Cormac McCarthy’s contribution to the theory of increasing returns

I really enjoyed this anecdote about the writing of W. Brian Arthur’s classic article on increasing returns from 1996:

As we are wrapping up the interview, he [Arthur] tells me an anecdote about the creation of that Harvard Business Review article. “I don’t know if you know the writer Cormac McCarthy,” he begins, “but I was very good friends with him at the time. I mailed the draft down to Cormac, who was in El Paso or somewhere like that. When I didn’t hear from him, I called him up and said, ‘Did you like my increasing returns article? It’s for the Harvard Business Review.’ There was kind of a silence on the line. And then he said, ‘Would you be interested in some editing help on that?’ Next time he’s in Santa Fe we spent four days on that piece. He took apart every single sentence, deleted every comma he could find. I said, ‘You can add that piece to your Collected Works, it will be somewhere in between Blood Meridian and All the Pretty Horses.’

“Let’s say the piece was better for all the hours Cormac and I spent poring over every sentence. The word got back to my editor at Harvard Business Review. She called me up, in a slight panic, and says, ‘I heard your article’s getting completely rewritten.’ And I said, ‘Yeah!’ She says, ‘By Cormac McCarthy? What did he do to it?’ And I said, ‘Oh, well, you know, pretty much what you’d expect. It now starts out with two guys on horseback in Texas, and they go off and discover increasing returns.’ And for a couple of seconds she was aghast.”

The full piece is from Fast Company, which has more on how the concept of increasing returns was used and abused by the technology industry in the years since its popularization. And indeed Arthur’s HBR article–it’s worth rereading–is extremely well-written, with many more simple, punchy sentences than are the norm for business or economics writing. It is hard to see any way to improve on the clarity of sentences like:

Increasing returns are the tendency for that which is ahead to get further ahead, for that which loses advantage to lose further advantage.

My guess is that McCarthy probably doesn’t deserve all the credit for the virtues of the prose, as Arthur is himself a very clear thinker and good writer (his book on technology is still one of my favorites). But everyone benefits from a good editor.

Can economics offer more than a counsel of despair to struggling places?

I just finished Enrico Moretti’s The New Geography of Jobs, an admirably clear book about one of the most important trends of the day: the increasing concentration of American jobs, wealth and economic activity in a small number of urban centers. He argues that technology boomtowns like Seattle and San Francisco are what they are today in large part because of historical accidents that set off positive feedback loops, rather than because of any particularly enlightened policy. This means that it is not very obvious what all the cities that are instead trapped in negative feedback loops, losing population and jobs, should do:

People often have unrealistic expectations of their governments. The role that local governments can play in revitalizing struggling communities is less extensive than most voters realize and most mayors would like to admit. The reality is that a city’s economic fate is in no small part determined by historical factors. Path dependency and strong forces of agglomeration present serious challenges for communities without a well-educated labor force and an established innovation sector.

He is careful not to say that there is nothing to be done in the face of the pitiless onslaught of market forces, but it’s also clear that he thinks, probably quite rightly, that many local development policies (like tax subsidies to large employers) are ineffective and a waste of money. In the end he proposes mostly national policies: substantial increases in research and development funding, improved education, more openness to highly skilled immigrants. Rather than try to hold back the forces that are concentrating the economy in a small number of urban centers, in other words, the US should try to supercharge them, in hopes that even more centers will develop and allow more people to benefit.

The conclusion that benign neglect is the only real option for dealing with regional inequality seems to be the consensus wisdom of the economics profession. Since the US election though, there has been a pretty dramatic backlash against this counsel of despair. Here are three pieces that I found excellent, all of which are worth reading in full.

Adam Ozimek has a quite measured and detailed post:

The level of nihilism espoused by economists about what we can do to help struggling places in the U.S. is, quite frankly, strange. Whenever the issue of helping places is raised, critics jump straight to the most extreme examples, such as former mining towns. But the fact that some places need to shrink, and the costs of helping some places sometimes outweighs the benefits, is a far less powerful point than these critics imagine. Other places have survived the loss of major industries and gone on to thrive. Understanding why this happens sometimes and doesn’t happen other times, and what policymakers can do to help replicate the successes, are crucial policy issues that cannot be pushed aside by pointing out the impossibility or desirability of saving every place.

