Engagement’s Second-Order Catastrophes

Scott Rozelle and Natalie Hell’s book Invisible China is an interesting if somewhat dry look at the development challenges China faces over the next 20 years. Rozelle and Hell are proudly pro-China; they devote an entire chapter of Invisible China to convincing readers that China’s rise to primacy is a boon for all the world and that all the world should all work to speed its rise. Chinese officials are perhaps the book’s main target audience. Rozelle and Hell worry that unless Communist officialdom takes drastic action soon, China will be stuck in what has been called the “middle income trap.” China was able to climb into middle income status with an economy built around infrastructure construction and manufacturing. There is very little unique about this in the Chinese case except scale: dozens of countries have used variations on this exact same model to pull themselves out of poverty and into the middle-income zone.

Moving from middle income to high income is a more difficult transition.

A country like China will eventually reach a point where wages have climbed high enough that factories begin decamping to cheaper shores and infrastructure construction run up negative returns. At this point the country’s economy must transition to higher skilled work. Rozelle and Hell believe that the key to this transition is education. The countries that made the transition successfully, like South Korea, Taiwan, and Ireland, were those that spent large sums and concerted political effort to raise education levels and otherwise improve human capital during the transition from poor to middle income status. Middle school and high school graduation rates are a good way to track this. China fairs poorly on this metric. 30% of China’s labor force has a high school education or higher. That looks less like Taiwan or South Korea in 1985 than the Mexico or Thailand today.1


Scott Rozelle and Natalie Hell, Invisible China: How the Urban-Rural Divide Threatens China’s Rise (Chicago: University of Chicago Press, 2020), 5. In fact, Thailand, Mexico, South Africa, and Turkey all have higher high school graduation rates than China does.

But this post is not about China. Rather, it is about what China did to Mexico. Mexico is the cautionary tale Rozelle and Hell want to scare Chinese officials with. “If you don’t reform now,” they seem to argue, “what you did to Mexico will be done to you!” That is a boogeyman worth fearing.

 Mexico was not always the economically stagnant, narcotics ridden country that it is today. As Rozelle and Hell tell it, in the late ‘80s and early ‘90s it was a busy manufacturing hub with a bright future. In 1994 Mexico was invited into the OECD precisely because it seemed on the verge of jumping into the high income bracket. This upward climb was derailed when the United States granted “permanent normal trade relations” to the People’s Republic of China. Americans often debate the impact this had on the U.S. industries, but its effect on the Mexican job market is much clearer—and more catastrophic:

From 2001 to 2004, Mexico lost an estimated 400,000 jobs to China. Over the course of just three years, the once-dominant Mexican textiles industry was supplanted by China as the number one exporter to the US market, and more than a third of Mexican factories assembling clothing were shut down. Many low-wage industries also contracted sharply. By 2000 “Made in Mexico” was fast disappearing from Walmart’s shelves….

As the factories began to leave, Mexico’s economy faltered, then stagnated, and today has still never recovered. For the past two decades, per capita economic growth has hovered just above 1 percent (in the most optimistic accounting rules), much slower than would be expected for a country with Mexico’s level of development. In the metrics that matter most for a growing economy, Mexico has also fallen behind. Rising productivity—getting more economic output from fewer inputs through new technologies and better management approaches—is a sign of a healthy economy. But productivity in Mexico has been consistently stagnant or declining across this period. 2


ibid, 53.

So what happened when Mexico hemorrhaged all of these jobs? In his interview with Jordan Schneider on the ChinaTalk podcast, Rozelle describes one part of the story:

Within a couple years, 10 million people lost their jobs and that was 20% of the Mexican labor force. What did those 10 million people do? About half of them or a third of them went across the border to the U.S. China’s not gonna be able to do that. The second thing is they went into the informal economy; they started hawking fruit on the street, washing car windows, doing landscaping.3


Jordan Schneider interview with Scott Rozelle, “Invisible China: How the Urban-Rural Divide Threatens China’s Rise,” ChinaTalk, podcast, 15 April 2021.

