Urban Economics Association

Urban Economics

1st Meeting of the Urban Economics Association

53th Annual North American Meetings of the Regional Science Association International (RSAI)
November 16-18, 2006
Toronto, Canada

Program Committee:

Accepted Papers

Session 01. Land Use and Spatial Interactions

Alejandro Jofre* (Center for Mathematical Modeling & DIM, Universidad de Chile), Pedro Jara (DELTA, and College de France, Paris.), Francisco Martinez (Dept. Civil Engineering, Transportation. Universidad de Chile)

A Land Use Equilibrium Model With endogenuos incomes

In a new land use equilibrium model based on a goods-and-housing exchange process with private ownership and hence endogenous incomes, we define two possible market equilibrium conditions. In the first one, households bid for all the available dwellings and market equilibrium adjusts utilities and allocates households in order to satisfy that at each available location only best bidders are localized. In the second one, a competitive equilibrium is defined; where households choose their dwellings by maximizing their utility, given the prices of goods and dwellings. In the second part of this paper we give conditions that allow us to identify equilibrium points that satisfy the equilibrium conditions of both approaches. The particular structure of the model motivates a new definition for willingness to pay (or bid) functions obtained as a result of an optimization problem and extending the usual definition (inverse on the rent of indirect utility) to any possible constraint in the consumer’s consumption sets. Finally, in the third part of the presentation, we establish an existence result for a simplified version of the utility maximization equilibrium model.

Pascal Mossay* (Universidad de Alicante), Picard Pierre (University of Manchester)

A Spatial Model of Social Interactions

A major source of spatial heterogeneity stems from the existence of social interactions. Social interactions through face-to-face contacts are at the essence of our societies and explain the gathering of individuals in villages, agglomerations or cities. They translate a psychological need for maintaining relationships with one another, and favour a constant exchange of ideas (Krugman 1991; Glaeser and Scheinkman 2003; Fujita and Thisse, 2002). In this paper we address the issue of the emergence of multiple agglomerations as the result of the interplay between social interactions and competition in the land market. The present paper builds on Beckmann's (1976) model. This model provides a simple rationale for the spatial agglomeration of agents as the result of spatial interaction externalities. Agents are distributed along a geographical space and benefit from social interactions with other agents. These social interactions provide individual benefits while entailing an individual cost as each one must access to distant agents. Moreover the return of the spatial interaction is also balanced by a cost of residence since agents compete for land space. When the benefit of social interactions is larger than the commuting and residence costs, agents prefer to be located close together, which leads to the formation of agglomerations. Beckmann considered the case of a one-dimensional spatial economy modelled along a line segment. The resulting equilibrium consists in a unimodal symmetric - bell-shaped - spatial distribution, where agents agglomerate around the city center and progressively disperse away from the city center (Fujita and Thisse, 2002). We revisit Beckmann (1976)'s framework along a line segment and extend it to the case of spatial economy extending along a circle. While the modelling along a line segment seems appropriate to describe the internal structure of cities, the formulation along a circumference provides a natural framework to analyze the emergence of multiple agglomerations. Circular spatial frameworks have already been studied in `racetrack economy' models. Those are new economic geography models based on economic interactions between agents that reside along the perimeter of a circumference, e.g. Fujita et al. (1999), Mossay (2003), or Picard and Tabuchi (2003). Yet, because of the complexity of market interactions, this strand of literature characterizes only the `flat earth' spatial equilibrium corresponding to a spatially constant distribution of agents. The present social interaction model has simpler analytical properties and allows us to discuss the structure of all spatial equilibria. The only change with respect to Beckmann's model concerns the utility function. We consider an hyperbolic preference for residential space rather than the logarithmic preference. Our results are the following. First we determine the first best and the equilibrium spatial distributions of agents along the line segment. In accordance to Fujita and Thisse (2002), the first best distribution is more concentrated than the equilibrium's; see also Tabuchi (1986). At equilibrium agents choose a too large lot size because they do not internalize other agents' preferences when making their own lot choice. Moreover, we show that the equilibrium includes a single city irrespective of the residential space preference. In this case, social interactions generate heterogeneity within a city but no heterogeneity among cities since the city is unique. Furthermore we also show that the implementation of the first best spatial distribution does not require the use of spatial transfers. Second we provide the conditions characterizing the spatial equilibria for a spatial economy extending along a circle. We consider the case of equidistant, symmetric, non-connected cities meaning cities sharing the same spatial structure and being separated by equal-size empty hinterlands. We show that there exists no spatial equilibrium with an even number of such cities. In contrast, there exist multiple equilibria with odd numbers of such cities. In that case the equilibria can be Pareto-ranked. Our findings contrast with Beckmann's result. Along a circle, spatial equilibria involving multiple agglomerations emerge. Indeed, the periodicity of the geographical space adds new location possibilities for agents because they may interact with others through additional alternative paths. Such issues arise in other spatial interaction models formulated around a circumference as in Krugman et al. (1999), Mossay (2003), Picard and Tabuchi (2003) and others. The relative simplicity of our model allows us to deal analytically with these aspects.

