Urban Economics Association

5th Meeting of the Urban Economics Association

At the 57th Annual North American Meetings of the Regional Science Association International (RSAI)
November 10-13, 2010
Denver, Colorado

Program Committee: Tom Holmes (Minnesota), Esteban Rossi-Hansberg (Princeton), Kurt Schmidheiny (Pompeu Fabra), and Matthew Turner (Toronto)

Conference Program

01A. Thu 8:00-10:00. Topics in Urban Economics (Pikes Peak)

Chair: Wen-Chi Liao

Tomoya Mori* (Institute of Economic Research, Kyoto University), Tony E. Smith (Department of Systems Engineering, University of Pennsylvania)

Central place regularities across economic regions

Regularities of city systems have typically been analyzed at the national levels. However, the present paper shows that these may not be specific to the national level, but they may exhibit fractal structure across economic subregions. Using Japanese data, we demonstrate that common Central Place Regularities (a version of the Rank-Size Rule and the Number-Average Size Rule) hold across three nesting subregions in Japan (derived from passenger traffic patterns), while these regularities disappear if subregions are randomly assembled.

Discussant: Nathan Schiff

Nathan Schiff* (Sauder School of Business, University of British Columbia)

Exploring Retail Concentration: Restaurant Clusters

This paper examines the location choices of restaurants in Manhattan and documents the prevalence of same cuisine geographic clusters. I use spatial point pattern techniques to describe global location patterns and to define individual clusters. I then set up a simple discrete choice model to predict the number of restaurants of each cuisine in census tracts. I find that many clusters cannot be explained with area observables, suggesting that demand externalities may expand the market catchment area of clustered restaurants. In an appendix I test for a hypothesized negative relationship between price and the number of immediate competitors but find little evidence for increased price competition in clusters.

Discussant: Stephan Heblich

Oliver Falck (Ifo Institute for Economic Research at the University of Munich), Christina Guenther (Max Planck Institute of Economics), Stephan Heblich* (Max Planck Institute of Economics), William R. Kerr (Harvard University and NBER)

From Russia with Love: The Impact of Relocated Firms on Incumbent Survival

We identify the impact of local firm concentration on incumbent performance with a quasi natural experiment. When Germany was divided after World War II, many firms in the machine tool industry fled the Soviet occupied zone to prevent expropriation. We show that the regional location decisions of these firms upon moving to western Germany were driven by non-economic factors and heuristics rather than existing industrial conditions. Relocating firms increased the likelihood of incumbent failure in destination regions, a pattern that differs sharply from new entrants. We further provide evidence that these effects are due to increased competition for local resources.

Discussant: Wen-Chi Liao

Wen-Chi Liao* (National University of Singapore)

Inshoring: the Geographic Fragmentation of Production and Inequality

The advent of information technology facilitates geographic separation of production tasks, which is referred to as offshoring in international contexts and inshoring in domestic contexts. Although the literature on offshoring has flourished, the research on inshoring is limited. This paper makes a contribution, by examining inshoring on both empirical and theoretical fronts. Empirically, it shows that business support services have been increasingly sent to small localities for cost savings and being separated from their downstream industries, which have been consistently concentrated in big cities. Theoretically, it analyzes welfare impact of inshoring. Contrary to a prediction in the offshoring literature, (domestic) support workers can be better off. This is primarily because inshoring allows support workers to benefit from higher urban productivity without bearing urban costs.

Discussant: Tomoya Mori

01B. Thu 8:00-10:00. Housing Economics I (Longs Peak)

Chair: Edward Coulson

Christian Hilber* (London School of Economics), Teemu Lyytikäinen (London School of Economics), Wouter Vermeulen (CPB Netherlands Bureau for Economic Policy Analysis / VU University)

Capitalization of Central Government Grants into Local House Prices: Panel Data Evidence from England

We explore the impact of central government grants on local house prices in England using a panel data set of local authorities from 2001 to 2008. Utilizing strategic political considerations affecting grant allocation as a source of exogenous variation in grants, we find that grants have a sizeable positive and significant causal effect on local house prices. This effect is larger in locations in which housing supply is constrained by physical barriers. In relatively constrained locations grants are roughly fully capitalized into property prices. One implication of full capitalization is that local governments in England do not seem to waste grants. Another implication is that the English grant system may have important unintended distributional consequences.

Discussant: Andrew Hanson

Andrew Hanson* (Georgia State University), Zackary Hawley (Georgia State University)

Do Landlords Discriminate in the Rental Housing Market? Evidence from an Internet Field Experiment in U.S. Cities

This paper tests for racial discrimination in the rental housing market using matched pair audits conducted via e-mail for rental units advertised on-line. We reveal home-seekers’ race to landlords by sending e-mails from names with a high likelihood of association with either whites or African Americans. Generally, discrimination occurs against African American names; however, when the content of the e-mail messages insinuates home-seekers with high social class, discrimination is non-existent. Racial discrimination is more severe in neighborhoods that are near “tipping points” in racial composition, and for units that are part of a larger building.

Discussant: Thomas Davidoff

Thomas Davidoff* (University of British Columbia)

Financing Retirement with Stochastic Mortality and Endogenous Sale of a Home

This paper extends the analysis of lifecycle savings when leaving a home is costly (Artle and Varaiya (1978)) to the case of stochastic mortality. I show necessary and sufficient conditions under which a combined annuity and forward mortgage achieve first best. Reverse mortgages, such as HECM, induce moral hazard on the choice of when to leave the home. When there is asymmetric information, reverse mortgages are more favorably selected than the otherwise first best product on the dimension of mortality, but more adversely on the dimension of taste for remaining in the home.

Discussant: Edward Coulson

Edward Coulson* (Penn State), Lynn Fisher (University of North Carolina)

Structure and Tenure: The roles of free-riding and uncertainty

We document the empirical fact that owner-occupancy declines, and then rises, with the number of units in a structure (unit count). We propose a model of building management. Optimal ownership structure (i.e. landlord vs. condominium associations) depends on the extent to which stochastic contributions to maintenance expense (possibly due to free-riding behavior) pose significant risk to maintenance levels at low levels of unit count. Anecdotal evidence about the choices of condo managers support the implications of the model.

Discussant: Christian Hilber

02A. Thu 13:30-15:30. Trade and Regional Economics (Pikes Peak)

Chair: Jacques-François Thisse

Thomas Holmes* (University of Minnesota), John J. Stevens (Federal Reserve Board of Governors)

Exports, Borders, Distance, and Plant Size

The fact that large manufacturing plants export relatively more than small plants has been at the foundation of much work in the international trade literature. We examine this fact using Census micro data on plant shipments from the Commodity Flow Survey. We show the fact is not entirely an international trade phenomenon; part of it can be accounted for by the effect of distance, distinct from any border effect. Export destinations tend to be further than domestic destinations, and large plants tend to ship further distances even to domestic locations, as compared with small plants. We develop an extension of the Melitz (200model and use it to set up an analysis with model interpretations of ratios between large plant and small plant shipments that can be calculated with the data. We obtain a decomposition of the overall ratio into a term that varies with distance, holding fixed the border, and a term that varies with the border, holding fixed the distance. The distance term accounts for more than half of the overall difference.

Discussant: Hajime Takatsuka

Hajime Takatsuka* (Kagawa University), Dao-Zhi Zeng (Tohoku University), Laixun Zhao (Kobe University)

Globalization and the Resource Curse

This paper examines the relationship between resource endowments and industrialization. Natural resources are extracted and transformed to resource goods, which are used as intermediate goods in the manufacturing sector and/or as final goods for consumption. When trade costs are high, the country with a more valuable natural resource has a more-than-proportionate share of firms and enjoys a higher welfare than the other country. However, when trade costs decrease, firms begin to move out of the country, resulting in a resource curse in terms of welfare as well as firm shares. Further, the resource curse becomes more severe if resources are used in consumption only but not in manufacturing production. The model also explains why economies poor in resources have adopted almost completely free trade policies in recent decades and become successful in developing their manufacturing industries.

Discussant: Keith Head

Keith Head* (University of British Columbia), Ran Jing (University of International Business and Economics), John Ries (University of British Columbia)

Import Sources of Chinese Cities: Order versus Randomness

We utilize very detailed import information for Chinese cities to assess the empirical validity of prominent trade models. We find that key predictions of the Ricardian comparative advantage and a Armington love of variety model are contradicted in the data. We observe that cities within a province often do not purchase a narrowly defined product from the leading foreign source of product in the province. A model of random sourcing describes well the likelihood that a city sources from the top source country in its province. We also find that firm orientation towards particular source countries has a significant impact on the sourcing decisions of cities.