Finally, it’s important to note that the competition between thriving metropolises and the now-struggling parts of the country need not be zero sum. Increasing the human, social and physical capital of struggling places in this country can reduce the need for economic transfers at the federal level and can help make an overall more tolerant and open society that is better able adjust to the dynamism and globalism needed for a growing modern economy. It may help prevent residents in these places from desperately voting for policies that will only make things worse, like a trade war or immigration restrictions. These policies don’t make any economic sense, but when the best ideas for helping struggling communities consists of getting their most able residents to move away, it becomes a little easier to understand.

In a long and interesting piece, Steve Randy Waldman argues that not all of the self-reinforcing dynamics of urban concentration are necessarily positive, and that the political downsides are now pretty obvious:

Cities are great, but I think the claim that everybody moving to the very largest cities would yield a massive, otherwise unachievable, productivity boost is as implausible as it is impractical. Historically, economic activity was far less concentrated during the decades when America enjoyed its strongest growth. Perhaps technology has changed everything. But perhaps much of the apparent productivity advantage enjoyed by large, powerhouse cities over medium-sized cities is due to creaming, sorting, and particularly high-powered coalitions of rent-extractors, rather than hypothesized quadratic-returns-to-scale human connectivity effects.

Then, of course, there is all the stuff that economic analysis tends to overlook: Community, history, attachment to family, attachment to the land itself, the perhaps quaintly aesthetic notion that a civilized country should not be composed of gleaming islands in a sea of decay and poverty. And politics. Politics seems to be a thing now. Rightly or wrongly (and I think the question is more complicated than many of us acknowledge), the United States’ political system enfranchises geography as well population. …In the American system, piling people into a few, dense cities is a sure recipe for disenfranchising most of the humans. A nation of mid-sized cities distributed throughout the country would both spread the wealth geographically and yield a more balanced politics than the dream of hyperproductive megacities.

Finally, a fantastic and impassioned piece by Ryan Avent also tackles the regional inequality question, among many other recent failures of economics:

The economic literature is pretty clear that moving people from low productivity places to high productivity places is very good for both the people that move and the economy as a whole. It’s also pretty clear that place-based policies designed to rejuvenate regions which have lost their economic reason for being tend not to work very well. And one logical conclusion to draw from these lines of research is that government ought to care about people rather than places, should focus aid to struggling places on things like cash transfers or retraining schemes or efforts to boost the housing capacity of booming regions, and should not be sentimental about the prospect of once proud industrial cities emptying out. And maybe that logical conclusion is the right one.

But maybe that’s not the right conclusion at all. Maybe the right question, once again, is which is likely to be more corrosive of the legitimacy of valuable macroinstitutions: the long-run decline of whole regions of advanced economies, or the inevitable waste and inefficiency that would accompany an effort to revive those declining regions. And perhaps benign neglect would win that argument. Yet the argument ought to take place; economists should not ignore the relevance and importance of macroinstitutions and assume that the inefficiency is the clinching argument.

The resource curse is alive and well in Mongolia

Catching up on my reading after a break, I see that the excellent Bill Bikales has written a nice summary of the unfolding economic crisis in Mongolia, where the currency is plunging, borrowing costs are rising and boom-era debts are suddenly looking very doubtful:

The crisis traces back to 2012, when a new Mongolian coalition government took office facing extremely favorable economic conditions, including high mineral prices and strong demand from China. Gross domestic product had grown by 17.3% in 2011 and by another 12.3% in 2012, making the country a global leader.

Investment flowed into Mongolia as a result of an agreement with Rio Tinto to develop the massive Oyu Tolgoi copper-and-gold resource in the Gobi Desert. There was also strong interest in the equally massive Tavan Tolgoi coal deposit in that region, along with other coal, iron and copper deposits.

But the new government had won election by making highly populist promises, and this led to a contradictory agenda. On the one hand, the government attempted to renegotiate the already signed Oyu Tolgoi agreement, and in general started seeking better terms from foreign mining firms. This led to a quick drop in investment, growth and revenues. At the same time, the government rapidly expanded spending on housing, government salaries, social welfare and pensions.

