Report

Teaser, summary, work performed and final results

Periodic Reporting for period 2 - DYNURBAN (Urban dynamics: learning from integrated models and big data)

Teaser

As economies grow and develop, the number and population sizes of cities naturally increase. However, the tight link between urbanisation and economic growth also runs in the opposite direction, with cities and urbanisation being a primary engine of economic growth. After all...

Summary

As economies grow and develop, the number and population sizes of cities naturally increase. However, the tight link between urbanisation and economic growth also runs in the opposite direction, with cities and urbanisation being a primary engine of economic growth. After all, the learning and human capital spillovers occurring in cities are fundamental to the creation of new ideas and the entrepreneurship which underpin economic growth. Despite widespread interest, isolating the aggregate implications of the number and population sizes of cities on economic growth and aggregate income has proved elusive. As part of this project, we propose a new model of how cities and urbanisation interact with aggregate income and economic growth. Our model relies on strong microfoundations to represent individual cities, matches key empirical regularities at the city and economy-wide levels, and generates novel predictions for which we provide evidence. Most importantly, this model is amenable to a quantification that relies a small number of parameters and remains transparent regarding the mechanisms at work. We directly estimate important parameters, which determine the magnitude of urban benefits and costs. This then allows us to assess quantitatively the effect of cities and urbanisation on economic growth and aggregate income and examine a variety of counterfactuals.

One of the key decisions individuals make in cities is the choice of neighbourhood where to live. We spend about two-thirds of our time at home and around one-third of our income buying or renting it. Depending on our residential location choice, there are also substantial differences in the extent to which jobs, education opportunities and amenities are within reach and in who we can interact with. As circumstances change, so do our residential location choices, and in many countries 5% or more of the population moves each year.

A key advantage of cities driving location decisions is that they offer more valuable experience and opportunities to workers, the more so the bigger the city. Since these benefits are greater for more able workers and given that housing costs in bigger cities are higher for everyone regardless of ability, one might expect that when workers move, more able ones are more likely to go to a big city. And yet, this is not the case. We believe that flawed self-assessment is partly to blame. When young migrants choose a location, even if they consider the heterogenous rewards of bigger cities depending on ability, they may be fooled by a very imperfect assessment of their own ability. We develop a model urban sorting by workers with heterogenous self-confidence and ability and test its key predictions on panel data from the National Longitudinal Survey of Youth 1979, which allows us to track individuals’ location and labour market activities and also contains individual measures of ability and self-confidence. Consistent with our theory, it is not the most able young workers who choose to locate in bigger cities, but instead the ones with most self-confidence. For more experienced workers, ability plays a stronger role in determining location choices, but the lasting impact of earlier choices dampens their incentives to move.

The sorting of workers across cities is not only driven by job opportunities and housing costs. They also care about proximity to family and friends. We show that gathering information on a person’s network of friends and family helps understand the heterogenous neighbourhood choices of apparently similar individuals much better. We use information about changes in individuals’ neighbourhood of residence and about each individual’s social network, derived from anonymised cellphone Call Detail Records for the universe of customers of a Swiss telecommunications operator, in combination with demographic and location attributes. Our analysis shows that taking into account where each person’s contacts live doubles our abi

Work performed

Three research papers have been completed or are close to completion for submission to a journal.

“City of dreams” (joint with Jorge De la Roca and Gianmarco Ottaviano) was submitted to a journal, a first round of revisions requested by the journal was completed and the paper resubmitted, and a second round of revisions will be completed by July 2019. Bigger cities offer more valuable experience and opportunities in exchange for higher housing costs. While higher-ability workers benefit more from bigger cities, they are not more likely to move to one. Our model of urban sorting by workers with heterogenous self-confidence and ability suggests flawed self-assessment is partly to blame. Analysis of NLSY79 data shows that, consistent with our theory, young workers with high self-confidence are more likely to initially lo- cate in a big city. For more experienced workers, ability plays a stronger role in determining location choices, but the lasting impact of earlier choices dampens their incentives to move.