Discussant: Jacques-François Thisse

Evgeny Zhelobodko (Novosibirsk State University), Sergey Kokovin (Novosibirsk State University), Jacques-François Thisse* (CORE-UCLouvain)

Monopolistic Competition: Beyond the CES

We propose a general model of monopolistic competition and derive a complete characterization of the market equilibrium based on an Arrow-Pratt measure of concavity of the utility, interpreted as the relative love for variety. When the relative love for variety increases with the consumption level, the market displays standard competitive effects. On the contrary, when it decreases, the equilibrium price increases with the number of firms and the market size, while the CES is the borderline case. Finally, we apply our setting to trade theory and uncover several new properties hindered by the CES, such as dumping and reverse dumping.

Discussant: Thomas Holmes

02B. Thu 13:30-15:30. The Housing Crisis (Longs Peak)

Chair: Raven Molloy

DeCoster Gregory (Bowdoin College), William Strange* (University of Toronto)

Developers, Herding, and Overbuilding

Recent years have seen the most pronounced turbulence that real estate markets have ever experienced. There have been wild swings in prices, a wave of foreclosures, countless failed investments, and massive overbuilding. This paper will be primarily concerned with overbuilding. Of the many forces that may have combined to produce this situation, the paper will focus on rational overbuilding carried out by developers whose decisions are made under uncertainty. We will establish the possibility of both statistical and reputation-based herding. The former refers to developers learning from each other, and so tending to copy. The latter refers to developers copying each other in order to reduce the probability of a loss of reputation that can result from making an unconventional choice.

Discussant: Paul Calem

Jan Brueckner (UC Irvine), Paul Calem* (Federal Reserve Board), Leonard Nakamura (Federal Reserve Bank of Philadelphia)

Subprime Mortgages and the Housing Bubble

This paper provides theoretical and empirical analysis of the connection between subprime mortgage lending and the recent U.S. housing bubble. The goal is to show theoretically and empirically that an upward shift in expectations regarding house-price appreciation can spur lending to borrowers with poor credit ratings, a consequence of lower concern about mortgage default. The key element of the model is borrower default costs, denoted C, which capture the costs incurred when a borrower defaults on a mortgage (credit impairment and guilt, for example). Recognizing that a low C makes default more likely, the single-period mortgages that are offered to low-C borrowers in the model require a larger repayment amount, B, implying that a higher interest rate is charged (C is assumed to be observable). A payment-to-income constraint puts an upper limit on B. With B and C inversely related, this upper limit in turn implies a minimum level of C, denoted C*, below which potential borrowers cannot get a mortgage. The paper’s main result is that greater expected house-price appreciation leads to a reduction in C*. Thus, the implication is that emergence of a housing bubble should encourage subprime lending, with more low-C borrowers becoming eligible for mortgages. The empirical part of the paper is devoted to testing this prediction, investigating whether counties with higher price appreciation have more subprime lending.

Discussant: Stuart Rosenthal

Stuart Gabriel (UCLA), Stuart Rosenthal* (Syracuse University)

Local Scale Economies in Loan Sales and Access to Mortgage Credit: 1994-2008

Markets for loan sales finance a huge share of lending in the United States but evidence of their impact on borrower access to credit is remarkably limited. In part that is because it is difficult to separate out variation in access to secondary markets from other factors. We address this issue by drawing on spatial variation in credit market activity. We also shed new light on the importance of proximity between borrowers and lenders. Central to our analysis is the idea that locally active secondary markets deepen networks that reduce the transactions cost of selling loans. We examine implications of this and related ideas for U.S. mortgage markets over the 1994 to 2008 period. In doing so we make three sets of contributions. Most important, we demonstrate that improved access to secondary market financing increases borrower access to mortgage credit, and especially so for high-risk market segments. Second, we show that for conforming size loans, proximity to locally active secondary markets has diminished in importance since the 1990s, consistent with the increasing use of information technology in that sector. Third, and in contrast, we show that for the more idiosyncratic jumbo loan sector locally active secondary markets continue to improve borrower access to credit throughout the 2000s. Analogous to small business loans, we argue that risk assessment in the jumbo loan sector may rely on “soft” information that is more easily conveyed among nearby agents for whom a closer relationship is possible, even in the age of the internet.

Discussant: Raven Molloy

Raven Molloy* (Federal Reserve Board of Governors), Hui Shan (Federal Reserve Board of Governors)

The Post-Foreclosure Experience of U.S. Households in the Current Housing Market Downturn

With foreclosures on residential mortgages soaring to historic highs, information about the post-foreclosure experience of former mortgage holders is crucial to our understanding of how the current housing downturn has affected the economy. We propose to take a first look at the post-foreclosure experience using the FRBNY credit-report data - a new quarterly panel dataset based on a large, nationally representative sample of all individuals with credit reports in the United States from 1999 to 2010. We plan to investigate a number of issues including whether post-foreclosure households buy a new house, rent, or move in with other previously-established households, the distance that households move, and the characteristics of the new neighborhood compared to those of the old neighborhood.

Discussant: William Strange

03A. Thu 16:00-18:00. Urban and Labor Economics (Pikes Peak)

Chair: Jason Faberman

Nathaniel Baum-Snow* (Brown Univeristy), Ronni Pavan (University of Rochester)

Inequality and City Size

Between 1969 and 2007 a strong monotonic relationship between wage inequality and city size has developed. In this paper, we investigate the causes of the city size inequality premium and its relationship with the growth in overall wage inequality. We find that one-quarter to one-third of the overall increase in hourly wage inequality in the United States from 1979 to 2007 is explained by city size independent of observable skill. While this influence has occurred throughout the wage distribution, the fraction of the increase in the lower half of the wage distribution explained by city size is at least 50 percent larger than that in the upper half of the wage distribution. More rapid growth in within skill group inequality in larger cities has been by far the most important force driving these city size specific patterns in the data. Differences in the industrial composition of cities of different sizes explain to 32 percent of this city size effect. Our evidence on the evolution of wage inequality in cross-sections of cities improves understanding of the role of urban agglomerations in contributing to the expansion of inequality in marginal products of labor over time.

Discussant: Mark Kutzbach

Mark Kutzbach* (U.S. Census Bureau), Fredrik Andersson (U.S. Controller of the Currency), John Haltiwanger (University of Maryland), Henry Pollakowski (Harvard University), Daniel Weinberg (U.S. Census Bureau)

Job Accessibility and Spatial Mismatch with Longitudinal Employment Data

The paper examines spatial mismatch using confidential longitudinal employment data. Its purpose is to increase our understanding of the implications of spatial barriers to access to low-wage work in U.S. metropolitan areas. A central assumption of the spatial mismatch literature is that accessibility to jobs is crucial for obtaining employment. The present analysis examines the importance and nature of this relationship using confidential Census longitudinal employee-employer linked microdata. In doing so, it addresses self-selection, uses a series of dependent variables based on longitudinal data, and proceeds at a detailed spatial level. In particular, it examines the opportunities available to displaced workers in terms of the spatial distribution of potentially-available jobs by industry and earnings. This research is made possible by the availability of a new confidential Census data source: the Longitudinal Employer-Household Dynamics (LEHD) data. These data provide, for over 1million U.S. workers, confidential detailed information on job location and residential location. They are longitudinal, permitting estimation of models explaining rehiring after a separation. Likewise, measures of job accessibility are defined not only on the spatial distribution of employment but also based on the spatial distribution of new hires. Models are estimated for five “Great Lakes” metropolitan areas. These areas vary in size, changes in manufacturing employment, and opportunities in alternate industry sectors.

Discussant: Henry Overman

Steve Gibbons (London School of Economics), Henry Overman* (London School of Economics), Guilherme Resende (London School of Economics)

Real Earnings Disparities in Britain

This paper estimates housing cost-earnings differentials across labour market areas in Britain. In line with economic theory, we treat these cost-earnings differentials as estimates of the value of amenities in different locations, or 'quality of life'. Wage data comes from an individual level panel covering of the UK workforce allowing us to control for observable and unobservable characteristics of individuals as well as to correct for the effect of non-linearities in tax schedules. When controlling for observables, we show that housing costs rise one for one with the earnings of the average household. However, when we use movers to control for unobservables, we find that sorting leads us to understate the cost effect of living in the highest house price locations. We rank labour markets, according to cost-earnings differentials and explore the factors leading to high 'quality of life'. In common with the existing literature we find that climate and low levels of pollutants (or their sources) appear to be one of the main drivers.

