way more than the fall in tariffs.

The flip side of that, however, is that much of the fall in trade costs is not reversible. Abolishing NAFTA, therefore, might have remarkably little effect, taking trade costs up from 33% to 36%. Abolishing NAFTA plus a 20% border adjustment tax will reduce Mexican incomes (via depreciation) but would also have remarkably little effect on trade volumes. 

It would take a straight-up tariff in the 20%+ range to seriously undo integration in any reasonable time period, at least for manufacturing. And even then the impact would be slower and smaller than we think. In fact, most of the impact would come in the form of European cars becoming suddenly competitive in the American market. 

Agriculture would be a different story. I will discuss more if there is interest. 

The border tax and climate change

I am opposed to a 20% tariff aimed at Mexican manufactured goods. I am more ambivalent about the 20% border adjustment tax (BAT) floated by the Trump administration. In fact, I may support it. 

The basic idea is simple. U.S. companies would pay a 20% income tax. Companies could deduct domestically-sourced inputs from their income but not imported goods or services. Those imports would then, in effect, have to be at least 20% cheaper than equivalent domestic competitors. 

The idea is to make the U.S. corporate income tax look like a value-added tax (VAT). If a corporation in Mexico buys an input from overseas, it pays the Mexican VAT on the purchase. If it sells something domestically, then it also pays the Mexican VAT. But if it sells something to a company in France, then it does not pay the Mexican VAT ... but the French buyer will pay the French VAT. Under the BAT, if a corporation in the U.S. buys an input from overseas, then it pays the U.S. BAT on the purchase. Ditto it pays the BAT on all earnings from domestic sales. But if it sells something to a French buyer, then it will not pay the U.S. BAT ... but the French buyer will pay the French VAT.  

As it stands, the BAT will apply to oil imports. So an American refinery that buys foreign crude would have to pay the 20% BAT on those purchases! Right now, WTI trades around $54 per barrel. A refinery using American crude would pay $54; a refinery using imported crudes at a similar grade would pay $65. American oil prices would rise, world oil prices would fall.

And that is good for the world! Higher U.S. oil prices mean less U.S. oil consumption. Moreover, the burden will fall particularly on dirtier forms of oil. Most American crude oil imports (60% in 2015) are heavy grades. Almost half of that consists of heavy crudes from Canada, which  release about significantly more carbon during the production process than other grades of oil. (So much so that the Section 526 of the Energy Security Act of 2007 theoretically banned U.S. federal agencies from purchasing it, although that has had no concrete effect.)

Lower world prices will, of course, stimulate more world consumption. But there are two wrinkles that will reduce that impact. First, transport links lock in most Canadian output to American markets. Last year, the Trudeau government approved an expansion of the Trans-Mountain pipeline from 300,000 to 890,000 bpd, which would allow Canada to divert as much as 28% of its current U.S.-bound exports to other markets. But that will take a long time and is a drop in the bucket of world production. Second, American refineries are optimized to use heavy crudes. They will switch away, but slowly, even if the cost of their primary input rises. Finally, inasmuch as world prices do fall, that will discourage foreign exploration and development, keeping some carbon in the ground for the long-run. 

Winners:

1. Our children;
2. American heavy oil plays, of which there are not many outside Alaska;
3. American light tight oil plays (hooray!);
4. Andrés Manuel López Obrador (see below).

Losers:

1. American oil consumers, who will likely see prices rise about 30₵ per gallon;
2. Refiners who use imported crude oil—although not by very much;
3. Canadian, Ecuadorean, Venezuelan and Mexican heavy oil producers, who are locked into the U.S. market;
4. Mexican consumers, who rely on gasoline produced in the United States from Mexican heavy crude.

As far as energy is concerned, this does not strike me as a bad policy, even if it will impact Mexico quite seriously. At 2015 prices, a 20% BAT would have raised about $2.0 billion from Mexican crude oil imports, although most of that burden would fall on American consumers. The real hit for Mexico is that the increase in U.S. gasoline prices will be passed on to either Mexican consumers or the Mexican government. Inasmuch as the Peña administration hands Mexican voters a second gasolinazo, then it will be another fillip for AMLO’s chances in the 2018 elections.

So a good policy for energy: a net plus for domestic production and a net plus for decarbonization. Probably pretty good Harley-Davidson, as well.

But as far as other products markets are concerned, I’m still undecided as to whether the BAT is a good idea. Thoughts?