In the book, Rozelle and Hell have more to say about this “informal sector” of the economy. As they define it:

The informal sector describes the part of the economy where workers do not have regular salaried jobs with benefits and set contracts. These are generally small businesses (or microfirms), often one person working alone. In some cases that person may employ a family member, a friend, or a neighbor or two. Think of a small food stand or tiny moving company or even the individuals in so many countries who stand at intersections and wash windshields for a few coins. These jobs are open to just about anyone, because they don’t rely on credentials or job interviews…. It’s likely that many people [in China] will choose the informal economy [when their factories shut down]—selling fried noodles in alleys or working as trash collectors, recyclers, window washers, or gardeners.

Again, these workers will be stuck with stagnant wage rates and falling earnings. And when you work for yourself on the sidewalk, it will be a job without stability or benefits. For ninety-nine out of a hundred, there will likely be no prospect of upward mobility. 4


Rozelle and Hell, Invisible China, 53.

Which brings us back to Mexico c. 2000:

Indeed, for more than two decades Mexico’s informal sector has been steadily growing. Right as the factories began to leave Mexico in droves, employment in the informal sector rose, then skyrocketed. In fact, from 1998 to 2013, employment in that sector grew by an astonishing 115 percent. Today, fully half of Mexico’s population is employed in the informal sector. In poor areas, things are even worse. In Mexico’s poorest state (Chiapas), fully 80 percent of the population works in the informal sector

Two influential economists—Santiago Levy at the Inter-American Development Bank and Daniel Rodrik, a Harvard professor—have argued that the growth of the informal sector has been a big factor in Mexico’s lackluster economic performance. The informal sector is much less efficient than the formal sector, it actively undermines the formal sector, and it weakens the social safety net. The cumulative employment growth in Mexico’s informal sector from 1998 to 2013, 115 percent, is much higher than the growth in the formal economy in the same period, 6 percent. Because it is much harder to tax, Mexico’s informal sector has dramatically reduced the revenue the government can draw on to deal with social problems or to make long-run investments. In the state of Chiapas, for example, the fact that 80 percent of the local population is in the informal sector means that the state government can recoup only a paltry 1.5 percent of its budget through taxes. The result is major downward pressure on the economy by the forces operating among the unskilled masses.5


ibid, 56.

The worst industry in the informal sector is the drug trade and the extortion rackets:

 As predicted, many other people who were excluded from Mexico’s economy turned to crime. From 2000 to 2010, business theft and extortion increased dramatically. From 2007 to 2014, more than 150,000 people were killed in Mexico as part of the violence that now permeates large segments of the country….

In Michoacán, a Mexican state where organized crime has a large and violent presence, an unskilled worker might earn about US$175 a month in difficult, unsteady work in the informal sector. But that same worker might earn three times as much by working for a local gang.

 The effects on the economy—and on human well-being—have been devastating. In Mexico, high levels of violence have pushed many to emigrate for safety. Rising crime rates have also suppressed investment, reduced consumption in the service sector, and in many cases forced people out of the labor market altogether, driving up unemployment rates. The government also has had to bear immeasurable costs of policing and damage repair, taking money away from more growth-oriented investments. With the rise of crime and the drug trade comes an associated rise in corruption of the police force and other government entities, making it that much more difficult to maintain the rule of law that a healthy economy depends on. 6


ibid, 59-60.

Rozelle and Hell understate the problem. A recent Congressional Research Service report describes the situation in Mexico like this:

According to one Mexican think tank that publishes an annual assessment, the top five cities in the world for violence in 2019 were in Mexico. Increasing violence, intimidation of Mexican politicians in advance of elections, and assassinations of journalists and media personnel have continued to raise alarm. From 2017 through 2019, a journalist was murdered nearly once a month on average, leading to Mexico’s status as one of the world’s most dangerous countries to practice journalism.