The only way the government could finance the resulting large budget deficit was by borrowing. For the first time, Mongolia became a significant global issuer of commercial paper. Between 2012 and June 2016, the government raised $3.6 billion, roughly one-third of GDP, on global bond markets, paying high interest rates. There was also a massive buildup of domestic debt. In a throwback to the planned-economy era, the banking sector once again became a major financier of government programs. Total loans in the economy doubled in the first two years of the 2012 government’s term, and the money supply expanded at an extraordinarily rapid pace. Nonperforming loans began to build up. …

By 2014, international financial institutions expressed measured but clear concern about the deteriorating economic situation. The central bank slowed monetary expansion and budgets were tightened somewhat. This coincided with a continued collapse in foreign investment and a steady decline in global mineral prices due to China’s slowdown. As a result, Mongolia’s growth slowed sharply to 2.3% in 2015 and is likely to be zero or negative in 2016.

But the current economic downturn isn’t primarily due to a decline in global commodity prices. It is the result of the government borrowing heavily against future export earnings while taking actions that deferred the day when those exports would materialize. Instead of preparing for an inevitable cyclical downturn in commodity prices, the government took steps that magnified that downturn’s impact.

A sovereign debt default now looks very much like a live possibility for Mongolia. This sad narrative fits very well the best current understanding of the resource curse–which is not that possession of natural resources mechanically causes lower growth. There are enough countries that manage to do well while having large energy or mining industries (such as the US) so that attempts to find correlations between resource endowments and growth outcomes have had decidedly mixed results. Rather, the problem with having a big resource sector is that it exposes a country to the huge boom-bust cycles typical of commodity markets–and it is rare for countries to be able to make good decisions at either end of a commodity cycle. The temptations to make borrow too much and make bad investments in the boom days is particularly strong; note this sentence from a recent World Bank paper: “Credit growth has been most pronounced, and nearing the pace associated with past credit booms, in commodity exporting countries.” As Bikales shows, Mongolia has problems because it made bad decisions, not just because it had a mining boom. One useful recent summary of the literature on the resource curse is Cullen Hendrix and Marcus Nolan’s Confronting the Curse: The Economics and Geopolitics of Natural Resource Governance, who conclude:

Natural resources are neither discovered nor exploited in an institutional vacuum. Preexisting institutions are the key moderating factor. If these institutions are strong and the size of the mineral sector does not dwarf the rest of the economy, resource wealth provides additional capital for productive investment. Even if Dutch disease dynamics come into play, these resources can be invested in ways that promote intergenerational equity and the accumulation of long-term wealth. Under these circumstances, resource income is growth promoting, and the “curse” becomes more of a blessing. This condition seems to be the equilibrium path of Norway, the United Kingdom, the Netherlands, and the United States. If preexisting institutions are weak and the mineral sector is much larger than the rest of the economy (as in Angola, Nigeria, and Saudi Arabia), the resource curse dynamic emerges.

An interesting comparison is available just over the border in the Chinese province of Inner Mongolia, which is a huge coal producer and has experienced a similar boom-bust cycle along with commodity prices. Both Mongolia and Inner Mongolia are currently enjoying nominal GDP growth of less than 4%, down considerably from their recent highs–though Mongolia’s peak nominal growth rate neared 50% while Inner Mongolia was–only!–around 25%.


The province of Inner Mongolia of course does not have its own currency and does not borrow internationally, so it is not going to experience the same type of fiscal and currency problems as the independent nation of Mongolia. And while the governing institutions in the two places are quite different, neither place is exactly pursuing Norway-type best practices for managing their resource wealth. So there is also some evidence of the resource curse playing out in Inner Mongolia, essentially meaning bad economic decision-making during commodity booms. The famous “ghost city” of Ordos could be one supporting anecdote; however that example is probably overplayed, as Wade Shepard reports: “The real story consists of a mining boomtown building a new district on a long-term timeline in a period when hundreds of other cities across the China were doing the same thing.” On a macro level however it seems pretty clear that investment got even more out of hand in Inner Mongolia than in the rest of China, and is correcting harder:


Is China’s growth now increasing rather than reducing global inequality?