Christian Redfearn* (University of Southern California)

Determinacy in Urban Form: Fixed Investment and Path Dependence in the Spatial Distribution of Metropolitan Employment

Currently the economics of agglomeration receives a great deal of research attention, focusing on a variety of externalities to explain the evolution of cities. Much of this research is ahistorical, with little attention paid to the cumulative history of investment decisions that are manifested in the urban form researchers seek to understand. This paper addresses path dependence in the location of economic activity within the Los Angeles metropolitan area and presents evidence that during a remarkably dynamic period – in terms of the growth and transformation of the region’s population and employment, as well as other fundamental modeling variables such as transportation costs and communication technology – the “shape” (the spatial allocation of economic activity) of the Los Angeles metropolitan area remained broadly unchanged. Over the twenty-year sample period, the number of employment centers and their share of total employment have undergone only small changes relative to their aggregate dynamics. This stability appears to have its origins in the large fixed investment in structures and transit networks made decades earlier. Indeed, where employment concentrations are not situated astride one of the arteries in the current highway network (largely established by 1960), their location can be attributed to the freeway system as it stood prior to WWII. The dynamics of the metropolitan area’s employment appear to be explained in no small part by a path dependence related largely to the region’s freeway system as it existed prior to 1960. This finding has important implications for our understanding of location choice and the causes of agglomeration within metropolitan areas.

David Frame* (Carnegie Mellon University)

Local public goods and the dynamics of urban decay

Urban decay can be thought of as the physical and social degeneration of parts of a city, or even entire cities. Structural economic change resulting in a loss of local population, property abandonment, crime, and other social problems. Why is the urban structure in and around cities such as Detroit, Pittsburgh, and Cleveland abandoned despite proximity to city centers? Other cities, such as Denver, for example, have made significant transformations in the past 10 to 15 years. Why have some city centers flourished, while others experience persistent decay? Local governments and the level of public service provided by cities are often identified as important components of decision making among different neighborhoods and/or jurisdictions within metropolitan areas. What policies undertaken by local governments encourage and speed urban decay? What policies can be undertaken to stop or even reverse it? What is the role played by developers at the fringe of existing cities? The model developed in this paper attempts to capture the dynamics of urban decay and renewal where existing urban areas, run by local governments, compete with private developers offering newer housing allow with public amenities.

Session 02. Land Supply and Local Policies

Yusuke Teraji* (Graduate School of Economics, Kyoto University)

An Economic Analysis of Municipal Consolidation with Heterogeneous Preference

This paper studies the conditions for the consolidation of political jurisdictions. We focus on the tradeoff between the scale merit and the heterogeneity of preference among residents. That is, large jurisdictions enjoy the benefit of scale but must incur the cost due to the heterogeneity among their populations. In addition to this, the land public ownership yields the strategic interaction: the agreement of the consolidation of some towns induces the consolidation of remaining towns. Our model implies that i) the consolidation might be too much observed under the decentralized regime; ii) observed number of the consolidation might be inefficiently large.

Shin-Kun Peng (Academia Sinica and National Taiwan University), Ping Wang* (Washington University in St. Louis and NBER)

Housing Quality, Housing Development and Public Policy

We construct a general equilibrium model to characterize housing quality and prices in a competitive spatial equilibrium framework in which the finance of local public amenities plays a crucial role. We consider an array of housing-related tax policies, including a developer revenue tax, a property tax, a land tax and a development license fee. In competitive spatial equilibrium, all households optimize and reach the same utility in locational equilibrium, all monopolistically competitive developers optimize and receive zero profit, and both housing and land markets clear. We characterize the schedule of housing quality, housing prices, land rent, and the population and housing density. Moreover, we evaluate the effects of tax policies on the housing market, the urban structure, as well as the welfare of the local economy.

John Hartwick* (Queen's University)

The Control of Land Rent in a Fortified Farming Village

We consider village defense as a local public good and take up the matter of funding this good with local land rent from farms, rent administered by a local civic leader. Rotating office-holding is the democratic alternative. We argue for this administrator becoming an exploitative local herediary lord of the community and consider constraints on his degree of predation on the incomes of the farmers. We draw on the Henry George theory of using land rent to finance the public sector and take up a new critique of Henry George theory.

Francois Ortalo-Magne* (University of Wisconsin - Madison), Andrea Prat (London School of Economics)

The Political Economy of Housing Supply

We model an economy where people choose housing consumption, housing investment, and vote at a local level on the number of new building permits to be issued. Inefficiencies arise due to a vicious circle between households' desire to own housing as a hedge against income shocks and their support for housing supply restrictions that make housing a better hedge against income shocks. The equilibrium supply of housing depends on the rule that governs the appropriation of the proceeds from issuing new building permits. There is hysteresis in housing supply: the more restricted the initial housing supply, the smaller the city size selected by the voting process. We analyze the effects of a number of policies: (1) A reform of the housing permit allocation process; (2) Incentives for home ownership; (3) Centralization of planning decisions; and (4) Elimination of barriers to fractional ownership. The model highlights the tension between the two housing policy objectives of increasing homeownership and housing affordability.

Session 03. Housing

Christian A. L. Hilber* (London School of Economics), Yingchun Liu (Fannie Mae)

Explaining the Black-White Homeownership Gap

Blacks in the United States are considerably less likely to own their homes compared to Whites. Differences in household income and other socio-economic characteristics can only partially explain this gap and previous studies suggest that the 'unexplained gap' has increased over time. In this paper we use the 2001 Panel Study of Income Dynamics (PSID) intergenerational data, which allows us to better control for household wealth, parental wealth (and other parental externalities) and location type (which proxies the relative cost of owning). We find that the remaining homeownership gap can be entirely explained by differences in endowments and location choice.