Maybe NAFTA is safer than we think

It is true that President Trump has made threatening noises. And it is true that he has appointed people with protectionist bents to run Commerce and the USTR office. 

But said Secretary of Commerce, Wilbur Ross, founded a major Mexican autoparts manufacturer. International Automotive Components makes parts for doors, side panels, consoles, exteriors, floors and sound systems for BMW, Fiat Chrysler Automobiles, Ford, General Motors, Honda, Mercedes, Nissan, Toyota, Volkswagen, Hyundai/Kia, Volvo Truck and Freightliner. He also owns the International Textile Group, with plants in Morelos.  

Talk about irony!

Now, Secretary Ross will presumably liquidate those holdings. So he should not have a financial stake in preserving NAFTA. But he will viscerally understand the many companies that do. 

I now suspect that the much-ballyhooed NAFTA renegotiation is going to end with stricter rules of origin, say 80% instead of 62½. There may also be some weakening of the dispute-resolution mechanism and higher labor standards. But I increasingly think that the Trump administration will refrain from blowing it all up. 

Although I still have no idea what fakery they will use to pretend that Mexico is paying for the wall.

Border adjustment, the Mexican peso, and Hoyt Bleakley

It now looks like the administration won’t (for now) be using its Economic Powers Act authority to impose a tariff on Mexican goods. Instead, he is going to ask Congress to create a border adjustment tax, which would stop American companies from deducting imports from their taxable income. At the same time, it would leave their exports untaxed. (Full explainer from Forbes, here.) 

So what would that do to the auto industry? Well, let us proceed in two stages.

1. What is the immediate impact on an American auto producer? Well, the border adjustment tax is less harmful than a tariff. The auto companies would not be able to deduct the cost of goods coming in from Mexican affiliates and suppliers, but they also would not be taxed on the cost of goods going out to Mexican affiliates and suppliers. But they would still have an incentive to source more inputs at home.
2. Except that the peso would fall (or the dollar would rise, same thing) in response to the tax. A 20% depreciation in the peso puts the auto company back where it started. 

But would Mexico allow the peso to fall that much? In the past, the Mexican central bank has acted terrified that depreciation would hammer Mexican corporations that had borrowed in dollars. That is a reasonable fear ... in theory.

In practice, it might not be. My friend and colleague, Hoyt Bleakley (with Kevin Cowan) looked at 450 non-financial firms in five Latin American countries. 

We consistently find that, contrary to the predicted sign of the net-worth effect, firms holding more dollar debt invest more than their counterparts in the aftermath of a depreciation. We show that this result is due to firms “matching” the currency denomination of their liabilities with the exchange rate sensitivity of their profits. Because of this matching, in equilibrium, the negative balance sheet effects of a depreciation on firms holding additional dollar debt were more than offset by the larger competitiveness gains of these firms.

In other words, depreciation might not be that harmful.

The other reason the central bank might not want to let the peso fall is fear of inflation. But that depends a lot on the reputation of the central bank and the structure of wage bargaining. Banxico is very credible and Mexico has weak unions, so while a fall in the peso will lead to a one-time increase in costs, it seems likely that Mexican workers would just have to suck up the fall in their living standards. It is hard to see how they could demand the kind of compensating wage increases that you would need to set off an inflationary spiral.

So yes, the peso could fall.

In short, the Trump tax would likely make Mexicans poorer with no compensating gains for American workers. But also no losses for fans of big American cars like the amazing Dodge Challenger, like my son. In other words, a gratuitous punch in the nose of those perfidious foreigners! And you can pretend that the tax flows are coming from those foreigners, when they are really coming from American consumers. 

Yeah, I can see why this might actually happen. 

The Trump tariff and the American auto industry

I have no idea if the Trump administration plans to impose a 20% tariff on Mexico. It seems now that the surprise announcement had to do with plans for a border adjustment tax, which is a different thing. But who knows? This is the Trump administration! Why are you asking me to make predictions?

Anyway, yesterday I mentioned that any tariff would impose serious problems on the American auto industry. The big problem is twofold: first, Mexican suppliers provide a large chunk of the inputs into American auto production. Second, many of those parts in fact cross the border multiple times. The end result is that a tariff would raise the cost of producing domestic cars by quite a lot. 

