By Juan Pablo Chauvin (June 13, 2018)

Originally published on the LSE American Politics and Policy Blog.

By the year 2100, the planet will have, according to forecasts, twice as many people living in cities as it does today. The lion’s share of this transformation will take place in the developing world, where the urban population could go from 2.6 billion to close to 8 billion in less than a century. What will this new chapter in the history of urbanization look like? Will it be a “re-make” of what we saw in the 20th century in now-rich countries? Will it come accompanied by stark improvements in living conditions as it did before? How will this depend on the actions of local and national policymakers over the next couple of decades?

These are broad and complex questions, which have major policy implications. They are ultimately about the economic opportunities that will be available -or not- for billions of people. And about who among them will have access to those opportunities. Yet we are still far from having enough satisfactory answers. Economists that study cities have traditionally focused on the US and other high-income countries, and we are only beginning to explore to what extent what we know about cities in the US is useful to understand cities in lower-income countries.

In a recent work, Ed Glaeser, Yurean Ma, Kristina Tobio and I delve into this question. We look at some of the key facts that researchers have learned about urbanization by studying the US and assess whether they are also valid in three large countries with lower income levels: Brazil, China, and India. We find that the main patterns that characterize US urbanization can also be found in the developing world, but not in all countries, and critically, not with the same intensity.

Take the “spatial equilibrium” concept, one of the most powerful tools to understand the urban system in the US. It is the idea that cities don’t “offer free lunch”. If the city is attractive along certain characteristics, like offering access to high-paying jobs, it will be unattractive along others, like having expensive real estate or traffic congestion.

Such trade-offs are repeatedly found across American metropolitan areas. For instance, as the graph in the upper-left corner of Figure 1 below shows, high-wage cities tend to also have high housing rents. But as seen in the rest of the graphs of that figure, the same is not always true in lower-income countries. Brazil looks the most like the US, in terms of the tight positive connection between wages and housing rents across cities. In China, this relationship exists, but it is much looser. And in India, we were not able to find the same pattern in the data at all.

This may reflect, at least in part, differences in within-country mobility. When a city improves its economic opportunities or its quality of living it attracts migrants from other cities. They, in turn, drive up housing costs and congestion, reducing the attractiveness of the city. But if there is little internal migration, these adjustments may not take place. Internal migration has only picked up in China in recent years with the reforms of the hukou system, while internal migration in India remains surprisingly low.

Whether the predictions of spatial equilibrium hold in a country matters for policy. For instance, governments around the world continue to invest in “place-based policies”, geared towards improving economic opportunities in economically-lagging areas. With spatial equilibrium forces at play, such policies may disproportionally benefit migrants and home-owners, who are likely to see the value of their properties increase. Without these forces, workers originally residing in the region are more likely to enjoy the benefits of the policies.

Crowded Old Delhi seen from the Jama Masjid – India” by Erik Törner is licensed under CC BY NC SA 2.0

While spatial equilibrium can explain how firms and workers locate across space, it tells us little about what drives urban success. In addressing the latter, economists frequently point to the existence of “agglomeration economies”. People and firms become more productive if they locate close to one another, and more economic opportunities are created in densely-populated places.

In line with prior studies of agglomeration economies in US cities, we find that doubling population density corresponds to an increase of around 5 percent in wages. Replicating the analysis in the other countries we find a similar figure for Brazil, a larger number for India (8 percent), and a much larger number for China (19 percent). Other studies have convincingly shown that at least part of the connection between density and productivity is causal.

If agglomeration economies are stronger in lower-income countries as our results suggest, policies that seek to limit internal migration and the size of cities may end up hurting productivity growth and the creation of economic opportunities. Conversely, policies that facilitate housing supply and accessibility in places that already have a significant concentration of economic activities, will likely foster productivity growth.

Another major driver of economic success in US cities is human capital. In the US and other high-income countries, individuals that work in better-educated places tend to become themselves more productive and earn higher wages. Partly because of these “human capital externalities”, more educated cities tend to grow faster.