Paul Cheshire* (London School of Economics), Christian Hilber (London School of Economics)

Office space supply restrictions in Britain: the political economy of market revenge

According to surveys by real estate intermediaries rents and occupations costs of office space in Britain are - and have been since comparable data have been available - the highest in the world (KingSturge, 2003; 2004 & 2005: CBRE 2004 & 2005). While London is a large and prosperous city with an international reach as a financial centre and location for HQs so might be expected to be expensive, it is not immediately obvious why the total occupation costs of office space per square metre in a not so prosperous, medium sized city such as Birmingham, England, should be 44 percent higher than Manhattan or 124 percent higher than richer and really land-constrained Singapore (KingSturge, 2004). Yet this is consistent with research on the effects of planning constraints in the residential sector showing that these are high and have had a significant net negative welfare effect in the prosperous south east of England (Cheshire and Sheppard, 2002). This paper first investigates the causes of high costs of office space in Britain. We provide the first systematic estimates of the value of the ‘Regulatory Tax’ (see Glaeser, Gyourko and Saks, 2005) for non-residential real estate. We have estimated this for office space in 15 British office locations. We provide a central estimate and then subject this to sensitivity tests. This work shows that even on the most conservative assumptions there is a very substantial cost of regulation in Britain. The regulatory tax relative to the marginal costs of construction is several orders of magnitude greater than that estimated for Manhattan condominiums by Glaeser et al (2005). The values are estimated over time with the oldest data being for the City and West End of London in 1961. This cross sectional time series data then allows us to investigate the political economy of regulatory restriction on office development. The British tax system provides no fiscal incentive to local authorities to permit commercial development since there is a uniform business rate with rates being set nationally and revenues redistributed according to nationally determined formulae. The result is that in all cases except one, permissions are controlled by ‘homevoters’ (Fischel, 2001). The exception is the City of London where the inheritance from the medieval system of government which allowed local determination to the City Corporation, coupled with the fact that there are only 4,000 residents in the City has meant that local business interests control the land regulatory system. The City of London is also the only office location in Britain where the value of the Regulatory tax has fallen over time. This fall seems to be related to explicit changes in planning restrictiveness in the City of London, triggered first by the development of competing back office locations in the 1960s and 1970s and then by the development of London Docklands during the 1980s. This provided significant competition to the City as an office location. In all other locations the value of the regulatory tax has increased significantly since the early 1990s. Finally, we construct alternative measures of regulatory restrictiveness for each of the fifteen locations and show how these relate to estimates of the regulatory tax. We then investigate how our measures of the regulatory tax relate over time and across locations to local economic indicators, home ownership rates and varying conservation requirements imposed by national government.

Nathaniel Baum-Snow* (Brown University), Justin Marion (UC Santa Cruz)

The Effects of Low Income Housing Units on Neighborhoods

This paper exploits a discontinuity in the allocation of tax credits to low income housing developments that generates pseudo-random assignment in the number of low income housing units built in similar sets of census tracts. Preliminary estimates indicate that a 30 percent increase in the tax credit is associated with an increase of approximately 5 low income housing units on a base increase of 12 units per tract on average. Among large metropolitan areas, we find a measurable negative impact of these additional low income units on local housing values.

Morris Davis* (University of Wisconsin), Morris Davis (Georgetown University), Jonathan Heathcote ()

The Price and Quantity of Land in the United States

A house is a bundle comprising a physical structure and the plot of land upon which the house is built. Thus changes in house prices reflect changes in the cost of structures and value of land. In this paper we apply this insight to construct the first constant-quality price and quantity indexes for the aggregate stock of residential land in the United States. We document that the value of residential land exceeds annual GDP, and that the dynamics for the prices of residential land and residential structures are quite different. For example, the real price index for residential land almost tripled between 1975 and 2005, while the real price of structures increased by only 24 percent. Fluctuations in house prices at business cycle frequencies, including the recent boom, are primarily driven by changes in the price of land.

Session 04. Real Estate

David Brasington* (Louisiana State University), Diane Hite (Auburn University)

A Mixed Index Approach to Identifying Hedonic Price Models

Recent literature suggests identifying house price hedonic models by using instrumental variables, spatial statistics, the borders approach, panel data, and other techniques. We introduce a mixed index model to identify house price hedonic regressions. We compare the performance of the mixed index model to a traditional hedonic model and to a hedonic model that includes characteristics of the buyer of each house. We find the mixed index model outperforms the other models based on bootstrap distributions of predicted housing values, prediction variance, and predicted policy effects. The mixed index model distributions are less skewed and kurtotic than the other models, suggesting that the mixed index model more closely satisfies the classical linear regression assumption of normally distributed errors. Compared to the mixed index model, the traditional hedonic overstates the importance of school quality to house price and understates the importance of environmental quality to house price.

Indrashis Chaudhuri* (Department of Economics, The Ohio State University)

Consumer Borrowing Behavior of U.S. Homeowners

The paper is aimed at developing a clear understanding of consumer borrowing behavior of U.S. homeowners experiencing house price shocks with a special focus on differences in behavior comparing non-Hispanic Whites and African-Americans. Some prominent economists attributed the increase in home equity and subsequent refinancing of mortgages to be the stabilizing force during the last recession. The theoretical model of my paper is aimed at analyzing the circumstances under which increased equity in homes results in increased consumption by homeowners. It is shown that liquidity constrained households and households that smooth consumption over time are the likely candidates to react strongly to any housing wealth windfall. In addition, collateralized borrowing should be the instrument of choice for cashing out equity from homes and spending on consumer goods. Given that on average African-Americans are more likely to have a smaller asset base and more volatile income than non-Hispanic Whites, the theoretical model suggests that Blacks should react strongly to any house price shock, as they are more likely to be financially constrained. The instrumental variable regression results point to the fact that the consumer credit behavior of homeowners differ by race in various consumer credit markets. Another finding is that Blacks are not behaving in line with theoretical predictions. The paper also found some evidence that Blacks refinance primarily to cash out equity where as Whites usually achieve the dual objectives of lowering interest rate and cashing out equity from refinancing.