Some figures: in 2012, the state of Michigan published a report on the Ford Motor Company’s Michigan Assembly Plant. The report found that the plant purchased about $3.1 billion from Tier I suppliers; of that, $831 million (or 27%) came from Mexico. (Roughly $3,800 per vehicle.) In 2014, the U.S. auto industry consumed $42 billion in intermediate inputs from Mexico (pages 10-11), approximately $3,500 per vehicle. So at first glance, the tariff would raise the cost of American-built small cars by $760 and all vehicles by about $710.

That is a lower bound, however. In the course of making a motor vehicle, auto parts cross the border as much as eight times. If the tax was applied at each crossing, the cost would go up commensurately. If you assume that components cross from Mexico four times, doubling in value-add each time, then the total cost per vehicle would rise from $710 to $1,300. 

The cost increase would hit small cars more than large ones ... but consumers would see the price hike in the large car sector. The reason is that auto companies in the United States need to sell a certain percentage of small cars in order to hit their fleet fuel economy standards. They will therefore hold down the price on small cars, transferring the impact to their larger vehicles.

The problem is that the U.S. producers more larger vehicles than small cars. 

How much would production shift back to the United States? I suspect less than the current studies predict. Why? Well, the Mexican peso would likely fall. The real exchange rate would go down and stay down. Mexican workers, who consume a lot of imports, would become poorer but they would also become cheaper. A 20% nominal decline in the peso means a drop from 22 per dollar to 26 per dollar. That is certainly conceivable. Hell, 30 per dollar is conceivable. A messy drop would potentially hurt the Mexican economy, if it turns out that Mexican companies have a lot of hidden dollar liabilities, but would nonetheless increase competitiveness.

And so, you would have a temporary increase in the price of large automobiles and light trucks, followed by a drop in Mexican costs. Some production would return to the U.S. (in order to avoid uncertainty) but in unsteady and unpredictable way. It would not be an economic apocalypse for the United States, but it would cause havoc across the Mexican economy with little (if any) employment gain in the USA and big losses for shareholders. And unless the tariff is extended to all countries, the big winners would be car manufacturers from Europe and Asia. 

That said, this is all theory. Who knows what the Administration will do?

A 20% tax on Mexican imports? Why, yes.

Now the Trump administration is floating the imposition of a 20% tax on Mexican imports to pay for the wall. Obviously, this would violate NAFTA, but the U.S. President likely has the unilateral authority to impose such a tax.

The International Emergency Economic Powers Act of 1977 gives the President the power to “investigate, block during the pendency of an investigation, regulate, direct and compel, nullify, void, prevent or prohibit, any acquisition, holding, withholding, use, transfer, withdrawal, transportation, importation or exportation of, or dealing in, or exercising any right, power, or privilege with respect to, or transactions involving, any property in which any foreign country or a national thereof has any interest by any person, or with respect to any property, subject to the jurisdiction of the United States.”

Presumably, the tax would be considered regulation of importation under the above paragraph.

Now, the Economic Powers Act also compels the President to “consult with the Congress,” but it is not clear that imposes a significant limit on his tax powers.

Alternatively, President Trump could use the Trading with the Enemy Act of 1917, which Richard Nixon invoked to impose a 10% import surcharge in 1971. But the Trump tax could be appealed to the  U.S. Court of International Trade USCIT). It turns out that there is some case law on this.  In United States v Yoshida International, USCIT’s predecessor ruled that the courts would not challenge the validity of a Presidential declaration of a state of war or national emergency but would check to see that the remedy imposed was appropriate to the nature of the emergency. I suppose that the emergency would be the existence of illegal immigration from Mexico and the remedy would be the wall. (Judging from U.S. v Yoshida, I do not think that the tax would stand under Trading with the Enemy but it might under the Economic Powers Act. Either way, it could be in place for years.)

The courts would have to review if this violates the NAFTA Implementation Act of 1994, but my guess is that the Implementation Act would not be held to supercede Presidential powers under the Trading with the Enemy Act or the Economic Powers Act. 

And even if it did, the Trump administration could announce that it was withdrawing from NAFTA. The logic would be that the NAFTA agreement allows withdrawal and the NAFTA Implementation Act gives the president the power to proclaim tariffs as laid out in the NAFTA. Poof, a rationale for unilateral Presidential powers. 

Of course, if you abrogated NAFTA, you could easily enough pay for the wall with the imposition of pre-NAFTA levies. It looks like the Trump administration is looking for more than that, however: they seem to want to damage U.S.-Mexico commerce as much as possible. 

Or at least drive down the value of the peso even further. 