In the run-up to the 2018 local and national elections, some 37 mayors, former mayors, or mayoral candidates were killed, and murders of nonelected public officials rose above 500. Over many years, Mexico’s brutal drug-trafficking-related violence has been dramatically punctuated by beheadings, public hangings of corpses, and murders of dozens of journalists and officials. Violence has spread from the border with the United States into Mexico’s interior. Organized crime groups have splintered and diversified their crime activities, turning to extortion, kidnapping, oil theft, human smuggling, sex trafficking, retail drug sales, and other illicit enterprises. These crimes often are described as more “parasitic” for local communities and populations inside Mexico, degrading a sense of citizen security.

The violence has flared in the Pacific states of Michoacán and Guerrero, in the central states of Guanajuato and Colima, and in the border states of Tamaulipas, Chihuahua, and Baja California, where Mexico’s largest border cities are located.Drug traffickers exercised significant territorial influence in parts of the country near drug production hubs and along drug trafficking routes during the six-year administration of President Enrique Peña Nieto (2012-2018), much as they had under the previous president. Although homicide rates declined early in Peña Nieto’s term, total homicides rose by 22% in 2016 and 23% in 2017, reaching a record level. In 2018, homicides in Mexico rose above 33,000 for a national rate of 27 per 100,000 people. According to the U.S. Department of State, Mexico exceeded 34,500 intentional homicides in 2019, for a national rate of 29 per 100,000. Thus, for each of the most recent three years, records were set and then eclipsed.7


June Beitell, “Mexico: Organized Crime and Drug Trafficking Organizations,” Congressional Research Service report no. R41576 (28 June 2020).

Where lies the source of all these ills? Why a two decade floodtide of illegal immigration to the United States, hundreds of thousands slaughtered in drug wars at home, the degeneration of the Mexican state, and the stagnation of the Mexican economy? Rozelle and Hell blame this all on low levels of basic education in Mexico. Perhaps that is the ultimate root of Mexico’s ills. But by their own account there is another source of these problems, the proximate cause of Mexican decline.

Search for the origins of catastrophe and what do we find? Well-dressed suits shaking hands in Washington’s fluorescent conference rooms. American officials pushed to bring China into the WTO; American legislators voted China favored trading status. In doing so, they signed the death warrant of a thousand factories across Mexico. Joblessness, stagnation, violence, and flight soon followed. These sort of second order consequence were not considered when Americans sat down to determine the shape of China’s relationship with the United States. The minds behind engagement did not ponder what horrors their tradecraft might work.

Perhaps they should have.


Readers interested in learning more about the failures of engagement might find the posts “Yes, We are in an Ideological Competition With China” “When Engagement Backfired,” “Two Case Studies in Communist Insecurity,” Mr. Science, Meet Mr. Stability,” “Do Mil-Mil Exchanges With the Chinese Do More Harm Than Good?” and “Give No Heed to the Walking Dead” of interest. To get updates on new posts published at the Scholar’s Stage, you can join the Scholar’s Stage mailing list, follow my twitter feed, or support my writing through Patreon. Your support makes this blog possible.


Leave a Comment


This is silly, Tanner, the Reagan amnesty was in 1986, Mexico was the first victim of the Latin American debt crisis of 1982, and the stagnation of the China trade after 2011 did not lead to an improvement in its economic performance. China probably did take manufacturing jobs from Mexico in the first decade of the twenty-first century, but NAFTA really did result in an improvement in Mexico’s economic complexity of exports during the 1990s and even later, and a one decade China shock surely can’t be responsible for four lost decades. Also, the gains in China from trade with the rest of the world were surely much larger than Mexico’s losses. Surely you don’t want the Chinese to starve, Tanner (or, for that matter, become a Soviet-style autarkic great power)?

” the stagnation of the China trade after 2011 did not lead to an improvement in its economic performance.”

Yes, after the drug war had started, security conditions had deteriorated, half of the economy had gone informal, state budgets had dropped, and so forth. That’s Rozelle’s entire point: middle income is a “trap” once the economy goes informal.

The counterfactual of course is what would have happened had China not been given preferential treatment in the USA. My wager is that eventually it would have fallen off anyway–but it makes a difference if that happens because of the China shock in 2000 vs. the Vietnam/Bangaldesh shock in 2015.