Discussant: Jason Faberman

Jason Faberman* (Federal Reserve Bank of Philadelphia), Matthew Freedman (Cornell University)

The Urban Density Premium and the Distribution of Earnings across Establishments: Evidence on the Role of Entry and Exit

A vast array of research has focused on the earnings premium greater economic density appears to afford to workers in a given location. In this paper, we explore the role of establishment entry, exit, and growth in generating the positive relationship between earnings and density across locations observed in the data. We focus our analysis on how these dynamic processes interact with the earnings-density relationship across the distribution of earnings within a metropolitan area. We motivate our work with recent research by Combes et al. (2009), who explore whether the positive relationship between productivity and city size is driven by agglomeration economies or by firm selection and conclude that is the former. We use a rich panel of establishment micro data for the U.S. Our preliminary findings show that the earnings of both entrants and exits exhibit a positive relationship to urban density. This is true in absolute terms, as well as when their earnings are normalized by the earnings of incumbents. Since both entrants and exits have lower earnings than incumbents do, on average, these results suggest that entry and exit generate a greater compression of the earnings distribution in more dense metropolitan areas.

Discussant: Nathaniel Baum-Snow

03B. Thu 16:00-18:00. Culture, Networks, and Cities (Longs Peak)

Chair: Yves Zenou

Jens Suedekum* (University of Duisburg-Essen), Stephan Heblich (MPI Jena), Oliver Falck (ifo Institute, University of Munich), Alfred Lameli (University of Marburg)

Dialects, Cultural Identity, and Economic Exchange

We evaluate linguistic micro-data from a unique language survey conducted between 1879 and 1888 in about 45,000 German schools. The recorded geography of dialects provides an ideal opportunity to comprehensively portray deep cultural differences at the local level. Dialects have been shaped during centuries in a process of linguistic evolution and reflect domains of cultural identity that would otherwise be immeasurable. We then show in a gravity analysis that cross-regional migration in the period 2000-2006 is positively affected by historical dialect similarity. This points at highly time-persistent cultural borders that impede economic exchange across regions of the same country. Finally, we also investigate the effects of dialect similarity on intra-national trade flows in Germany.

Discussant: Beatrix Brügger

Beatrix Brügger* (University of Lausanne), Rafael Lalive (University of Lausanne), Josef Zweimüller (University of Zurich)

Does Culture Affect Unemployment? Evidence from the Roestigraben

This paper studies the role of culture in shaping unemployment outcomes. The empirical analysis is based on a spatial regression discontinuity design with local comparisons across a language barrier in Switzerland. This 'Roestigraben' separates cultural groups, but neither labor markets nor political jurisdictions. Local contrasts across thel anguage border identify the role of culture for unemployment. Our findings indicate that differences in culture explain differences in unemployment duration on the order of 20%.

Discussant: Marcus Berliant

Marcus Berliant* (Washington University in St. Louis), Masashisa Fujita (Konan University)

Is the `Wonderland of No Spatial Dimension' the Ultimate Goal of Regional Science?: Culture and Diversity in Knowledge Creation

What are the costs and benefits of knowledge diversity? Does the creation of a culture of ideas in common among a population raise or lower productivity? What role is played by interregional interaction among researchers? Can a real technological increase in the cost of collaboration and cost of public knowledge flow between regions increase welfare? In our framework, a culture is a set of ideas held exclusively by residents of a location. In general in our model, the equilibrium path generates separate cultures in different cities. When we compare this to the situation where all workers are resident in one city (or equivalently, when there is no distance discount in the productivity of collaboration), R & D workers become too homogeneous and there is only one culture. As a result, equilibrium productivity is lower relative to the situation when there are multiple cultures and workers are more diverse. In other words, it is better to differentiate workers spatially so that they become more diverse and develop more diverse cultures. This increases productivity relative to the situation where there is only one culture and workers are relatively homogeneous.

Discussant: Yves Zenou

Yves Zenou* (Stockholm University), Helsley Robert W. (University of California, Berkeley)

Social Networks and Interactions in Cities

We develop a model where both the urban space and the social space are explicitly taken into account. Agents are located in a social network. Each agent derives utility from consumption but also from interactions with their direct friends. Each agent optimally decides how many visits to make to the city center given that there is a trade off between the positive externalities they obtain from interacting with other agents in the center and the transportation costs. There are only two locations in the city, the center and the suburbs, and only agents living in the suburbs pay transportation costs to commute to the center. We show that more central agents in the social space tend to live in the city center and interact more with other agents that peripheral agents. We also show that denser networks lead to more aggregate social activities. We finally show that the market equilibrium is not optimal because of social externalities. We determine the value of the social-interaction subsidy that can restore the first best outcome.

Discussant: Jens Suedekum

04A. Fri 8:00-10:00. Theories of Agglomeration (Pikes Peak)

Chair: Wen-Tai Hsu

Nathan Seegert* (University of Michigan)

A Dynamic Model of City Formation

Understanding how cities are formed and grow is imperative to knowing whether there are an efficient amount of cities and whether these cities are efficiently sized. The answers to these questions are essential in assessing urban policy. A dynamic model is used to produce the life-cycle of a city. An important innovation of this dynamic model is that cities are formed by individuals without large agents such as developers. The previous literature on self-organized city formation has been limited by a coordination problem, which this dynamic model has solved. The dynamic model is able to match the empirical facts that almost all cities grow through time, city growth tends be sequential, and cities experience periods or rapid growth. In addition the behavior of cities that are able to limit mobility of migrants is analyzed. A system of cities that are able to limit migration through zoning greenbelts around the city and density within the city are found to be undersized relative to the socially efficient population level. A system of cities that are unable to exclude migrants are found to be oversized relative to the socially efficient population level. However, a system of cities that are able to charge migrants a fee for entering the city are found to be efficiently distributed.

Discussant: Pierre M. Picard

Pierre M. Picard* (University of Luxembourg), Takatoshi Tabuchi (University of Tokyo)

City with Forward and Backward Linkages

This paper considers the spatial structure of a city subject to final demand and input-output linkages. Individuals consume differentiated goods (or services) and firms purchase differentiated inputs (or services) in product (or service) markets where firms compete under monopolistic competition. Workers rent their residential lots in an urban land market and contribute to the production of differentiated goods and inputs. We show that firms and workers co-agglomerate and endogenously form a city. We characterize and discuss the spatial distribution of firms and consumers in such cities on one- and two-dimensional spaces (linear city and planar city). We show that final demand and input-output linkages raise the urban density and reduce the city spread. We finally show that a city is too much dispersed compared to the social optimum.

Discussant: Victor Couture

Victor Couture* (University of Toronto)

Knowledge spillovers in cities: An auction approach

I propose a model of a city in which uncompensated knowledge transfers to entrepreneurs are bids by experts in auctions for jobs. The model derives from the key ideas about how knowledge differs from other inputs of production, namely that knowledge must be possessed for its value to be assessed (Arrow 1962), and that knowledge is freely reproducible. The model identifies conditions under which knowledge 'spills' in non-market interactions, as opposed to being transacted in markets. Endogenous agglomeration economies result from growth in the number of meetings between experts and entrepreneurs, and from heightened competition for jobs between experts.

Discussant: Wen-Tai Hsu

Thomas J. Holmes (University of Minnesota), Wen-Tai Hsu* (Chinese University of Hong Kong), Sanghoon Lee (University of British Columbia)

Plants and Productivity in Regional Agglomeration

This paper presents a model of agglomeration and selection by extending the Eaton-Kortum model of trade, BEJK (Bernard, Eaton, Jensen, and Kortum 200in particular, to incorporate entry of entrepreneurs and labor mobility. The selection mechanism can be decomposed into two stages: (i) Entrepreneurs engages in head-to-head Bertrand competition for the same variety, and the limiting distribution of productivity of those who survive the local competition is Frechet; (ii) trade opens up competition across regions. In a larger market, the mill prices are lower, the average plant size is larger, the selection rate is smaller, but the ex post average productivity is larger. We make several points different from the literature: Selection and agglomeration are intertwined, the variance of ex post productivity distribution can be larger in the larger market, and the selection pressure across firms with different productivity differs in a continuous way and hence does not imply a minimum cutoff for survival.

Discussant: Nathan Seegert

04B. Fri 8:00-10:00. Education and Sorting (Longs Peak)

Chair: Stuart Gabriel

Dennis Epple (Carnegie Mellon University), Maria Marta Ferreyra* (Carnegie Mellon University), Brett Gordon (Columbia University)

Charter School Entry in Market Equilibrium: The Case of Washington, DC

We develop and estimate a structural model of charter school entry. In this model, charter schools offer an alternative to neighborhood public schools. In the model’s first stage, the potential entrant chooses his/her physical location and characteristics based on the perceived opportunities not exploited by conventional public schools. In the second stage, students choose among the available schools. We estimate the model using data for the DC school district. We use our parameter estimates to investigate the potential effects of changes in charter school legislation on charter school entry, and on post-entry outcomes such as enrollment and achievement.