Looking at US cities we find results that are consistent with what other researchers have documented in the past. When the college-educated share of the adult population increases by 10 percentage points, wages in the city increase by about 10 percent. The equivalent figures are, however, significantly larger in our three comparison countries: 19 percent in India, 47 percent in Brazil, and 53 percent in China. These results should be treated with caution, since we are unable to show with our data to what extent these connections reflect causal relationships. But they clearly suggest that human capital may be an even more important force behind the success of cities in countries in lower-income countries.

The notion of human capital externalities has been used as an argument in favor of a broader public investment in local education. However, it is unclear that this alone would be the best policy response from the city perspective. Investments in education may not effectively increase local human capital since more educated people will eventually leave if they cannot find adequate employment in their city. What makes difference in productivity is to have educated people living and working in the city, which highlights the importance of policies geared towards attracting and retaining highly-educated individuals.

While spatial equilibrium can explain how firms and workers locate across space, it tells us little about what drives urban success. In addressing the latter, economists frequently point to the existence of “agglomeration economies”. People and firms become more productive if they locate close to one another, and more economic opportunities are created in densely-populated places.

In line with prior studies of agglomeration economies in US cities, we find that doubling population density corresponds to an increase of around 5 percent in wages. Replicating the analysis in the other countries we find a similar figure for Brazil, a larger number for India (8 percent), and a much larger number for China (19 percent). Other studies have convincingly shown that at least part of the connection between density and productivity is causal.

If agglomeration economies are stronger in lower-income countries as our results suggest, policies that seek to limit internal migration and the size of cities may end up hurting productivity growth and the creation of economic opportunities. Conversely, policies that facilitate housing supply and accessibility in places that already have a significant concentration of economic activities, will likely foster productivity growth.

Another major driver of economic success in US cities is human capital. In the US and other high-income countries, individuals that work in better-educated places tend to become themselves more productive and earn higher wages. Partly because of these “human capital externalities”, more educated cities tend to grow faster.

Looking at US cities we find results that are consistent with what other researchers have documented in the past. When the college-educated share of the adult population increases by 10 percentage points, wages in the city increase by about 10 percent. The equivalent figures are, however, significantly larger in our three comparison countries: 19 percent in India, 47 percent in Brazil, and 53 percent in China. These results should be treated with caution, since we are unable to show with our data to what extent these connections reflect causal relationships. But they clearly suggest that human capital may be an even more important force behind the success of cities in countries in lower-income countries.

The notion of human capital externalities has been used as an argument in favor of a broader public investment in local education. However, it is unclear that this alone would be the best policy response from the city perspective. Investments in education may not effectively increase local human capital since more educated people will eventually leave if they cannot find adequate employment in their city. What makes a difference in productivity is to have educated people living and working in the city, which highlights the importance of policies geared towards attracting and retaining highly-educated individuals.

I have mentioned only a few ways in which urbanization in developing countries may be different from what we have seen in the US and other rich countries. This is part of a still-small but growing body of studies focusing on the economy of developing-country cities. But the questions that remain open are vast, and the stakes of getting the answers right are high. Just think about the fact that two-thirds of the urban infrastructure in Africa will be built between now and 2050. Getting urbanization right could substantively improve the lives of millions of people in some of the poorest countries on earth in just a few decades. Getting it wrong could seriously constrain the development opportunities in those places for generations.

What should the policy priorities be in these and other rapidly urbanizing contexts? What can governments and firms do to take advantage of the opportunities and to mitigate the risks? How can these transformations be leveraged to improve the distribution of economic opportunities? What can policymakers in places like Africa and South Asia learn from countries that urbanized more recently and faced similar challenges, like most of Latin America? We need better answers to these increasingly urgent questions.

By Juan Pablo Chauvin (February 12, 2019)

Originally published on Ideas Matter, the blog of the IADB’s Research Department. It is also available in Spanish here.

When it comes to urban economic development, everything is a question of tradeoffs. There are, as economists like to say, “no free lunches.”