Here is an interesting tidbit from Branko Milanovic’s latest book, Global Inequality: A New Approach for the Age of Globalization. Much of the book is about the recent, unusual combination of a trend for inequality to rise within countries (as the upper classes take a larger share of each nation’s income) and a trend for inequality between countries to fall (as rising incomes in developing countries narrow the gap between the haves and have-nots on a global basis). China has been the main driver of the latter dynamic, but we may already be at a turning point in that trend–one that will require India to keep growing if global inequality is to keep falling:

Population-weighted intercountry inequality has been uniformly decreasing since the late 1970s, since about the time when China introduced the “[household] responsibility system” (de facto private ownership of land) in rural areas and growth picked up. Moreover, convergence (the decrease in intercountry, population-weighted Gini values) has been remarkable and has accelerated in the first decade of the twenty-first century. We have already seen that this movement was the key factor behind the decrease in global inequality and the broadening of the global middle class. …

China’s role as the main engine driving the reduction in global inequality becomes less important as the country gets richer. In 2011, China’s mean per capita income, calculated from household surveys and expressed in international dollars, was 22 percent below the global mean and was greater than the mean incomes of 49 percent of the people in the world (assumed to have the mean incomes of their countries).

The world will very soon be in the position where China’s high growth rate begins to add to global inequality, not detract from it. India’s mean income is currently ahead of only 7 percent of the world population, and India cannot be expected to “turn the corner,” that is, to become, in average per capita terms, richer than more than 50 percent of the world population, in the next twenty years. Thus it will, if it grows fast, take over from China as the main engine of global income equalization.

The technicalities are interesting and worth citing in full:

Footnote 16: In the case of the Gini coefficient (with which we work here), the point at which a unit begins to add to inequality depends on its rank (let’s call it the “turning point rank”), that is, the number of units from which it has a higher income, but also on the initial Gini. The turning point rank formula is i > ½ (G + 1)( n + 1) which for a large n simplifies to i > ½ (G + 1) n, where i = the turning point rank (the rank i runs from 1 to n), n = total number of units, G = Gini coefficient. Note that the turning point is n/ 2 (i.e., the median) only when the Gini is zero. For the derivation of the formula, see Milanovic (1994).

With the current level of population-weighted global Gini being around 0.54, the turning point rank is 0.77n. That means that China’s mean income has to be such that, when all individuals in the world are ranked by the mean incomes of their countries, 77 percent of the world population is left behind China. But because China’s population is 20 percent of world population, for a Chinese person to be at that (“turning”) point, he or she needs to leave behind only 57 percent (77 − 20) of the world population. Currently, as we have seen, China’s mean income exceeds the mean income of 49 percent of world population. This means that China needs to leave behind just an additional 8 percent of people in the world to begin adding to global population-weighted inequality. This could already be happening by the time this text is being read.


When “It’s the economy, stupid” falls short

I read the news today, oh boy. For some clear thinking if not reassurance, I recommend an essay by Yascha Mounk at Project Syndicate, in which he surveys various people’s takes on the recent global political instability, and comes down fairly hard against the economists’ argument that all this stuff is the result of stagnating incomes. Here is a condensed excerpt, though the whole thing is worth reading:

In what sense are politicians as different as Trump, Erdoğan, and French National Front leader Marine Le Pen connected? Does the anger that has set so many countries’ voters against their political establishment have common causes – and, if so, are there common remedies that can halt the rise of populists? …

According to Bill Emmott, former Editor of The Economist, the reasons are straightforward: “the interests of ordinary people have been subordinated to those of the elite.” … Harvard’s Dani Rodrik agrees, arguing that “the internationalization of markets for goods, services, and capital drives a wedge between the cosmopolitan, professional, skilled groups that are able to take advantage of it and the rest of society.” …

The problem with these explanations is the mismatch between the root causes of popular anger and the form this anger takes in most countries. Supporters of Trump or the Netherlands’ Geert Wilders may blame free-trade agreements for eroding job opportunities; but the bulk of their energy and anger is directed not at prevailing economic orthodoxy, but at social policy. The defining feature of their political brand is hatred of immigrants, not of the World Trade Organization. …

Joschka Fischer, a former foreign minister of Germany, has a more radical view of the economic causes of an essentially cultural rage. He argues that the “White Man’s World” is under attack from the “globalization of labor markets, gender parity, and the legal and social emancipation of sexual minorities.” Immigration is the “issue that brings that prognosis home (not just metaphorically) to today’s angst-inspired nationalists.”