Discussant: David Albouy

David Albouy* (University of Michigan)

Evaluating the Efficiency and Equity of Federal Fiscal Equalization

In theory, federal transfers that make household location decisions efficient will offset differences in federal-tax payments and local tax revenues on capital, but not local tax revenues from residents. Transfers that redistribute resources equitably across regions will likely target areas with individuals of low earnings potential or low real incomes. Examining these metrics in practice, federal transfer differences across Canadian provinces are neither efficient nor equitable, but exacerbate pre-existing inefficiencies and underfund minorities. Total locational inefficiencies cost the economy 0.percent of income annually and cause Atlantic and Prairie provinces to have populations 31 percent beyond their efficient long-run levels.

Discussant: Keren Horn

Keren Horn* (New York University)

Homeowner Investment Decisions and School Quality

This paper will explore the relationship between public elementary school quality, as measured by test scores, and the homeowner’s decision to invest in their home. Previous research has shown that school quality is capitalized into house prices (Black, 1999; Figlio and Lucas, 200and thus homeowners are likely to be sensitive to the quality of the local public school when making the decision of whether or not to invest in their property. Furthermore, based on estimates from the 2003 American Housing Survey (AHS), American households spent approximately $250 billion dollars on home improvements over the course of two years, (Leventis, 2007) compared to $5.6 billion annual investment from CDBG and HOME. Thus homeowner investment decisions play a large role in determining the trajectories of urban neighborhoods. Understanding the interactions between urban schools and urban homeowners can help inform policy decisions surrounding urban revitalization. For this analysis I will primarily rely on data from SABINS, State Departments of Education and the confidential version of the American Housing Survey. This paper will be the first to utilize the internal version of the AHS with the expanded section on housing investments to study the homeowner’s decision as well as the first paper to link the household investment decision directly to school quality.

Discussant: Stuart Gabriel

Yuming Fu (National University of Singapore), Stuart Gabriel* (University of California at Los Angeles)

Migration and Economic Growth in China: The Role of Knowledge and Human Capital Spillovers

We apply a skill-based directional migration model to examine the importance of knowledge and human capital spillovers to China’s economic development. Those spillovers augment private investment in idea discovery and human capital and are central to modern theories of spatial equilibrium and endogenous economic growth (Romer 1986 and 1990, Lucas 1988 and 200and Glaeser and Gottleib 2009). Upon accounting for regional differentials in skill-based compensation, cost-of-living, and natural and other amenities, model estimation indicates that high-skill migrants attach significant importance to knowledge and human capital spillovers arising from FDI and human capital concentration in destination regions. Among low-skill migrants, the importance of spillover effects is considerably damped, reflecting institutional and other barriers to social interactions and human capital investment in destination cities. Findings provide new insights as regards the failure of labor migration to alleviate persistent regional disparities in China’s economic development. Moreover, results suggest important potential welfare gains from removing barriers for low-skill migrants to access spillover benefits.

Discussant: Maria Marta Ferreyra

05A. Fri 10:30-12:30. Theory and Empirics of Agglomeration I (Pikes Peak)

Chair: Esteban Rossi-Hansberg

Kristian Behrens* (University of Quebec at Montreal), Cem Ertur (University of Orleans), Wilfried Koch (University of Quebec at Montreal)

Measuring localization using micro-geographic data: Is it unobserved heterogeneity or externalities?

The aim of our paper is to develop a methodology that allows us to exploit the micro-geographic nature of firm location data while allowing us to disentangle the determinants of localization (unobserved heterogeneity vs interactions). To this end, we blend standard spatial econometric techniques with a Monte Carlo approach using firm-level data for a large sample of Canadian manufacturing firms.

Discussant: Holger Sieg

Holger Sieg* (University of Pennsylvania), Jeff Brinkman (Canegie Mellon University), Daniele Coen-Pirani (University of Pittsburgh)

Estimating a Dynamic Firm Location Model with Agglomeration Externalities

We develop a new dynamic general equilibrium model of firm location choice that can explain the observed sorting of firms by productivity and is consistent with the observed entry, exit, and relocation decisions of firms within an urban economy. We discuss existence of equilibrium of and characterize the stationary distribution of firms in each location. The parameters of the model can be estimated using a nested fixed point algorithm. We implement the estimator using data collect by Dunn and Bradstreet for the Pittsburgh metropolitan area. The data suggest that firms located in the city are older and larger than firms located outside the urban core. As a consequence they use more land and labor in the production process. However, they face higher rental rates for land and office space. As a consequence they operate with a higher employee per land ratio. We find that our model explains these observed features of the data well. Finally, we consider the impact of different relocation policies that provide targeted subsidies to new start-ups and superstar firms.

Discussant: Kurt Schmidheiny

Jan Eeckhout (Universitat Pompeu Fabra), Roberto Pinheiro (University of Pennsylvania), Kurt Schmidheiny* (Universitat Pompeu Fabra)

Spatial Sorting: Why New York, Los Angeles and Detroit attract the greatest minds as well as the unskilled

We propose a theory of skill mobility across cities. It confirms the well documented city size-wage premium: the wage distribution in larger and more productive cities nearly everywhere first-order stochastically dominates that in less productive cities. Yet, because this premium reflects higher house prices, this does not necessarily imply that this stochastic dominance relation also exists in the distribution of skills. Our model entails evidence quite to the contrary: instead of first-order, there is second-order stochastic dominance in the skill distribution. We show that this is due to patterns of sorting across cities in a model with mobility. Our model predicts a 'Sinatra' as well as an 'Eminem' effect: both the very high and the very low skilled disproportionately locate in the biggest cities. Based on our theory, the pattern of spatial sorting can be explained by a simple technology with varying elasticity of substitution by skill. Using CPS data on wages and Census data on house prices, this technology with the elasticity of substitution decreasing in skill density is consistent with the observed patterns of skills.

Discussant: Esteban Rossi-Hansberg

Klaus Desmet (Universidad Carlos III), Esteban Rossi-Hansberg* (Princeton University)

Urban Accounting and Welfare

We study the determinants of the size of cities in the U.S. We first calculate frictions in labor supply and demand decisions as well as productivity in most MSAs in the U.S. using a neoclassical model of a system of cities with labor supply decisions. Using this theory, we can relate productivity, frictions and amenities, to land rents and MSA population size. We then implement this decomposition in the data. The results show that there is a mechanism by which more efficient cities are larger but, as result, also exhibit more frictions. That is, size involves frictions through congestion and other negative effects of agglomeration. Cities that exhibit larger frictions than the ones implied by this size-productivity mechanism are smaller and have lower rents. We show that frictions and productivity are the main determinants of differences in city sizes. Counter-factual exercises allow us to calculate city sizes absent efficiency, excessive frictions, or amenity differences. We show that a large city, like New York, would lose around 9percent of its population if it were to lose its productivity advantage and 12% if it were to lose its amenities, while excessive frictions play a negligible role. In contrast, Bloomington, IN, would increase its size by about 60% if it had average excessive frictions.

Discussant: Kristian Behrens

05B. Fri 10:30-12:30. Housing II (Longs Peak)

Chair: Lu Han

Carlos Garriga* (Federal Reserve Bank of St. Louis), Don Schlagenhauf (Florida State University)

House Prices Boom and Bust: The Case of Spain

This paper describes a quantitative model developed to account for the change in the level of house prices (boom-and-bust cycle) in Spain. The driving forces behind the housing boom are residential investment, immigration, current account de…cits, and the elimination of land regulation. The model emphasizes the interaction of housing supply (determined by the existing stock of residential investment and new construction) with market demand. A calibrated version of the model for the Spanish economy replicates the key aggregate of the economy in 1995. The model predicts that a change in observed fundamentals can rationalize at least 84 percent of the recent boom in the value of housing capital. The model suggests that without large current account de…cits and demographic changes the housing boom could have been much smaller. With respect to the housing bust, the model suggests that the combination of increasing mortgage rates, unemployment, and low productivity can have large e¤ects in the value of housing capital. Some conservative predictions quantify adjustments that range between and 29 percent.