The tax would also have to exempt autoparts. More on this later ... 

Peña-Trump meeting cancelled!

Over the past week, it seems to have sunk in to the Mexican commentariat that Trump is serious about building a wall and renegotiating NAFTA. Newspaper articles have gone from “Hey, maybe there is an upside here, we can modernize NAFTA!” to “We are at war.” 

Then President Peña cancels his meeting with the American president.

As I have pointed out, many Mexicans blame Peña for Trump’s election. They’re probably not wrong. The logic is that Peña could have humiliated Trump. Then he recently brings back Videgaray, who had to resign over the Trump summit, on the theory that Videgaray is down with the Trump administration. But then the American president orders construction on the wall and continues to insist that Mexico pay for it. It leaves Peña with the worst of all worlds. 

The signs are that the Mexican president is under great strain. His approval rating is down to 7%. “He stays late in his office and sleeps little,” says an administration insider. 

Department of corrections: the Gasolinazo did not raise tax revenue

Here I am, in Mexico, as the 45th presidency of the United States begins. The talk among the intelligentsia is about NAFTA and, weirdly, the prospects of a war between the United States and China. So I am going to talk about the gasolinazo. Which still has a lot of people upset, as you can see from the image of this angry soldadera pointing a carbine at President Peña.

Now the notionally right-wing National Action Party is trying to escape blame for the price hike. It isn’t the reform what caused the spike, it’s “the tax structure what obscenely raises the price!” says Ricardo Anaya, the head of the party.

Well, not quite. On January 6th, I reported that the price rise was essentially a tax hike. Turns out that is wrong. All of the rise is going to Pemex, to cover the costs of refining or importing the stuff. 

El @gobmx no recibirá ni un centavo por el incremento en el precio de la gasolina⛽️ pic.twitter.com/JaQZ2f9Mpv

— SE México (@SE_mx) January 7, 2017

Moreover, decontrol of gasoline prices was part of the reform that the PAN did indeed vote for. Now, it is true that the price of gasoline in Mexico is now higher than in the United States. But that is because Pemex is a mess, both in terms of refining the stuff at home and transporting it from the United States at a reasonable price. 

Right now, taxes on internal fuel sales serve as a buffer against price swings. When the international price rises, government revenues fall and vice versa. The plan, logically, is for those swings to now be passed on to the consumer. Government revenue will stabilize. 

The problem is that Mexican voters, well, they liked price stability. Still, even with the recent riots, I suspect that the administration will follow through on its plans for full liberalization. 

⛽️ What have they go to lose? Besides the next election, but that ship, I suspect, has already sailed.

 

Riots in Mexico

Mexico has been beset by widespread looting and demonstrations over the past five days. So far, there have been 987 arrests and two deaths. The incidents devolved into outright riots in Ixmiquilpan, Hidalgo, and Monterrey, N.L.; but stores have been sacked and roads blockaded all over the country. I heard about the looting from friends living in Ecatepec, Mexico State, a poor suburb of Mexico City, but it has now made its way into American news.

The trigger? The Peña administration decided to raise gasoline prices on January 1st. Prices rose to heretofore unseen levels in real terms. The below figure show the retail price per gallon for Magna gasoline in Mexico and the equivalent blend in the United States, measured in real pesos of November 2016. 

Why the rise? It was essentially a tax hike. As you can see, Mexican retail gasoline prices have been above American ones since the middle of 2014. U.S. prices have risen recently, but Mexican prices remained lower.

The hike was not about raising Mexican prices to equal American prices. Rather, the Mexican government wanted to raise prices now in advance of its plan to fully liberalize prices later this year. The higher prices are in advance of liberalization, the more room the government has to impose a high excise tax. Since the current administration has promised to freeze said tax once it is introduced, it wants to be able to set the initial tax as high as possible.

As policy, none of this is a bad idea. Price liberalization will save the government money and excise taxes should be high. But as politics, it is terrible. A more popular government facing a better economic situation could have gotten away with raising gasoline prices. But the Peña administration is not popular and the economic situation is not good. 

The latest blow to the Peña administration has been the recent re-emergence of the man behind Trump’s visit to Mexico during the presidential campaign. Trump, as you can imagine, is not popular south of the border. Public opinion was horrified that the Peña administration invited him to meet the Mexican president. There was some hope that President Peña intended to lure candidate Trump into a trap, but instead Peña allowed Trump to roll all over him in a joint press conference. (The rules were no questions, but Trump invited them anyway; then Peña neither shut Trump down nor called out the GOP candidate when he contradicted what Peña said had gone down in the meeting.) 