Paul Anglin* (University of Guelph)

Local dynamics and contagion in real estate markets

Prices usually assumed to adjust so that the market clears in such a way that any dynamic characteristics of a market are independent across different segments within a market. In this paper, I propose a model with three locations: one in the center and two on the periphery. In each market separately, the price is determined by a combination of auto-correlated shocks and non-instantaneous adjustment. Prices are also determined by the potential for buyers to move between locations based on information. Thus, the evolution of the three prices is governed by a simple difference equation characterized by three parameters: the degree of autocorrelation, the speed of adjustment and a measure of contagion between regions. While local parameters may be hidden by aggregation, I argue that changes in the micro-structure can alter macro-phenomena: that contagion between regions alters the dynamic properties of aggregate measures. Most of the analysis uses numerical simulation to quantify the effects. The paper ends with a discussion of empirical issues since the contagion parameter cannot be estimated directly using commonly-available city-wide data on average prices. This analysis will also use data drawn from Toronto Canada data for over 80 regions and 10 years. I present evidence of significant correlation over time and over space.

John McDonald* (University of Illinois at Chicago), Yuliya Yurova (UIC)

Property Taxation and Selling Prices of Industrial Real Estate

This study shows that variations in the local property tax are fully capitalized into the selling prices of industrial real estate in the O'Hare Airport market area. The property tax is 1.67% in DuPage County and 4.36% in Cook County. Both counties abut O'Hare Airport.

Session 05. Agglomeration and its Effects

Thomas Klier (Federal Reserve Bank of Chicago), Daniel McMillen* (University of Illinois at Chicago)

Clustering of Auto Supplier Plants in the U.S.: GMM Spatial Logit for Large Samples

A linearized version of Pinkse and Slade’s (1998) spatial probit estimator is used to account for the tendency of auto supplier plants to cluster together. By reducing estimation to two steps – standard probit or logit followed by two-stage least squares – linearization produces a model that can be estimated using large datasets. Our results imply significant clustering among older plants. Supplier plants are more likely to be in counties that are near assembly plants, that include interstate highways, and that are near other counties with supplier plants. New plants show no additional tendency toward clustering beyond that shown by older plants.

Gerald Carlino* (Federal Reserve Bank of Philadelphia), Robert Hunt (Federal Reserve Bank of Philadelphia)


This paper extends the research in Carlino, Chaterjee and Hunt (2004) to consider variations in the effects of scale (urbanization economies) and industrial clustering (localization economies) on industry level rates of innovation, as measured by patents. We estimate separate equations for more than a dozen U.S. industries over the 1990s. We find both effects matter, but the magnitudes vary significantly across industries. We also verify that our earlier results in terms of the selectivity of worker matches, as measured by density of workers, obtains in most R&D intensive industries.

Li Gan (Texas A&M University), Qinghua Zhang* (Peking University)

The Thick Market Effect on Housing Markets Transactions

This paper provides a search model for housing market where the number of buyers and/or sellers plays very important role. The model makes three testable predictions: (1) the unemployment rate has a negative impact on the trading volume and the sale prices of the housing market; (2) a larger housing market has a lower average sale price, shorter time-to-sale and smaller price dispersion, in addition to a lower vacancy rate. (3) In a larger housing market, when the unemployment rate goes up (or down), the sale price decreases (or increases) by a smaller percentage than in a smaller market. All three predictions are supported by a panel dataset of the Texas city-level housing markets.

Fabiano Schivardi* (Banca d'Italia), Luigi Guiso (Universita' Tor Vergata)

What determines entrepreneurial clusters?

We contrast two potential explanations of the substantial differences in entrepreneurial activity observed across geographical areas: entry costs and external effects. We extend the Lucas model of entrepreneurship to allow for heterogeneous entry costs and for externalities that shift the distribution of entrepreneurial talents. We show that these assumptions have opposite predictions on the relation between entrepreneurial activity and productivity: with different entry costs, in areas with more entrepreneurs their average productivity should be lower and vice versa. We test these implications on a sample of Italian firms and unambiguously reject the entry costs explanation in favor of the externalities one. We also investigate the sources of external effects, finding robust evidence that learning externalities are an important determinant of cross-sectional differences in entrepreneurial activity.

Session 06. Transportation

Pierre M Picard* (CORE UCL; SoSS University of Manchester), Thierry Brechet (CORE UCL)

Airports and Tradeable Noise Permits

This paper presents a market design for the management of noise disturbance created by aircraft traffic around large airports. A market for tradable noise permits allows noise generators to compensate harmed residents. We show that the noise permit markets allow to achieve the planner's optimal allocation of flights provided that she/he does not over-weight the benefit of economic activity compared to the disutility of noise disturbances. The fact that zones are likely to be strategic players does not fundamentally alter this finding. Because of market auctionneer's information constraints, noise permits are likely to redistribute windfall gains to residents located in non-critical zones. This entices landlords to increase their land/house rents there and to design smaller houses in the long run.

Se-il Mun* (Kyoto University), Shintaro Nakagawa (Kyoto University)

Cross-border Transport Infrastructure and Aid Policies

We investigated resource allocation concerning the provision of cross-border transport infrastructure, which is used for trade of goods between two neighboring countries. Since the level of infrastructure is sub-optimal under the circumstances that two governments choose the levels of infrastructure independently, we focus on the role of foreign aid to improve the efficiency of infrastructure provision. In this paper, we examine the welfare effects of aid policies, and show that aid can make both countries better off, i.e., Pareto improvement. Furthermore, Pareto improvement is more likely if the stage of development in recipient country is very low or sufficiently high.