Discussant: Ping Wang

Ping Wang* (Washington University in St. Louis and NBER), Danyang Xie (Hong Kong University of Science and Technology)

Housing Dynamics: Theory Behind Empirics

We construct a two-sector optimal growth model of housing where housing is produced by both land and housing structure/household durables. We explicitly model locational choice. Housing services derive positive utility but are decayed away from the city center. Our model enables a full characterization of the dynamic paths of housing as well as housing and land prices. The model is particularly designed to be calibrated to fit some important stylized facts, including faster growth of housing structure/household durables than housing, faster growth of land prices than housing prices, and a locationally steeper land rent gradient than the housing price gradient. The calibrated model is then used to quantitatively assess the long-run and short-run consequences of changes in preferences and technologies.

Discussant: Gabriel Ehrlich

Gabriel Ehrlich* (University of Michigan), David Albouy (University of Michigan)

Land Values and Housing Productivities across Cities

This paper uses a cost-function approach to estimate parameters from the housing production function using new cross-city data on land values and construction wage-levels. According to our estimates, land accounts for percent of housing costs, labor 64 percent, and mobile capital 11 percent. The estimated elasticity of substitution between factors of production is greater than one, although it is not precisely estimated. Furthermore, we are able to estimate differences in total factor productivity in the housing sector across cities. A one-standard deviation increase in the regulatory environment is associated with an 8 percent increase in housing costs, while a similar increase in undevelopable land is associated with a 7 percent increase. Overall, the most productive housing environments are found in the West Central (North and South) regions, while the least productive are in the Pacific region. Productivity in housing is negatively correlated with measures of productivity in the tradable sector.

Discussant: Lu Han

Lu Han* (University of Toronto), David Genesove (Hebrew University of Jerusalem)

Search and Matching in the Housing Market

Housing markets clear, in part, through the time that buyers and sellers spend on the market. We show that demand generally leads to shorter seller time on the market and fewer homes that buyers visit, while buyer time on the market is much less sensitive to demand. Furthermore, seller time on the market and homes visited are much more sensitive to demand growth than its level, consistent with sellers responding to demand with a lag. Those same findings also provide an estimate of the elasticity of the matching function.

Discussant: Carlos Garriga

06A. Fri 14:45-16:45. Theory and Empirics of Agglomeration II (Pikes Peak)

Chair: Sanghoon Lee

Vernon Henderson* (Brown University), David Weil (Brown University), Adam Storeygard (Brown University)

Measuring Economic Growth from Outer Space

GDP growth is often measured poorly for countries and rarely measured at all for cities. We propose a readily available proxy: satellite data on lights at night. Our statistical framework uses light growth to supplement existing income growth measures. The framework is applied to countries with the lowest quality income data, resulting in estimates of growth that differ substantially from established estimates. We then apply the framework to look sub-national economic growth as related to the new economic geography, examining at what time periods hinterland versus primate/coatal cities in sub-saharan Africa grew more quickly

Discussant: Jeffrey Lin

Jeffrey Lin* (Federal Reserve Bank of Philadelphia), Hoyt Bleakley (University of Chicago Booth School of Business)

Portage: Path dependence and increasing returns in U.S. history

How much do initial conditions and increasing returns matter for the location of economic activity? Us- ing population data from the decennial censuses between 1790 and 2000, we examine portage sites in the U.S. South and Midwest, including those near the fall line, a geomorphologic feature along the east coast of the United States that divides the Piedmont from the coastal plain. In colonial times, the waterborne transport of goods required portage around the falls at these points, while during early industrialization, the falls were a source of water power. These factors attracted commerce and manufacturing. While these original sources of advantage have long since been made obsolete, we find evidence of continuing - and even increasing - importance of these portage sites over time. Our results provide evidence for path dependence in the presence of strong local aggregate increasing returns to scale as a reason for present-day differences across locations in productivity and density.

Discussant: Gilles Duranton

Pierre-Philippe Combes (University of Aix), Gilles Duranton* (University of Toronto), Laurent Gobillon (INED)

The Costs of Agglomeration: Land Prices in French Cities

We exploit a unique French data set to investigate the cross-sectional determinants of land prices with a focus on their elasticity with respect to city size. Even though this elasticity is large at about 0.8, the city size elasticity of urban costs associated with the scarcity of land is small because expenditure on land represents only a tiny share of local incomes. Our data also allow us to estimate a production function for housing.

Discussant: Sanghoon Lee

Sanghoon Lee* (University of British Columbia), Qiang Li (Shanghai University of Economics and Finance)

Uneven Landscape and City Size Distribution

This paper proposes a new explanation for a robust empirical pattern in city size distribution. The key result is that city size distribution converges to log-normal distribution if city size is determined by a product of numerous small random factors. The key message is that we can not use the empirical pattern to test a model of cities. A model tends to focus on one force it aims to highlight. Our result shows that the one force may not generate the empirical pattern, but when we have many of them as in reality, they may generate the pattern.

Discussant: Vernon Henderson

06B. Fri 14:45-16:45. Neighborhoods and Individual Behavior (Longs Peak)

Chair: Jordi Jofre-Monseny

Amanda Ross* (Syracuse University)

Crime, Police, and Truth-in-Sentencing: The Impact of State Sentencing Policy on Local Communities

This paper considers two related questions: the impact of spatial variation in crime prevention policies on the migration of criminal activity into nearby locations and the tendency for higher government anti-crime policies to be partly offset by a scaling back of local crime deterrent efforts. A key source of identification is to draw upon variation in the timing of adoption of state-wide Truth-in-Sentencing (TIS) legislation during the 1990s. The effect on crime and police expenditures are then analyzed using a border methodology. The empirical design compares activity in adjacent counties on opposite sides of state borders in the 59 urban areas that cross state lines. Results indicate that adoption of the stiffer sentencing policy in one state prompts migration of criminal activity into adjacent counties of the neighboring state. Adoption of TIS also tends to lower the overall level of criminal activity in the urban area. These patterns indicate that localities benefit from stiffer state-level sentencing policies in two ways: by reducing the number of crimes committed throughout an urban area and by prompting migration of criminal activity into adjacent counties. Findings further indicate that following imposition of TIS, lower levels of government tend to reduce their expenditures on police protection. This suggests that some of the deterrent effect of higher level government anti-crime policy is offset by a scaling back on anti-crime efforts at the local level.

Discussant: Matthew Freedman

Matthew Freedman* (Cornell University), Emily Owens (Cornell University)

Low-Income Housing Development and Crime

This paper examines the effect of rental housing development subsidized by the government's Low-Income Housing Tax Credit program on local crime. We take advantage of changes in the formula used to determine the eligibility of census tracts for Qualified Census Tract (QCT) status, which affects the size of the tax credits developers receive for building low-income housing. QCT status attracts real estate development from other parts of the county, differentially improving the housing stock in the poorest census tracts. Low-income housing development, and the associated revitalization of neighborhoods, brings with it significant reductions in violent crime that are measurable at the county level. There are no detectable effects on property crime, perhaps because of changes in reporting behavior among residents.

Discussant: Eleonora Patacchini

Eleonora Patacchini* (La Sapienza University of Rome), Yves Zenou (Stockholm University )

Neighborhood Effects and Parental Involvement in the Intergenerational Transmission of Education

We analyze the intergenerational transmission of education focusing on the interplay between family and neighborhood effects. We develop a theoretical model suggesting that both neighborhood quality and parental effort are of importance for the education attained by children. This model proposes a mechanism explaining why and how they are of importance, distinguishing between high- and low-educated parents. We then bring this model to the data using a longitudinal data set in Britain. The available information on social housing in big cities allows us to identify the role of neighborhood in educational outcomes. We find that the better is the quality of the neighborhood, the higher is the parents’ involvement in their children’s education.

Discussant: Jordi Jofre-Monseny

Matz Dahlberg (Uppsala University), Peter Fredriksson (Stockholm University), Jordi Jofre-Monseny* (Universitat de Barcelona)

On the dynamics of segregation

Card et al (2008a) formalize a model of ethnic residential segregation where ethnically mixed neighborhoods are dynamically stable until they reach a threshold (the tipping point). Once the neighborhood has surpassed the tipping point, it will experience massive white flight. These authors propose methods to identify tipping points and, using population counts at the US Census tract level, find that tipping is a salient feature of neighborhood dynamics. The objective of this paper is to use individual register data from Sweden to provide a more complete and informative description of neighborhood tipping behavior. We find that tipping is explained by both increased out-migration and decreased in-migration of whites, although increased out-migration seems to be more important. Tipping seems to be driven by relatively rich individuals and by individuals with kids, suggesting that tipping behavior may increase segregation of whites in a number of dimensions. School grades of white students are lower in neighborhoods that have tipped, indicating that families with kids that do well in school leave neighborhoods that are tipping.