The meeting was such a disaster that President Peña felt he had no choice but to fire Luis Videgaray, the man who had planned it.

But now Videgaray is back as the new foreign minister ... precisely because Peña believes that Videgaray can curry favor with President Trump. So this government brings the man who sold them out to Donald Trump and then turns around to raise gasoline prices? 

The economic situation makes the situation worse for the Administration. The “gasolinazo” happened while Mexican inflation is rising. (See the below chart.) Why is inflation rising? Well, the value of the peso is collapsing, driving up the price of imports. Why is the peso collapsing? Well, in one word, Donald Trump. The fear that he will impose trade barriers on Mexico causes investors to sell pesos, and so drives down the exchange rate.

But Donald Trump is a man whom many Mexicans believe would never have won election had President Peña had the cojones to humiliate him in that press conference way back when.  

But Donald Trump is a man whom many Mexicans believe would never have won election had President Peña had the cojones to humiliate him in that press conference way back when. 

Let us add it up. The Peña administration curries favor with candidate Trump. Trump wins the election, indirectly causing inflation to accelerate in Mexico. The Peña administration then pours fuel on the inflationary fire by hiking gasoline prices while bringing back the sellout who arranged the Trump meeting in order to curry more favor with the U.S. president-elect. 

Result: rioting.

Andrés Manuel López Obrador cannot believe his luck. 

(Filed under “2016 American election” because Donald Trump is not yet president and the roots of the crisis date back to his campaign.)

Things can get worse

One thing to consider is that the U.S. election results just make it a lot more likely that Andrés Manuel López Obrador (AMLO) will be elected the 21st president of the United Mexican States under its current constitution. AMLO is a nationalist and an opponent of the current president, the one who prostrated himself before an American candidate.

AMLO was a fine mayor of Mexico City, but has shown himself to have doubtful conviction about democracy. His election would be a major risk, particularly in the context of a hostile relationship with the USA.

Read the link, please. It is a sobering thought. 

Establishing the rule of law is not easy

The rule of law is something that is very hard to define. We know when it is not there at all: chaos and warlords. We also know when it barely exists: bribery and impunity. But it is very possible for a state to operate without outright bribery and impunity and yet still lack the benefits of the rule of law. 

Consider the recent Casas Geo bankruptcy in Mexico. (Hat tip: Xóchitl Herrera.) Casas Geo started by building small modern houses for poor Mexican families. The houses would be built in long rows, sharing foundations, with modern utilities already hooked up. A small four-bedroom house went for around $46,000. Reforms that made mortgage credit widely available fueled the company’s growth.

 

In February 2013, however, the new administration announced a change in housing policy: federal subsidies would be redirected from suburban expansion to more center-city housing. The reason was that Mexican suburbs had outrun their cities’ transportation nets, leaving increasing amounts of vacant housing around the outskirts and burdening poor households with crushing commutes. The problem was that the housing companies had accumulated a lot of land outside existing built-up areas. Geo was particularly hard hit. In April, the company announced a 38% decline in quarterly sales. (Page 2.) By May, the company had begun to default on some of its bonds. To be fair, housing construction had been in trouble before then and the reform was sorely needed. But it sent Geo over the edge into bankruptcy. Geo sought protection under the 2013 amendments to Mexico’s 2000 Bankruptcy Protection Act. 

The Mexican congress modeled the Bankruptcy Protection Act on American bankruptcy law. (A comparison can be found here.) As in the United States, federal courts run bankruptcy proceedings. Creditors cannot make an end-run around the bankruptcy proceeding by appealing to state or lower federal courts. 

Only they did! A bunch of would-be homeowners whose made down payments but never received a house wanted cash instead of the promise of a future house. So they sued in a different federal court. And now the whole reorganization is at risk; Geo might be forced to liquidate, which would leave everyone worse off.

That is not supposed to happen; you are not supposed to be able to get around a bankruptcy court decision by filing in a non-bankruptcy court. But since it did happen, well, now it is not clear if Mexico even has a functioning bankruptcy law. 

This sort of thing is not as obvious as corruption or an executive branch that flaunts judicial decisions. But it is nonetheless pernicious. The rule of law looks similar everywhere, but every place that lacks it lacks it in its own way.