Consider California. The astronomical wages paid by firms like Google, Apple and other digital firms draw thousands of people to the Silicon Valley/San Francisco area, and the high wages of the Hollywood studios lure people to Los Angeles. But those big salaries also result in soaring housing costs that make it virtually impossible for many low-income or middle-income people, like teachers, to buy a decent home or even pay rent. The state’s amenities—like nice weather—have a similar effect, attracting well-to-do sun seekers while making life unaffordable for many long-time residents. That, in a nutshell, is the urban tradeoff problem: good news on one front, typically results in bad news on another.

The tradeoffs in Latin America

The existence of such tradeoffs is a useful lens for thinking about internal labor migration in Latin America and the Caribbean, a region, which increased its percentage of city dwellers from 40% in 1950 to nearly 80% in 2011 to become the most urbanized area of the developing world. The movement to cities, which has resulted in eight megacities of more than 10 million people, is on the whole a positive phenomenon for the region. It brings together firms and people, creating immense economies of scale and a cross fertilization of ideas in both industry and services. It generates prosperity and over the long run helps pull people out of poverty. But the large influx of people into the major cities also causes housing shortages which drive rents up. According to a United Nations report issued in 2012, Latin America suffers from a housing shortage of around 50 million homes, with nearly 20% of the population living in shantytowns, many of them without basic services like electricity and sewage.

Why should policymakers care? As it turns out, because these tradeoffs exist, local development policies often have unexpected and even unwanted consequences. City governments around the world have historically promoted projects to boost employment and improve the livelihoods of the local populations that elected them. The establishment of an industrial park, for example, is likely to create jobs. Economic opportunities in the city, however, also lead to internal migration , and some of the new arrivals may take up informal housing in slums. Where there are institutional or physical barriers to the slums’ expansion, they will move to formal housing elsewhere in the city. That, in turn, will cause local rents to rise—perhaps to soar— harming the well-being of local residents. While migrants will have better economic opportunities than in their place of origin, the local population will lose out—that is, everyone except landlords.

Gender impacts migration

To complicate things further, gender is an important factor in this dynamic, as I found in a recent study. Whether job creation favors male or female employment in a city is likely to matter because individual families tend to move to places where the men, rather than the women, have the greatest job prospects. It turns out that this family decision can have major implications for who benefits from local economic development policies.

My study encompassed 539 localities in Brazil, which grew at different rates over the 1990s and the 2000s. To measure how local job creation favored males or females, I used gender differences in employment among industries in those localities. Clear and consistent patterns emerged. Growth in industries that typically hire men, like construction, transportation, petroleum refining and others, led to significant rises in both urban migration and urban housing rents, as migrant men brought their families with them to their new dwellings in the city. But the same was not true for growth in industries that typically hire women, like the garments industry.

Indeed, while a 1% increase in demand for male labor resulted in a 0.76% increase in local population during the 2000s and 0.63% in local rents during the two decades under consideration, the same increase in demand for female labor resulted in only a 0.09% increase in urban population and a 0.01% increase in rents. Job creation for men, in other words, was stoking urban migration. But increases in jobs for women were more frequently being filled by locals, without generating stress on local housing market markets or negatively affecting the local population.

Policymakers in cities face complex choices on job creation

This is, of course, only one of the myriad factors that determine who benefits from local economic development policies. But it illustrates the complex choices that policymakers at the city level must make. If they seek to improve economic opportunities for locals, they may choose policies that minimize potential migration responses—like promoting work in areas that employ women over those that employ men. Achieving this goal, however, would come at the cost of reducing opportunities for potential migrants and landlords, and stalling the potentially large contribution of these groups to future urban growth. There is indeed no “free lunch”, and explicitly considering the tradeoffs at play can help local officials design policies that are more effective in achieving their goals.

By Juan Pablo Chauvin (April 11, 2019)

Originally published on Ideas Matter, the blog of the IADB’s Research Department. It is also available in Spanish here.