Discussant: Amanda Ross

Fri 17:00-18:30. Plenary in Honor of Jacques Thisse

Chair: John Quigley (UC Berkeley)

Vernon Henderson (Brown University), Yves Zenou (Stockholm University)

Simon Anderson (University of Virginia)

Masahisa Fujita (Kyoto University)

Fri 18:45. Cocktail Party Urban Economics Association (by invitation)

D&F Clock Tower, Floors 17-21
1601 Arapahoe Street, Denver

07A. Sat 8:00-10:00. Topics in Agglomeration (Pikes Peak)

Chair: Elisabet Viladecans Marsal

Clément Bosquet* (GREQAM, Université Aix-Marseille), Pierre-Philippe Combes (GREQAM, Université Aix-Marseille)

Agglomeration Economies in Academic Research

We develop a strategy using individual data to identify agglomeration e ffects in academic research. It allows us to quantify through their impact on publication records the role of economies of urbanisation (size, field diversity and research access), of economies of localisation ( field of specialisation), of the faculty composition (position, age, gender, heterogeneity in talents), of research networks (openness in coauthorship and links to stars) and of the local institutional design (degree of self-recruitment). The application relates to an exhaustive data set of French economists over 1992-2008.

Discussant: Tatsuhito Kono

Tatsuhito Kono* (Tohoku University), Yuuki Mitsutani (Tohoku University), Kirti Kusum Joshi (Tohoku University)

Clustering of Stores in Center and Suburb with Consumers’ Imperfect Information

Consumers with imperfect information about the quality of goods tend to search multiple stores for comparative shopping. On the other hand, stores cluster together to benefit from the large pool of consumers, and consumers also benefit from such store agglomeration because their searching costs are reduced. However, existence of travel cost motivates multiple clusters to locate away from each other in order to monopolize over local consumers. In a spatial competition model with trade-off between clustering and dispersing forces, this paper analyzes how relative travel costs between center and suburb clusters affect (store density and (2) equilibrium price in each cluster as well as (3) consumers’ welfare.

Discussant: Megha Mukim

Megha Mukim* (London School of Economics)

Industry and the urge to cluster: a study of the informal sector in India

This paper studies the determinants of firm location choice at the district-level in India to gauge the relative importance of agglomeration economies vis-a-vis good business environment. A peculiar characteristic of the Indian economy is that the unorganised non-farm sector accounts for 43.of Net Domestic Product and employs around 71.6% of the total workforce. I analyse National Sample Survey data on the unorganised sector that covers over 4.4 million firms, in both sectors - manufacturing and services. The empirical analysis is carried out using count models, and I instrument with land revenue institutions to deal with possible endogeneity bias. As a robustness check I use data on over 140,000 manufacturing firms from the Annual Survey of Industries, and 20,000 manufacturing and services firms from the Prowess database to study the organised sector. I find that buyer-suppler linkages and industrial diversity make a district more attractive to economic activity, whilst the quality and level of infrastructure are also important. I conclude that public policy may be limited in its ability to encourage relocation of informal firms.

Discussant: Elisabet Viladecans Marsal

Raquel Marín López (Universitat de Barcelona & Institut d’Economia de Barcelona), Jordi Jofre Monseny (Universitat de Barcelona & Institut d’Economia de Barcelona), Elisabet Viladecans Marsal* (Universitat de Barcelona & Institut d’Economia de Barcelona)

On the mechanisms of agglomeration: Evidence from the location of new manufacturing firms in Spain

The objective of this paper is to quantify the relative importance of the three Marshall’s agglomeration theories by examining the location of new manufacturing firms in Spain. Recent literature has considered that agglomeration is not only explained by intra-industry relationships, but inter-industry relationships could be as well decisive. The relationships among industry pairs provide very rich information that can be used to identify the relevance of different agglomeration theories (Ellison et al., 2007). Thus, our aim is to identify the tendency of different industries to be co-located in an area. We approach Marshall’s mechanisms in terms of new manufacturing firms choosing locations in environments with the presence of: industries with which they share similar workers in terms of their skills (labor market pooling); 2) industries with which they have a supplier/costumer relationship (input sharing); and 3) industries with which they share technologies (knowledge spillovers). To capture these relationships among industries, a very disaggregated industry level is needed. We work with 75 three-digit manufacturing industries of the National Classification of Economic Activities.

Discussant: Clément Bosquet

07B. Sat 8:00-10:00. Empirics of Fiscal Policy (Longs Peak)

Chair: Teemu Lyytikäinen

Raphael Parchet* (University of Lausanne), Beatrix Brügger (University of Lausanne)

Culture and Taxes: Towards Identifying Tax Competition

We propose a new strategy for testing the existence of interjurisdictional tax competition and for estimating its spatial reach. Our strategy rests on differences between desired tax levels, determined by culture-specific preferences, and equilibrium tax levels, determined by interjurisdictional fiscal externalities as well as by preferences. Exploiting the fact that fiscal preferences differ systematically and demonstrably between French speaking and German speaking Swiss regions, and using propensity score matching, we find that local income tax burdens do not change discretely at the language border but exhibit smooth spatial gradients. The slope of these gradients implies that tax competition constrains tax choices of the jurisdictions with a preference for higher taxes at a distance of up to kilometres. This suggests that tax competition does shape income taxation of local governments, but only on a relatively small spatial scale.

Discussant: Gerald Carlino

Gerald Carlino* (Federal Reserve Bank of Philadelphia), Robert Inman (Wharton School, University of Pennsylvania)

States Fiscal Policies and the Economic Crises: Will the ARRA State Stimulus Stimulate the Economy?

The American Recovery and Reinvestment Act of 2009 (ARRA) is an important policy response to the current deep recession. State and local governments have suffered significant declines in their revenues and, for states in particular, a significant increase in the demand for services, particularly Medicaid funding, unemployment compensation, and welfare spending. ARRA will provide almost $150 billion in fiscal relief to state and local governments with the majority of aid provided in fiscal years 2009, 2010, and 201The hope is that this relief will encourage state spending and discourage state tax increases, thereby adding to national aggregate demand. This paper uses the U.S. experience of prior changes in federal aid to state and local governments to examine the likely consequences of this stimulus package for state budgets and FOR aggregate GDP. In an SVAR analysis we explicitly incorporate federal transfers to the state and local sector as part of federal fiscal policies. We find the response of GDP to federal aid shocks to be relatively small over the first two years after the receipt of aid (a one dollar shock to per capita aid raises real per capita GDP on impact by about $0.06, rising to $0.27 after one year and to $0.49 after two years). From a panel analysis of state budgetary behavior, we find that a potential reason for the weak response of GDP to aid is the propensity of state governments to “save” a disproportionate share (almost 60 percent) of their new assistance. They do so by replenishing their rainy day funds and by paying down their long-term debt positions. Such behaviors are potentially rational responses to debt-financed temporary assistance.

Discussant: Ricardo Politi

Ricardo Politi* (Sao Paulo School of Economics - Getulio Vargas Foundation), Enlinson Mattos (Sao Paulo School of Economics - Getulio Vargas Foundation)

Tax base choice and tax mimicking: evidence of ad-valorem tax interaction in Brazilian Basic Basket Food.

This paper investigates tax base choice of sub national governments in Brazil taking into account the correspondent decision made by their spatial neighbors. In particular, we estimate the determinants of tax rate exemptions and reductions, and tax basis choice, allowing for the presence of spatial dependence in ad-valorem tax in Brazilian states for goods which compound the states Basic Basket Food from 1992 to 2007. Our results suggest limited evidence of horizontal tax rate interaction. However, we do find strong evidence of tax base choice spatial interaction, i.e., the presence of one good in one neighboring state Basic Basket Food has a statistical significant and positive effect in the original state choice. These results seem to reinforce the positive model proposed by Hettich and Winer (1984) and Inman (1989), which suggest that political competition is multidimensional and can arise in a broader variety of forms. Our estimations recommend that such spatial interaction can occur through tax base choice, and tax deductions and exemptions for Brazilian states.