Kurt Van Dender* (University of California, Irvine)

Determinants of Fares and Operating Revenues at U.S. Airports

We estimate a system of equations to investigate the dependence of (a) average passenger fares for flights out of large U.S. airports, (b) airports’ average aeronautical revenues per flight, and (c) concession revenues per passenger, on the airports’ market environment and on service quality. The main indicators of market structure are whether there is a nearby airport offering similar services (horizontal relations), the degree of airline concentration at the airport, and network hub status (vertical relations). Service quality is measured by delays, which are partly the result of congestion. The system is estimated on annual data for large U.S. airports from 1998 through 2002. We find that aeronautical and concession revenues are only weakly related to indicators of market structure. However, passengers and airlines respond to fares, airport charges and delays as would be expected, and fares depend on vertical and horizontal features of market structure as well. The results imply that the anticipated increase in commercialization of airports is likely to strongly alter their decisions concerning aeronautical and concession charges and concessions; they also provide some guidance on airport and passenger behavior affecting such changes.

Marvin Kraus* (Boston College)

Returns to Scale in Networks

When demand increases in a congestible network, there are typically a variety of ways that a network authority can provide additional capacity. For example, in a highway network, existing roads can be widened or new ones added. This paper is concerned with the determination of the degree of local returns to scale in the cost function for the network’s outputs when there are multiple margins – including the network’s density – along which the network authority can make adjustments to capacity. The degree of local returns to scale is important, bearing on whether and to what extent an optimally priced and designed network should be subsidized. We prove: Theorem. Under the provision of a cost-minimizing network, the degree of local returns to scale is the same along all margins for adjusting capacity, both singly and in combination. This includes the network’s density. A particular implication of the theorem is that, subject to the optimality of the initial network, its density can be held fixed in evaluating local returns to scale.

Session 07. Honouring M. Fujita

Tony Smith* (University of Pennsylvania), Tomoya Mori (Kyoto University)


In this paper we give a brief overview of our recent attempts to develop an analytical framework for studying empirical patterns of spatial economic agglomeration. These efforts have been largely inspired by the theoretical work of Fujita and Krugman, who formulated a general equilibrium model of a one-dimensional continuum within which equilibrium patterns of land use could be derived. Unlike previous two-region models or non-spatial multi-region models, their continuous framework for the first time allowed one to study the equilibrium size and spacing of industrial agglomerations, as well as their possible coordination across different industries. Our work attempts to build on these theoretical contributions by developing analytical methods for studying these spatial agglomeration properties empirically. One key finding of the Fujita-Krugman model was that in many cases the equilibrium industrial patterns derived are strikingly similar to the classical Hierarchy Principle of Christaller. Our work with Japanese data lends empirical support to this finding. In particular, we have found a strong empirical regularity, designated as the “Number-Average Size Rule”, that is not only consistent with Christaller Hierarchies, but also reveals a new connection between such hierarchies and the classical Rank-Size Rule. In addition, we have developed a new statistical approach to the comparative analysis of agglomeration (localization) between industries in terms of Kullback-Leibler divergence. The decomposability properties of K-L divergence are particularly well suited to the study of agglomeration hierarchies at various level of spatial aggregation. Another key finding of Fujita-Krugman is the notion of an “agglomeration shadow”, which helps to explain the equilibrium spacing between agglomerations within a given industry. Our most recent work on agglomeration patterns can in principle help to measure this shadow effect. Here we begin with a statistical model of agglomeration patterns in terms of “cluster schemes”, and then develop methods for constructing and identifying the cluster schemes that best fit the regional concentrations of each industry. These “best fit” patterns not only yield empirical information about the size and spacing of agglomerations within industries, but also about their degree of coordination across industries. So it is our hope that the results of this analysis may ultimately suggest useful theoretical hypotheses for two-dimensional extensions of the Fujita-Krugman model.

Vernon Henderson* (brown university & NBER), Hyoung-Gun Wang (World Bank)

Urbanization and City Growth: the Role of Institutions

. This paper models the urbanization process and how urbanization in a country is accommodated by increases in numbers versus population sizes of cities, in an endogenous growth context where political institutions play a key role. The paper estimates the equations of the model describing growth in city numbers in a country and growth in individual city sizes, using a worldwide data set on all metro areas over 100,000 from 1960-2000. Institutions and the degree of democratization and fiscal decentralization, as well as technological advances, strongly affect growth in both city numbers and individual city sizes, with the effects on individual city sizes being heterogeneous. Technology improvements help bigger cities, with their complex infrastructure needs, relative to smaller ones; but increasing democratization levels the playing field across the urban hierarchy, allowing smaller cities greater ability to compete for firms and residents. In sum, these two opposing effects on the relative sizes of bigger versus smaller cities appear to have left the overall relative size distribution of cities worldwide unchanged over the time period.

Edward Glaeser* (Harvard University)

Why Have Housing Prices Gone Up?

Since 1950, housing prices have risen regularly by almost two percent per year. Between 1950 and 1970, this increase reflects rising housing quality and construction costs. Since 1970, this increase reflects the increasing difficulty of obtaining regulatory approval for building new homes. In this paper, we present a simple model of regulatory approval that suggests a number of explanations for this change including changing judicial tastes, decreasing ability to bribe regulators, rising incomes and greater tastes for amenities, and improvements in the ability of homeowners to organize and influence local decisions. Our preliminary evidence suggests that there was a significant increase in the ability of local residents to block new projects and a change of cities from urban growth machines to homeowners’ cooperatives.