When Brazil’s government decided in 1967 to create a free trade zone in Manaus, it had glittering visions of Amazonian development. Located at the meeting point of two tributaries of the Amazon river, the so-called “City of the Forest” had fallen on hard times since the end of the rubber boom nearly a century before. Generous reductions in import tariffs and other tax breaks, the government thought, would encourage manufacturing firms to set up there and turn the once elegant port city into an engine of regional prosperity again.

Today the policy remains controversial, and the evidence for that vision’s success is ambiguous at best. Government action does appear to have had a positive effect on Manaus’ GDP per capita.  However, the bulk of the growth came from the expansion of the services sector — primarily from tourists entering the city to buy consumer goods at cheaper prices.

Meanwhile, there was no significant impact on productivity in manufacturing. And in agriculture, productivity fell as workers left productive jobs in the countryside to work in factories that had become profitable through subsidies rather than through improved levels of efficiency.

The dilemmas of place-based policies

The story of Manaus is important in part because it illustrates a key dilemma of place-based policies, those government initiatives targeted at a particular geographic area, like a region, city or neighborhood, rather than focused on groups of people.

Such policies typically involve investments to create jobs and increase productivity in a given locale.  But they also create artificial incentives for firms and workers to relocate from places where they would have been more productive. As a result, they  may benefit the locality where they are implemented, but hurt other places, and the national economy as a whole.

Moreover, place-based policies may fail to even help the places they were intended to benefit. As I discussed in a previous blog, if a local development policy creates economic opportunities, the resulting influx of people into the city can also drive rents up and slow down wage growth. With the exception of homeowners, that would negatively impact local residents.  Migration, in this dynamic, can undermine policymakers’ good intentions of improving an area.

Unexpected consequences when people leave their hometowns

Unexpected consequences also may result from people leaving their hometowns.  Consider the case of educational investments. A wide body of evidence shows that more educated cities tend to have more economic growth. Educational spending  would seem to be precisely what cities need. However, in a recent study I examined a national program in Brazil, known as FUNDEF, which was initiated in 1997-1998 and which redistributed resources for primary and middle school education. As a result, some municipalities spent extra funds on education and others didn’t. How did they fare?

When I studied the beneficiaries of the program 12 years after its implementation, I found that many municipalities had lost educated people, and those that remained were actually doing worse. The program had worked out well for individuals. A one percent increase in the local education budget was associated with 2.6 percentage points higher likelihood of completing primary school, as well as a greater possibility of being employed and of receiving higher wages. But because many of the beneficiary municipalities didn’t have sufficient jobs for their more educated workers, there was a “brain drain” as people migrated to places of greater economic opportunity. Indeed, every one percent increase in education spending by a municipality, was associated with a 0.8 percentage points increase in the likelihood of leaving the area for somewhere else.

The success of a US-based program

There are, of course, successes and failures in place-based policies. We are still learning about the characteristics that make an intervention more likely to fit into one or the other category. But something that appears to give them a better chance of success is the existence of rules that make mobility more difficult.  A study by Matias Busso and co-authors shows remarkable success for an urban Empowerment Zones program that unfolded in the United States in the 1990s. The initiative, which funneled hundreds of millions of dollars in tax breaks and grants to encourage businesses to locate and hire in impoverished areas of cities like Atlanta, Baltimore and Chicago, could have sparked increased commuting into the recipient areas, undoing some of the program’s positive effects. But the benefits provided by the policy were restricted to individuals who lived inside the zones. That meant that people in other neighborhoods would have had to spend significant resources in relocating there and generally avoided doing so. As a result, employment and wages rose in those areas that benefited from the program, without major changes in population or housing rents.

How policymakers incorporate these lessons can greatly improve the effectiveness of local economic development efforts in Latin America and the Caribbean. What we know is that policies that target places can have unintended consequences when the effects of commuting and migration are not taken into account. But these unintended consequences can be prevented and even turned into opportunities if policymakers ask themselves if people are likely to migrate as a result of their initiatives and how that may affect the desired impact.