Discussant: Teemu Lyytikäinen

Teemu Lyytikäinen* (London School of Economics)

Tax Competition among Local Governments: Evidence from a Property Tax Reform in Finland

This paper uses a Finnish policy change to study tax competition among local governments. Changes in the statutory lower limits to the property tax rates are used as a source of exogenous variation to estimate the causal effect of tax rates in neighbouring municipalities on the tax rate choices of municipalities. I do not find evidence on strategic interaction in property tax rate choices among Finnish municipalities. The results are in contrast with the earlier literature that has mainly found positive interdependence in tax rates among neighbouring jurisdictions. I compare the causal estimates based on the policy change to spatial lag estimates and spatial instrumental variables estimates that are commonly used in the literature. In contrast to the estimates based on the policy change, the spatial lag and spatial IV models give positive and significant estimates for the neighbourhood effect in tax rate choices. This suggests that the spatial lag and spatial IV methods may have a tendency to overestimate the degree of spatial interdependence in tax rates.

Discussant: Raphael Parchet

08A. Sat 10:30-12:30. Empirical Urban Economics (Pikes Peak)

Chair: Carlo Menon

Jeanne Tschopp* (University of Lausanne)

Does Occupational Composition Matter? Evidence from Germany.

How do average wages respond to exogeneous industry-specific shocks that shift labour demand across industries? We extend the search and bargaining model of Beaudry, Green and Sand (2009) to the occupational dimension and show that this consideration is crucial. Omitting occupational composition leads to biased estimates of the adjustment of average wages to industry-specific shocks if the latter do not occur in isolation of occupation-specific shocks. Using German individual-level data for 1977-200we estimate that the bias resulting from the omission of occupational composition is substantial and statistically significant. Results for a specification that includes occupational composition suggest that the average wage adjustment to industry-specific shocks is twice as large as that predicted with industrial composition only.

Discussant: Gabriel M Ahlfeldt

Gabriel M Ahlfeldt* (LSE), Georgios Kavetsos (Cass Business School)

Form or Function? The Impact of New Soccer Stadia on Property Prices in London

This paper focuses on the channels through which stadium externalities capitalize into property prices. We investigate two of the largest stadium investment projects of the recent decade – the New Wembley and the Emirates stadium in London, UK. Evidence suggests positive stadium externalities, which are large compared to construction costs. Notable anticipation effects are found immediately following the announcement of the final stadium plans. Our results emphasize the role stadium architecture plays in promoting positive spillovers to the neighbourhood. We therefore recommend public funding of large-scale sports facilities to be made conditional on a comprehensive urban design strategy that maximizes the external benefits.

Discussant: David Cuberes

David Cuberes* (University of Alicante)

Internet Usage and City Size

This paper uses a new dataset on Internet flows and Internet-related infrastructure connections between American cities to explore the effect of the spread of this technology on city size. I first explore whether larger cities use the Internet more intensively, instrumenting current city population with historical city size. In the second part of the paper I study the reverse relationship i.e. whether variations in Internet usage lead to larger agglomerations of population. Following the strategy of Andersen et al. (2010) I instrument Internet usage with the intensity of lightning activity. Due to data availability, this second exercise needs to be aggregated at the state level. Finally, I discuss my estimated elasticities to infer whether the Internet and face-to face-contacts are indeed complements or substitutes.

Discussant: Carlo Menon

Carlo Menon* (Bank of Italy)

Stars and Comets: an Exploration of the Patent Universe

The patent distribution across inventors is extremely skewed, as many inventors - the comets - register one or a few patents, while a small number of inventors - the stars - register many patents. Using a rich database on US inventors, we provide evidence suggesting that the two categories of patents are localized in different places. We then test whether the activity of stars is beneficial for local comets, finding that 10% increase in the number of star patents leads to increase in the number of comet patents.

Discussant: Jeanne Tschopp

08B. Sat 10:30-12:30. Transportation and Cities (Longs Peak)

Chair: Jeffrey Zax

Adam Storeygard* (Brown University)

Farther on down the road: transport costs, trade and urban growth in sub-Saharan Africa

In this paper, I seek to determine the effect of changing transport costs on economic output in a sample of over a thousand hinterland cities with and without manufacturing in sub-Saharan African countries. Specifically, I consider whether periphery cities with lower transport costs to their country's primate city grew faster than those further away or with poorer road connections, in the context of dramatically rising oil prices between 2003 and 2008. This is an empirical implementation of a new economic geography (NEG) model in a geographic context where they have been rarely used, using newly available data on transport networks and proxies for city economic activity.

Discussant: Rosa Sanchis-Guarner

Rosa Sanchis-Guarner* (London School of Economics & SERC), Steve Gibbons (London School of Economics & SERC), Teemu Lyytikäinen (London School of Economics & SERC), Henry Overman (London School of Economics & SERC)

Productivity and employment impacts of agglomeration: evidence from transport improvements

This paper estimates the effect of road transport infrastructure improvements on firm productivity and employment. The study uses micro longitudinal datasets on firms and employees in Britain, linked by detailed geographical location to road transport improvements that occurred between 1998 and 200We measure to which extent new road infrastructure projects changed employment accessibility (or 'effective density') at locations close to the sites of the projects. We then estimate whether firms in locations that experienced large changes of this type showed productivity improvements relative to those that experienced smaller changes. We do not find evidence that improved accessibility increased total factor productivity within firms, or that firms increased employment at the plant level due after the improvements. However, we find evidence that places (postcode sectors) that experience accessibility improvements close to transport schemes attract more firms and employment, and weaker evidence of productivity improvements at this more aggregated level.

Discussant: Jeffrey Zax

Jeffrey Zax* (University of Colorado at Boulder), Ryan Hall ()

The effects of congestion on the propensities to telecommute and to commute off-peak

Congestion is a costly problem in many metropolitan areas. Many public policy responses are difficult to adopt and implement because they require substantial investments in infrastructure and extended disruptions to existing activities. However, individuals may also have the capacity to reduce the personal costs which they suffer from congestion by altering their travel patterns. For example, commuting costs can be altered by changing either workplace or residence locations. Even without such relocations, commuting costs may be reduced by either working from home or by commuting at times when congestion is less. This paper estimates the effect of metropolitan-level congestion on these two commuting margins. It matches individual-level data regarding telecommuting and time of departure for work from the 2000 Census with metropolitan-level measures of historical average commuting time and congestion from the 1990 Census and from the Texas Transportation Institute to estimate the effect of historical congestion on the propensities to work at home and to depart for work at off-peak times. The results demonstrate that telecommuting and off-peak departure times are significantly more likely in metropolitan areas with higher historical congestion levels. Moreover, the elasticities of these responses to congestion levels are substantive in magnitude. They suggest that policies which encourage or facilitate these responses may have noteworthy effects on aggregate congestion.

Discussant: Adam Storeygard

09A. Sat 13:30-15:30. Structure of Cities I (Pikes Peak)

Chair: Efthymia Kyriakopoulou

Wouter Vermeulen* (CPB Netherlands Bureau for Economic Policy Analysis)

Growth controls, agglomeration externalities and welfare in a system of heterogeneous cities

Cities that impose severe land use constraints effectively limit the size of their workforce and the level of economic activity, so that they may not reap the full benefits of agglomeration. How large are these indirect costs in comparison to the value of preserved open space and the foregone consumer surplus of housing that would have been created on it? This question is analysed in an applied general equilibrium model for a system of heterogeneous cities, which is calibrated to the strongly regulated Dutch Randstad area. We find that the introduction of a Marshallian production externality alters the second-best number of cities and their size distribution dramatically, relative to a constant returns to scale scenario. However, a land use policy that ignores its indirect impact on productivity yields almost the same social surplus, provided that it truly reflects the external value of open space.

Discussant: Alex Anas

Alex Anas* (State University of New York at Buffalo)

Metropolitan Decentralization and the Stability of Commuting Time

RELU-TRAN, a computable general equilibrium model of the Chicago MSA with endogenous road congestion, shows that as the metropolitan area grows according to the regional forecasts from 2000 to 2030, the interdependent suburbanization of jobs and residences without increases in highway capacity would increase suburban land development and sprawl, but not increase the per-worker travel time in commuting. The reason is that congestion per mile traveled would increase on average as more jobs and residences suburbanized, but the average distance traveled between job and residence would decrease as the decentralization of firms and workers adjusted to the higher congestion per mile. On the one hand, these results provide an explanation of empirical observations made by others in the last twenty years. On the other hand, they provide the first empirical test of the predictions of theoretical models in which the locations of jobs and residences are interdependent and simultaneously determined. The results show that suburban development, contrary to popular expectations, is likely to remain sustainable in the face of additional metropolitan decentralization in the future.