The other reason this post was silly was because China’s rise -through creating more available investment funds, lowering input costs, and creating new markets for exports- probably benefited the global economy overall. India, Bangladesh, Pakistan, the Philippines, Cambodia, Eastern Europe, Sub-Saharan Africa, and Turkey, for instance, had substantial growth accelerations in the first decade of the 2000s, and Indonesia crawled out of the hole left by the 1997 crisis. Most oil exporters benefited, surely. Even in Latin America, the growth spurts in Brazil, Peru, Colombia, Honduras, and, of course, Ecuador and Venezuela likely outweighed any damage done to Mexico. Some countries, like Portugal, the U.S., Mexico, and various commodities importers probably were hurt by the China shock, but the number of winners seems to have outweighed the number of losers.

Similarly, it’s important to note that while the Mexican stagnation has been historic and embarrassing, the country does not severely underperform economically relative to its national IQ. Chile (average IQ 91) is currently at $24,928 GDP per capita by PPP, Argentina and Uruguay (which likely benefited from the China shock; average IQ 89) are currently at $23,474 and $22,141, respectively, while Mexico (average IQ 87) is at $20,266. None of these are disastrous numbers.

“The counterfactual of course is what would have happened had China not been given preferential treatment in the USA.”

You get a more bifucerated world, with China exporting more to Russia, Asia, and Africa. If the U.S. institutes secondary sanctions, Europe, Japan, and Korea wouldn’t import much from China, either. Think the distribution of Chinese vaccines. China would also be more autarkic, including in matters of fossil fuels. China would produce less pollution and probably rely more on services and mining and less on manufacturing. The Chinese economic boom would likely still take place due to reform and opening up, as well as the transfer of Hong Kong to China. “Wolf warrior” diplomacy would have started much earlier. A lot of infrastructure would still get built. Economically, China might be ten years behind.

This might be true, but from an American perspective Mexico is, well, important. Increasing development and safety in Mexico matters far more to us than development and security in, say, Indonesia, or Kenya. That is the trade off suggested in my counterfactual: not what would happen to the global economy but what would happen to Mexico? Giving China permanent MFN status was a terrible policy all around, but the point of this post was to emphasize that it was also a decision put Beijing’s interests above priorities much closer to our own shores.

A disappointingly biased piece. TL;DR, “They stole our jobs!” (in Trump’s voice)

Following China’s admission to the WTO, manufacturing industries in many medium-income nations have been affected. We now see Mexico with high crime rates and huge drug issues. Why is this not happening in Eastern Europe, in India, or in other East Asian nations? Why just Mexico? Some analysis is needed. (Hint: Geography. Which neighbors does Mexico have? Where are those markets for drugs and guns?)

BTW, if you go to your local Walmart, you will find way more “Made in Vietnam” clothing than those “Made in China”. The low labor cost in Vietnam has pushed Chinese textile industry out of the market in recent years. Would the author see a similar drug and crime problem in China as in Mexico? Would he put a similar blame on Vietnam this time?

Well, that is actually Rozelle and Hell’s prediction — if they cannot rise up, they expect there will be a similar crime problem in China.

Time will tell.

P.S. Next time don’t use ‘anonymous’ as your tag or I will file your comment as spam.

> Well, that is actually Rozelle and Hell’s prediction — if they cannot rise up, they expect there will be a similar crime problem in China.

Do we have any examples of a East Asian middle income countries with big crime problems? I’m reasonably sure we don’t have any of East Asian countries on that income bracket with major public order problems for non-political reasons. Koreans riot for democracy but they don’t loot.

Middle-income trap for China? That sounds so last decade!

Scott Rozelle and Natalie Hell’s book is so out of touch with reality. China is classified as an upper-middle income country and it is the only developing country that is competing in AND in the top ranks of pretty much every hi-tech industry. It’s ranked 2nd in the authoritative Nature Index on high quality scientific publications after the US and has been closing the gap year after year. It is ranked as second in AI research and applications after the US, and it leads in quantum communication, quantum computing and supercomputers in the world. It’s now arguably the second leading space power, having successfully conducted a lunar sample return and sending a space station, the third-country to do both after the US and Russia, and has sent a spacecraft to orbit, land and rove the Mars, only the second country to do so after the US. In fact, it completed all in the first go, which took the US decades in many missions and the Soviet Union/Russia/ESA repeatedly failed to do.