Fischer’s view helps to explain a puzzle identified by Daniel Gros, Director of the Center for European Policy Studies in Brussels. If losses from globalization “account for the rise of populism, they must have somehow intensified in the last few years, with low-skill workers’ circumstances and prospects deteriorating faster vis-à-vis their high-skill counterparts.” But “that simply is not the case,” Gros shows, “especially in Europe.” While the economic transformations of recent decades help to explain falling trust in existing political institutions, it is facile to assume that it is primarily globalization’s immediate losers who support the populists, much less to expect that an upswing in the business cycle will halt the populists’ rise. …

The most optimistic observers emphasize that the costs of globalization stem from politics, not economics. As J. Bradford DeLong of the University of California at Berkeley argues, income stagnation was caused not by globalization, but rather by politicians who have “failed to implement policies to manage globalization’s effects.” … But if such desirable policies exist, why should our political systems, having failed to pursue them in the past, be able to do so now?

In my view, the choices facing us in the next decades may be far starker than the optimists admit. There are three reasons why attempts to shore up living standards are unlikely to stem the populist threat: the right policies might not be adopted; even if they were adopted, they might fail to redress economic grievances sufficiently; and, most important, even if they did redress economic grievances, they would not necessarily defuse the culturally-based anxieties of many citizens. As Harvard’s Ricardo Hausmann emphasizes – and as the Brexit vote clearly showed – the “sense of ‘us’” is more important to many people than achieving greater success through integration with “others.” …

This implies that redistribution and compensation of globalization’s losers will not be enough, and that liberal democracy is likely to become an increasingly unstable political compound.

I find this convincing albeit somewhat depressing. I think it’s pretty clear that the Trump phenomenon is mostly about white identity politics rather than economic issues, and that the Brexit vote was also mostly about English nationalism rather than economic issues. Greg Ip’s recent column has a good summary of research on anti-immigrant sentiment, and how it is in fact driven mainly by national identity rather than economic issues.

The benefit of an economic explanation of the populism/nationalism/whatever thing that is going on in so many different countries is that it can unify what is being explained: all these events are manifestations of the underlying trends in the global economy. I’m don’t find the economic explanations that convincing, but the alternatives are admittedly more complicated. If the working-out of national identity issues is central, as it indeed seems to be, that makes it harder to explain in terms of a cross-country phenomenon.

So I don’t have the unifying field theory for all this, but I do think it’s worth thinking more about how to understand nationalism. One basic conclusion is that politics is not in fact economics by other means, but actually about politics. And there is no more crucial political task than defining the nature and boundaries of the political community. Reflexively looking for the hidden economic interests underlying a political position is a Marxist fallacy if there ever was one–invoking “it’s the economy, stupid” is not always the smart move.

Explaining catch-up growth with China and commodities

The World Bank’s latest Global Economic Prospects report may be a 194-page document, but most of the attention it got was for one little infographic. The Financial Times focused its coverage on the chart, and the Economist also made it one of their charts of the day. As the bank helpfully made the underlying data available, it is easy to reproduce, so here is the original:


That’s indeed a very nice chart, showing that catch-up growth is not a constant phenomenon, but one that has risen and fallen over the last couple of decades. I like the chart too, but when I first looked at it, I thought: I’ve seen that curve somewhere before. Because I’m interested in regional growth patterns, I have been looking at catch-up growth within China: how quickly have poorer provinces been closing the income gap with the wealthier provinces? (I chose Shanghai as the reference point, since it has been the most developed part of China for many decades.) And when I took my provincial catch-up data and overlaid it with the World Bank’s global data, this is what I got:


I would say those trends are pretty much the same: fewer places experiencing catch-up growth in 1997-2001, a widening of catch-up growth to more places from 2002-2012, and more recently a sharp fall off. So that’s pretty interesting: catch-up growth within China, and catch-up growth across lots of other developing countries, seems to follow the same pattern.