Session 08. Honouring M. Fujita: Empirical New Economic Geography

Tomoya Mori* (Kyoto University), Tony E. Smith (University of Pennsylvania)

A Probabilistic Modeling Approach to the Detection of Industrial Agglomerations

Dating from the seminal work of Ellison and Glaeser in 1997, a wealth of evidence for the ubiquity of industrial agglomerations has been published. However, most of these results are based on analyses of a single (scalar) index of agglomeration. Hence it is not surprising that industries deemed to be similar by such indices can often exhibit very different patterns of agglomeration -- both with respect to the number, size, and spatial extent of individual agglomeration clusters. The purpose of this paper is thus to propose a more detailed spatial analysis of agglomeration in terms of multiple-cluster patterns. The key idea is to develop a simple probability model of multiple clusters, called cluster schemes, and then to seek a "best" cluster scheme for each industry by employing a range of standard model-selection criteria. Our ulitimate objective is to provide a richer characterization of spatial agglomeration patterns that will allow more meaningful comparisons of these patterns across industries.

Keith Head* (University of British Columbia, Sauder School of Business), Thierry Mayer (University of Paris I)

Decolonization and the erosion of market potential

We investigate the costs of independence to former colonies in terms of lost market potential. The idea is that trade linkages fell between the former colony and colonial power. Because the colonial power's market could not be easily replicated, the former colony experienced a net loss in market potential. New Economic Geography models predict a corresponding decrease in the incomes of factors in the former colonies. We investigate whether this mechanism finds support in the data.

Gilles Duranton* (University of Toronto), Henry Overman (London School of Economics)

Detailed Location Patterns of UK Manufacturing Industries

Using a point-pattern methodology, we explore a range of issues regarding the detailed location patterns of uk manufacturing industries. In particular, we focus on the location of entrants and exiters vs. continuing establishments, domestic- vs. foreign-owned, large vs. small, and affiliated vs. independent. We also examine co-localisation between vertically-linked industries. Our analysis provides a set of new stylised facts and confirmation for others.

Anthony Briant (Paris-jourdan Sciences-Economiques), Pierre-Philippe Combes* (GREQAM-University of Aix-Marseille, Paris-jourdan Sciences-Economiques and CEPR), Miren Lafourcade (University of Valenciennes and Paris-jourdan Sciences-Economiques)

Does the Modifiable Areal Unit Problem jeopardize linebreak economic geography estimations?

Any empirical research based on geographical data has to deal first with the choice of a given spatial nomenclature. When results differ from one study to another that does not use the same nomenclature, it might be the case that changes in the results are due to changes in the nomenclature, or to real differences in the underlying process under study. For instance, the determinants of local wages have been studied at various geographical scales such as countries (Redding and Venables, 2005), European regions (Head and Mayer, 2006), national regions (Brackman et al., 2004 ; Mion, 2004), or even smaller units such as US counties (Hanson, 2005), French (Combes et al., 2004) or Italian (Mion and Naticchioni, 2005) employment areas. Typically, the significance and weight of wage determinants vary across these studies. A huge body of empirical literature was also devoted to analyze the agglomeration scope of economic activities. In the same vein as before, it is difficult to infer from the results obtained under different geographical scales whether theoretical predictions can be challenged independently of the spatial scale chosen for empirical analysis. More generally, if different areal arrangements of the same data produce different numerical results, independently of subjacent theory, testing the validity of theoretical predictions becomes itself an extraordinary difficult task. This paper investigates whether some standard economic geography estimations are jeopardized by changes in the scale and aggregation of spatial units. We notably compare the biases arising from nomenclature choices with those resulting from economic mis-specification.

Session 09. Honouring M. Fujita: Theoretical New Economic Geography

Marcus Berliant* (Washington University in St. Louis), Fan-chin Kung (City University of Hong Kong)

Can Information Asymmetry Cause Agglomeration?

Various models, such as those used in the New Economic Geography, employ combinations of agglomerative and repulsive forces to generate equilibria with cities and agglomeration. Can classical asymmetric information in the labor market, in the form of adverse selection, result in an equilibrium that features agglomeration of agents? We use a model with two types, high and low ability, and two locations. The high type dislikes work more than the low type. Firms in both locations have the same technology for production of a single consumption commodity. They know the distribution of types, but the type of a particular agent is private information to that agent. The firms compete with both potential entrants and firms in the other location. Firms offer labor contracts that specify wages based on hours worked. In equilibrium, zero profit, voluntary participation, and incentive compatibility constraints must be satisfied along with feasibility. A further stability requirement is imposed, that the equilibrium be immune to small locational deviations of consumers. We have functional forms and some relatively mild restrictions on parameters such that the equilibrium separates types by location. Thus, high and low skilled workers agglomerate separately. This can be induced as a comparative static change from a symmetric equilibrium to an asymmetric one by varying some of the exogenous parameters.

Takatoshi Tabuchi* (University of Tokyo), Chao-cheng Mai (Tamkang University), Shin-Kun Peng (Academia Sinica)

Economic Geography with Tariff Competition

A simple two-country model of economic geography is constructed in order to examine the effect of tariff competition on the manufacturing distribution of firms. We show that tariff competition with free capital movements ends up with a coreperiphery economy, where at least one of the two countries imposes no tariff in Nash equilibrium. We also show that tariff competition without capital movements harms each other’s welfare, and that tariff competition with capital movements leads to the first-best outcome.