And it is not just in the leading-edge science and technology. China leads the world in many tech and strategic industries, including solar power, wind power, ultra high-voltage power transmission, electric cars, battery, which form an end-to-end clean energy supply and demand chain. China is also among the leaders in high-speed rail, telecommunication equipment/5G, nuclear power plant, etc. And we don’t even need to mention China’s digital and mobile industry, which is extremely dynamic, innovative and huge. The list goes on. In fact, one could argue that China is one of the two tech superpowers of the world, along with the US.

Comparing China to Mexico – I don’t know what exactly to put it – and do it in 2020, is, well, right there with the WSJ article about China’s Chernobyl Moment last year and only slightly better than Gordon Chang’s book of The Coming Collapse of China.

People should put their emotion and ideology bias aside when it comes to evaluate China. While it’s fair to question whether China would fall into the middle-income trap five or ten years ago, by now, it’s ludicrously out of touch with reality.

The trouble of course is that there are two Chinas–the China of rocket ships and quantum advance, and the China of little concrete flats, bad toilets, and parentless homes. Rozelle spends his time in that second China. I don’t think ideology has much to do with his worry: he is, as I said, about as resolutely pro-China as any Westerner can be. He sees every day the situation on the ground obscured by the glimmers of Shenzhen and Shanghai. I would not be so quick to dismiss his worries.

You’re talking to a paid Chinese student, Tanner. Otherwise known as the 50-cent Army. It’s noticeable by the generic points (“China is also among the leaders in high-speed rail, telecommunication equipment/5G, nuclear power plant, etc.”) and the fact that he only superficially engages with your essay.

Everything is more impressive in China because it’s such a huge country, with an authoritarian government that can deploy the country’s resources how it sees fit, and this has created some excellent infrastructure and impressive developments in high-tech.

On the other hand, the truth is that China’s average income and human development across the board is not that different from countries like Mexico or Chile, and very far from the world’s most developed countries. If China had a normally-sized population, it would probably be viewed more like a developing country.

Really provocative and interesting piece. Thank you. Given the strength of the claim, I’d like to see the evidence worked through more thoroughly. For example. Honduras and Colombia have undergone similar devolutions: was the cause the same, or did Mexico’s devastation cause theirs, or did theirs have separate causes (which would imply a challenge, though not a refutation, of this theory)?

I’ve been living outside China for over 20 years, so my knowledge of Chinese primary and secondary education may be outdated. In the 80s when I was going to school in China, students not at the grade level got retained. I had primary school classmates that were retained 3 times or more. I expect that many of them did not make it to high school graduation. Things might have changed drastically since then, but I would not be surprised if an average Chinese high school graduate is substantially more capable than an average US high school graduate. 30% high school graduation rate cited above needs to be understood in this context.

This is surely the right approach; the approach used in current American ghettos is simply unacceptable. A country’s education system must be based on standards, not some arbitrary metric of what percentage of the population must qualify.

Are you equating China to Mexico??

1. China has the 4 bigs: biggest economy (according to the US government), biggest market, biggest production base, and biggest talent base.

2. As Elon Musk said (I paraphrase) “China only has to have half of gdp-per-capita to have economy twice of the US. It’s just math ”

3. Do you see many Mexican STEM graduate students in the US (unlike 40% of STEM graduate students in the US are Chinese)?

There used to be a really good number crunchy industrial blog I used to read at about the same time I started reading your blog – long time ago.

I remember him talking about how Mexico was losing jobs to China.

But his point was that advances in technology were putting pressure on all industrial economies, including China. He felt that China had a struggle ahead of it much like you are talking about. Low wage work would go shift to even less expensive locals, and the high end stuff would lose jobs to technology.

To be honest, I am not convinced that the world has really figured out a way to skillfully navigate a post-industrial world.

> China fairs poorly on this metric


I would be delighted to proofread your posts if you would like that.