“Calling from the outside: The role of networks in residential mobility” (joint with Konstantin Büchel, Maximilian V. Ehrlich, and Elisabet Viladecans-Marsal) was submitted to a journal May 2019. Using anonymised cellphone data, we study the role of social networks in residential mobility decisions. Individuals with few local contacts are more likely to change residence. Movers strongly prefer places with more of their contacts close-by. Contacts matter because proximity to them is itself valuable and increases the enjoyment of attractive locations. They also provide hard-to-find local information and reduce frictions, especially in home-search. Local contacts who left recently or are more central are particularly influential. As people age, proximity to family gains importance relative to friends.

“Urban growth and its aggregate implications” (joint with Gilles Duranton) will likely be submitted to a journal in July 2019. We develop an urban growth model where human capital spillovers foster entrepreneurship and learning in heterogenous cities. Incumbent residents limit city expansion through land-use regulations so that commuting and housing costs do not outweigh productivity gains. The model builds on strong microfoundations, matches key regularities at the city and economy-wide levels, and generates novel predictions for which we provide evidence. It can be quantified relying on few parameters, provides a basis to estimate the main ones, and remains transparent regarding its mechanisms. We then examine a variety of counterfactuals to assess quantitatively the effect of cities on economic growth and aggregate income.

Substantial data preparation has already been performed for two other papers. One seeks to understand the dynamics of urban land use and involves combining gridded data of land use, population, businesses and roads for 3 decennial periods. The other paper aims to understand why work experience is substantially more valuable when it is acquired in bigger cities by looking at mobility across firms, tasks and position in the organisational hierarchy.

Final results

Despite widespread interest, isolating the aggregate implications of the number and population sizes of cities on economic growth and aggregate income has proved elusive. Given the general equilibrium nature of the problem, the micro-mechanisms that generate a productivity and innovation advantage of cities and the empirical estimates we have for them do not immediately map into aggregate implications. An alternative approach would be to estimate the contribution of cities to aggregate outcomes directly from aggregate data. Unfortunately, attempts such as regressing the rate of output growth of countries on characteristics of their cities, while suggestive, have fallen into the usual pitfalls of cross-country regressions.

In this project, we propose a new model of how cities and urbanisation interact with aggregate income and economic growth. Our model relies on strong microfoundations to represent individual cities, matches key empirical regularities at the city and economy-wide levels, and generates novel predictions for which we provide evidence. Most importantly, this model is amenable to a quantification that relies a small number of parameters and remains transparent regarding the mechanisms at work. We directly estimate important parameters, which determine the magnitude of urban benefits and costs. This then allows us to assess quantitatively the effect of cities and urbanisation on economic growth and aggregate income and examine a variety of counterfactuals. Let us develop these points in more detail.

Consistent with suggestions from the empirical literature, we model the agglomeration benefits of cities as arising from human capital spillovers. These spillovers foster entrepreneurship which, in turn, leads to higher city productivity. As cities grow in population, they facilitate learning and further human capital accumulation, magnifying economic growth. City population growth brings disadvantages as well as advantages, although urban economists have paid less attention to characterising and quantifying the costs of larger cities than their benefits. Our modelling of urban costs considers several components that vary in strength within and across cities, as well as over time. Transportation costs inside cities have been found to be an important determinant of urban growth. Within each city, as emphasised by the standard monocentric city model, more central locations feature better accessibility in exchange for more expensive homes. In practice, however, not everyone works centrally, so transport requirements do not increase in proportion with distance to the city centre. Our modelling takes this into account, as well as the fact that congestion causes transport costs to increase with the number of travellers. Across cities, the cost of housing with a given accessibility also increases significantly with city population. These elements affect the relationship between urban costs and city population in the cross section of cities and also the long-term evolution of the urban system. With both benefits and costs to city size, our model incorporates the “fundamental trade-off” of urban economics. To resolve this trade-off, models of urban systems often rely on city developers to deliver the socially optimal number and population sizes of cities. However, the equivalence between what is delivered by city developers, local governments, and a social planner breaks down once we allow for heterogeneity across cities. Taking this into consideration, we propose a political economy mechanism where endogenously-determined land-use regulations balance the greater commuting and housing costs associated with larger cities against agglomeration benefits. ‘Incumbent residents’ of more productive cities regulate land use to limit entry into their city, thereby maximising their own welfare at the expense of potential newcomers.

Our modelling of city formation through a local political process is intuitively appealing and

Website & more info

More info: https://diegopuga.org.