Massimo Del Gatto (U Cagliari), Giordano Mion (U Catholique Louvain), Gianmarco Ottaviano* (U Bologna)

Integration, Firm Selection and the Costs of Non-Europe

In models with heterogeneous firms trade integration has a positive impact on aggregate productivity through the selection of the best firms as import competition drives the least productive ones out of the market. To quantify the impact of firm selection on productivity, we calibrate and simulate a multi-country multi-sector model with monopolistic competition and variable markups using firm-level data and aggregate trade figures on a panel of 11 EU countries. We find that EU trade has a sizeable impact on aggregate productivity. In 2000 the introduction of prohibitive trade barriers would have caused an average productivity loss of roughly $13$ per cent, whereas a reduction of intra-EU trade costs by 5 per cent would have generated a productivity gain of roughly 2 per cent. Productivity losses and gains, however, vary a lot across countries and sectors depending on market accessibility and trade costs. We provide evidence that our results are robust to alternative distance and productivity measures.

Kazuhiro Yamamoto* (Osaka University)

Location of industry and market size in a world of imperfect capital mobility

This paper examines the impact of imperfect international capital mobility on industrial location when increasing returns are present. When the international capital mobility is perfect, agglomeration of manufacturing firms is progressed with decline in transportation costs of manufactured goods, and full-agglomeration in a large market country is observed at low transportation costs. In contrast, when international capital mobility is imperfect, agglomeration in the large market country is progressed with capital trade integration. When transportation costs of manufactured goods are low, all capital holders in two countries invest their capital into home market: the distribution of firms between two countries holds the same value to the distribution of capital holding between two countries.

Session 10. Honouring M. Fujita: Urban modeling

Robert Helsley* (University of British Columbia), William Strange (University of Toronto)

A Game-Theoretic Analysis of Skyscrapers

The development of structural steel and the elevator in the late 19th Century allowed buildings to be constructed economically at heights that had previously been impossible. But tall buildings have never been about economy alone. Ever since the development of the first skyscrapers, it has been clear that value has been placed not just on the pro forma attribution of value to leasable space but also on building height in and of itself. One dimension of building height that appears to have been central to builders is relative height. In 1909, the Metropolitan Life Building became the largest occupied building in the world, with only the Eiffel Tower reaching a greater height. In 1913, the Woolworth Building was completed, displacing the Metropolitan Life Building at the top of the list of the World’s Tallest Buildings. It private conversations with builder Louis Horowitz, Frank W. Woolworth made it clear that an accounting that related the costs of the building to its leasing revenues failed to capture the great value that accrued to being tallest: winning mattered in that it established the Woolworth brand as a corporation and it marked an achievement for Woolworth himself as an entrepreneur. In other words, Woolworth assigned value to being tallest that was independent of the narrow value of the skyscraper as a piece of real estate. This situation is one that has repeated itself several times, notably the skyscraper race between the Manhattan Company Building (40 Wall Street), the Chrysler Building, and the Empire State Building. The situation seems to be repeating itself today, with new holders of the world’s tallest title being built every few years and other contenders in the process of being built. A common feature of these situations is that builders appear to assign value to being biggest, which results in their topping each other with structures of undeniable symbolic significance but doubtful economy. This paper carries out a game-theoretic analysis of such a skyscraper contest. A builder is assumed to have a payoff function that depends in part on profits, as derived from a standard model of urban spatial structure. Builder payoff also depends on whether the builder has developed the tallest structure in his or her market. The paper considers both simultaneous-move and sequential skyscraper games. The main conclusion that emerges from the analysis is that the contest results in dissipation, with the value of the tallest-building prize at least partially lost in the poor economics of skyscrapers. In the simultaneous move version of the game, the dissipation manifests itself in a mixed strategy equilibrium where all but the highest-value builder gain no expected value whatsoever from competing in the skyscraper contest. Although the highest-value builder does enjoy positive expected surplus from the contest, there is partial expected dissipation of the fruits of victory for this builder as well. This sort of race is at least broadly consistent with the historical evidence sketched above. In a sequential version of the game, the dissipation takes the form of costly pre-emption, where the leader builds a tall-enough building to deter competing builders. This pre-emption also seems to be consistent with observation, in particular with the Empire State Building’s long stay at the top of the tallest-building list. These results are relevant to several significant issues in urban economics. First, they bear on agglomeration in that skyscrapers allow the concentrations of great numbers of workers and businesses in very close proximity. It is well-known that in the presence of positive externalities associated with agglomeration, there exists a tendency for density to be inefficiently low in the market equilibrium. The skyscraper contest results suggest that an opposing tendency may exist to build at excessive densities. Second, the results also bear on the related issues of the health of central cities and so-called “urban sprawl.” Opponents of sprawl argue that the tendency to decentralize spatially is inefficiently strong in equilibrium since suburbanites and exurbanites do not bear the full costs of commuting or decentralized public good and service provision. The skyscraper contest results suggest an opposing tendency, one that tends towards a more centralized urban spatial structure. Third, the results bear on the tendency towards overbuilding in real estate markets. Overbuilding has been identified as an important aspect of real estate cycles in nearly every city. It has been variously attributed to irrationality on the parts of builders and lenders, to incentive problems in banking, and to particular features of tax codes. This paper’s results suggest another possible foundation, one arising from strategic interactions between builders.

Jan Brueckner* (University of California, Irvine), Stuart Rosenthal (Syracuse University)

Gentrification and Neighborhood Housing Cycles: Will America's Future Downtowns Be Rich?

This paper identifies a new factor, the age of the housing stock, that affects where high- and low-income neighborhoods are located in U.S. cities. High-income households, driven by a high demand for housing services, will tend to locate in areas of the city where the housing stock is relatively young. Because cities develop and redevelop from the center outward over time, the location of these neighborhoods varies over the city’s history. The model predicts a suburban location for the rich in an initial period, when young dwellings are found only in the suburbs, while predicting eventual gentrification once central redevelopment creates a young downtown housing stock. Empirical work indicates that if the influence of spatial variation in dwelling ages were eliminated, longstanding central city/suburban disparities in neighborhood economic status would be reduced by up to 50 percent. Model estimates further predict that between 2000 and 2020, central-city/suburban differences in economic status will widen somewhat in smaller cities but narrow sharply in the largest American cities as they become more gentrified.