Discussant: Efthymia Kyriakopoulou

Efthymia Kyriakopoulou* (Athens University of Economics and Business), Anastasios Xepapadeas (Athens University of Economics and Business)

On the Structure of Cities: Emergence of Residential and Industrial Areas under Environmental Policy

We study the internal structure of a city, in a model of economic geography, where industry and housing compete for scarce land to locate. We analyze a spatial model of a city in which a single good is produced using land, labor, machinery and emissions of a pollutant, and in which people consume goods, choose residential land and dislike pollution. The agglomeration effects, caused by trade-offs between centripetal and centrifugal forces, in the form of pollution diffusing in space, stringency of environmental policy, transportation and commuting costs determine the emergence of industrial and residential clusters across space.

Discussant: Wouter Vermeulen

09B. Sat 13:30-15:30. Neighborhood Effects (Longs Peak)

Chair: Yong Suk Lee

Stephen Gibbons* (London School of Economics), Olmo Silva (London School of Economics), Felix Weinhardt (London School of Economics)

Do neighbours affect teenage achievements? Evidence from residential change

The importance of quality of peers in the neighbourhood on individual behaviour and educational achievements is widely upheld by sociologists and educational researchers. Economists instead tend to be more sceptical and attribute apparent correlations between individual outcomes and peer group/neighbourhood characteristics to residential sorting based on personal preferences and characteristics. To investigate these issues, we make use of information contained in pupil-level and school-level administrative data matched to detailed geographical information on pupils’ residence and geographical mobility, as well as other administrative data sources. We use this combined set of information to analyse to effect of quality of peers in the neighbourhood - and more generally quality of the neighbourhood - on individual progress through secondary education. To measure the quality of one student’s neighbourhood, we construct several aggregate indicators based on geographical areas that group a handful of postcodes surrounding pupil’s place of residence. These include aggregate information about the educational quality of the peers in the neighbourhood - such as average attainments - as well as proxies for the overall quality of the neighbourhood - such as the incidence of unemployment. Different strategies are used to account for the fact that place of residence is determined by unobservable family characteristics and preferences about school quality and neighbourhood amenity. All in all, our results show little evidence of sizeable and significant neighbourhood effects on young people’s educational attainments.

Discussant: Stephen L Ross

Stephen L Ross* (University of Connecticut), Jason Fletcher (Yale University)

Estimating the Effects of Friendship Networks on Health Behaviors of Adolescents

Researchers typically examine peer effects by defining the peer group broadly (all classmates, schoolmates, neighbors) because of the lack of friendship information in many data sources as well as to enable the use of plausibly exogenous variation in peer group composition across cohorts in the same school. This paper estimates the effects of friend’s health behaviors on own health behaviors for adolescents. A causal effect of friend’s health behaviors is identified by comparing similar individuals who have the same friendship opportunities because they attend the same school and make the same friendship choices, under the assumption that the friendship choice reveals information about an individual’s unobservables. We combine this identification strategy with a cross-cohort, within school design so that the model is identified based on across grade differences in the clustering of health behaviors within specific friendship options. This strategy allows us to separate the effect of friends behavior on own behavior from the effect of friends observables attributes on behavior, a key aspect of the reflection problem. Our results suggest that friendship network effects are important in determining both adolescent tobacco and alcohol use.

Discussant: Gregory Verdugo

Gregory Verdugo* (Banque de France)

Public Housing Magnets: Public Housing Supply and Immigrants' Location Choices

In continental Europe, the participation rate of non-European immigrants in public housing programs is particularly large with respect to natives. This paper investigates the influence of public housing on the initial location choices of immigrants across cities and regions. To identify this effect, I use the variations provided by the massive increase in the public housing supply in France between 1968 and 1990, which was negatively related to the distribution of immigrants across cities. To deal with the issue of reverse causality, I use past construction decisions and the variations of political divisions across cities to instrumentthe evolution of the public housing stock per capita. Results indicate that increasing a city’s public housing supply has a large ’magnetic effect’ on non-European immigrants, but no effect on European immigrants. Cities with high ethnic community public housing participation rates are attractive to non-European immigrants while high participation rates deter European immigrants.

Discussant: Yong Suk Lee

Yong Suk Lee* (Brown University)

Tracking vs Mixing in Secondary Education: Implications on Location Sorting and Intergenerational Mobility in Korea

I examine whether the regime shift from an exam based high school entrance policy (“tracking”) to a district based lottery allocation mechanism (“mixing”) affects location sorting and intergenerational education mobility. If ability and income are not randomly distributed over space and quality of education varies across location then the mechanism that allocates students to schools will likely have a direct impact on intergenerational mobility. A simple model predicts that, under mixing, the gradient on household income relative to one’s ability becomes larger in the achievement production function due to households sorting across districts. I empirically test these hypotheses utilizing data from Korea where the transition from tracking to mixing occurred during the 1970s. Using a nationally representative sample, I find that the difference in difference and difference in difference in difference estimates on income relative to ability increases with mixing. This effect is strongest in Seoul and the provincial capitals and the coefficient increases steadily until it peaks 4 to 5 years after the regime change to mixing. These results are consistent with the hypothesis of location sorting. I directly test for location sorting within Seoul utilizing a border discontinuity design on the change of land prices as well as district level panel data on population and tax revenues. The movement of top-tier high schools in Seoul during this period helps the identification. I find a discrete jump in the change in housing land prices in the top school district. I also find that the ratio of school aged children and real estate tax revenue increases in districts that receive an additional top-tier high school.

Discussant: Stephen Gibbons

10A. Sat 16:00-18:00. Structure of Cities II (Pikes Peak)

Chair: Jan Rouwendal

Hesham Abdel-Rahman* (University of New Orleans)

Specialization, Diversification, and Mixed System Cities with Multiple Types of Workers

When do cities specialize and when do they diversify in production? When do a specialized and a diversified city coexist? What is the impact of skill distribution on the formation of different types of cities with different industrial compositions? Does income disparity vary based on the type of cities in an urban system? The objective of this paper is to analyze these questions in a two-sector, spatial, general-equilibrium model of a closed economy. The economy is populated with a continuum of unskilled workers with heterogeneous potential ability. Workers can acquire either specialized or generalized training. Workers with generalized training and specialized training are used to produce two goods with the use of specialized and generalized technologies. Knowledge spillover among diverse skilled workers in cities results in higher productivity due to localization or urbanization economies. Within this framework, we examine the impact of different types of training and technological externalities on the distribution of workers in a system of cities as well as on the city’s industrial composition. The paper characterizes the condition under which various equilibrium configurations will emerge. Furthermore, comparative static results are presented for each equilibrium configuration.

Discussant: Dao-Zhi Zeng

Toshiaki Takahashi (The University of Tokyo), Hajime Takatsuka (Kagawa University), Dao-Zhi Zeng* (Tohoku University)

A Minimalist Model of the Home Market Effect

The concept of home market effect (HME) aims to examine firm agglomeration when differentiated goods are produced under increasing returns to scale and transported costly. This paper proposes a minimalist model observing the HME in the absence of the agricultural sector. The model includes one industry only with mobile capital and immobile labor. Typical results observed in the existing literature can be seen in this simple model. Particularly, the wage in the larger country is higher, which is equivalent to the HME in our model, and both the wage ratio and the industrial location pattern evolve in (inverted) U shape with the decreasing transport costs. Finally, despite the HME, we show that the trade integration might be beneficial to both countries in the sense of welfare.

Discussant: Chia-Ming Yu

Chia-Ming Yu* (Washington University in St. Louis), Marcus Berliant (Washington University in St. Louis)

Creative Capital and the New Geography of Classes

The creative class refers to workers who add economic value through their creativity. The growth of creative class forms the foundation for the creative economy, which also requires a pool of low-end service workers to fill their daily needs. Depending on the creative workers' sensitivity to environmental openness and the marginal utilities on housing, equilibrium distributions of creative and service workers can be different. When creative workers are not sensitive to environmental openness, in the unique stable equilibrium creative and service workers are evenly distributed; When creative workers are sensitive enough to environmental openness, the symmetric equilibrium is not stable and there are two stable equilibria each of which has a unbalanced class structure. This paper illustrates how creative capital is formed and explains both the inequalities across and within U.S. cities.

Discussant: Jan Rouwendal

Jan Rouwendal* (VU University), Hans Koster (VU University)

Mixed Urban Land Use

Urban areas are characterised by dispersed employment patterns and mixed land use. Lucas and Rossi Hansberg (LRH) [Econometrica 70 (2002), 1445-1476] developed a model where the urban spatial structure is determined by the external benefits of agglomeration and the longer commutes for workers. This paper reviews and tests the main implications of the LRH-model using semiparametric regression techniques. We show that in mixed urban areas, agglomeration is an important determinant of the land rent, while in residential areas proximity to a business area impacts the rent.

Discussant: Hesham Abdel-Rahman

* Presenter