One possibility is that catch-up growth is just a function of growth, and so when global/China GDP growth is slow, catch-up growth is less widespread. But this doesn’t explain why catch-up growth has faded so sharply in the last couple of years: while both global growth and trade volumes are not doing that great, they also have not gotten suddenly worse. What has declined very sharply are commodity prices, thanks to an oversupply generated by producers who thought China’s housing construction boom would go on longer than it actually did. So I think commodities may be more important for the pattern of emerging-market catch-up growth than the World Bank acknowledges.

This does not mean that I’m arguing commodity exports are actually a great thing and that it’s really too bad that commodity prices have fallen. I firmly agree with the conventional wisdom that commodity exports are not an effective or sustainable way for developing countries to become rich. But remember what is being measured in these lovely charts: not the number of people whose incomes are converging with developed-country standards, but the number of countries (basically a diffusion index). And my intuition would be that more developing countries are, if only by default, commodity exporters, simply because the alternative development model–exporting manufactured goods–is in fact quite hard to do.

The data support this intuition. If I split developing countries into two baskets, manufactures exporters and commodity exporters, on the simple criterion of having more or less than half their exports in manufactured goods (a concept I borrowed from Jon Anderson), the majority of developing countries are in fact commodity exporters. For the low and middle-income countries in the World Bank’s World Development Indicators database, only 33 of 96 countries had more than 50% of their exports in manufactured goods in 2011. The same pattern holds internally within China: while most of China’s population is concentrated along the coast, most of its provinces are not. Of China’s 31 provinces, only 10 are officially classified as “Eastern.” The diffusion index for catch-up growth within China will therefore be dominated by the central and western provinces, and these provinces have more commodity-driven economies. To be precise, I estimate that the mining and metals share of GDP is higher than the national average in all but four of the 21 central and western provinces.

This pattern of catch-up growth is not just a statistical artifact, but gets at a real phenomenon. The same economic role has been played by a group of provinces within China’s borders, and a large group of countries outside China’s borders. Both prospered by supplying materials for China’s housing boom (the underlying cause of the commodity boom), and both are seeing that prosperity erode now that the housing boom is fading. I keep discovering that housing is the answer to many economic questions about China; it seems that Chinese housing also explains a lot about the patterns of global growth.

Freeing up China’s service sector: the why and the how

I have a new paper out at the Paulson Institute, on China’s service sector. Obviously I would like people to read the whole thing, but here’s the short version:

The first part of the paper is a rather fun (for me anyway) piece of data work, comparing the development of China’s service sector over time with the patterns of Japan, South Korea, and Taiwan. The conclusion is that China’s “modern” service sector looks over-developed, while its household service sector looks under-developed. The over-development of modern services seems to caused mainly by a huge increase in the size of the financial sector, and therefore is not actually a cause for celebration. The under-development of household services suggests (though admittedly does not prove) that these sectors are being held back by high levels of state ownership and/or regulation. Therefore I argue that instead of doubling down on finance-driven growth, China would be better off liberalizing household and other service sectors.

In the second part I offer some suggestions on how to do this. My main idea is that China should make use of its well-developed planning institutions to drive liberalization. In simple terms, I propose that the Chinese government set quantitative targets for the market share of private companies in various sectors. This would make clear what the end goal of liberalization is, and impose accountability on officials for achieving that end goal–but it would also allow for flexibility in how the goal is reached. There is some precedent for this approach: in the twelfth five-year plan for the healthcare sector, the government set a target that private hospitals should account for 20% of hospital beds and service volume by 2015, and their market share did rise. Similar targets could work in other sectors. It’s worth a try, I think.

My thanks to Evan Feigenbaum and particularly Song Houze of the Paulson Institute for their help and comments on this paper, and I am also very pleased that the Institute’s resources allow the paper to be translated into Chinese.