Charles de Bartolome* (University of Colorado), Stephen Ross (University of Connecticut)

The Location of the Poor in a Metropolitan Area: A Positive Analysis

We seek to explain the stylized fact that poor households form the majority in the inner city of most American metropolitan areas. Using numerical simulations, we show that (1) typically there exist two equilibria: one in which the poor form the majority in the inner city and the other in which the rich form the majority; (2) when the metropolitan population is small, rich households “jump to the suburb” to obtain their desired public service, and this causes the growth path to select the equilibrium in which poor households are the inner city’s majority.

Yves Zenou* (Research Institute of Industrial Economics)

Why do Ethnic Minorities Search Less? A Transport-Mode Based Theory

The aim of this paper is to show that different transport modes between whites and nonwhites lead to different search intensities. We develop a theoretical model in which whites mainly use cars to commute whereas nonwhites use public transportation. We show that, for both whites and nonwhites, living in areas where employed workers' average commuting time is higher yield the unemployed to search more than in areas with lower commuting time. Because of different transport modes, we also show that white unemployed workers search more intensively than nonwhites even if both live in areas where employed workers have exactly the same average commuting time. This is because using a faster transportation mode allows unemployed whites to accept jobs that are located further away and thus to have a higher area of search than nonwhites.

Session 11. Honouring M. Fujita: Urban labour markets

Laurent Gobillon (Institut National d'Etudes Demographiques (INED)), Harris Selod* (Institut National de la Recherche Agronomique (INRA))

Acces to jobs, residential segregation and urban unemployment in the Paris Greater Area

In this paper, we first present a survey showing how residential segregation and disconnection from job opportunities can exacerbate urban unemployment. We then present a set of relevant descriptive statistics on the Greater Paris Area which characterize the region’s spatial disparities in unemployment, residential segregation, and access to jobs. These statistics are computed from the 1999 Census of the Population and traveling times provided by the French Ministry of Transportation. We then estimate the effect of indices measuring segregation and access to jobs on transitions out of unemployment using a rotational panel dataset from the French Labor Force Survey over the 1990-2002 period. Our results show that neighborhood segregation is a key factor that prevents unemployed workers from finding a job (without moving) and impedes residential mobility. The result remains when accounting for potential location endogeneity biases.

Eleonora Patacchini* (University of Rome "La Sapienza"), Yves Zenou (IUI, The Research Institute of Industrial Economics, Stockholm)

Spatial Dependence in Local Unemployment Rates

By explicitly considering the spatial dimension of local regional labor markets, we develop a simple dynamic model that explains the spatial correlation between unemployment rates, and test it using local autority-level data in England. Our evidence shows a significant spatial dependence that has been growing over time and characterized by a low distance decay. Highly localised effects are found to derive from commuting flows. These results are consistent with the theoretical model.

Hikaru Ogawa (Nagoya University), Yasuhiro Sato* (Nagoya University), Toshiki Tamai (Nagoya University)

Tax competition and public input provision with imperfect labor markets

Incorporating a labor market imperfection caused by fixed wages into a tax competition model, this paper analyzes what will happen when jurisdictional governments provide public inputs to cope with regional unemployment. The results show that capital taxation has an employment externality. Public input provision is also shown to cause an employment externality by shifting the production frontier. These employment externalities induce the governments to levy tax on capital even if a head tax is available. The degree of the employment externalities is shown to vary depending on the level of fixed wage. Composition of public spending is also discussed.

Stephen L. Ross* (University of Connecticut), Shihe Fu (Southwestern University of Finance and Economics)

Wage Premium in Employment Clusters: Agglomeration Economies or Worker Heterogeneity?

This paper examines whether wage premia that appear to be correlated with concentrations of firm activity arise due to agglomeration economies or workers sorting across work locations based on preferences that are correlated with unobserved productivity. A worker’s residential location is used as a proxy for their unobservable productivity attributes, and the paper examines whether the estimates of work location wage premia are robust to the inclusion of controls for residential location. The models in this paper are estimated using the 5% Public Use Microdata Sample (PUMS) from the 2000 U.S. Decennial Census. Preliminary analyses suggest that a substantial fraction of wage premium can be explained by worker residential location.

Session 12. Honouring M. Fujita: Local public economics

Alex Anas* (State University of New York at Buffalo), David Pines (Tel-Aviv University)

Anti-sprawl policies in a system of congested cities

Armed with recurring analyses since the mid 1960s, economists believe that the under pricing of traffic congestion in urban areas causes not only excessive travel but also excessively low land use densities and excessively spread out cities, a condition popularly known as urban sprawl. We consider congested cities of different population sizes linked to each other by migration. Under plausible general assumptions, we show that first-best-optimal congestion tolls reduce excessive travel and travel expenditure by shifting population from large to small cities, increasing aggregate land consumption in the city system. Thus more not less geographic sprawl is associated with economic efficiency. When tolls are not available, planners can achieve a second-best optimum by zoning or subsidizing smaller cities for expansion while simultaneously taxing or restricting land expansion in the large cities. In this case also, aggregate urban land across all city sizes expands as excess travel is reduced, provided the elasticity of substitution between land and other goods is sufficiently close to zero.

Yoshitsugu Kanemoto* (University of Tokyo), Satoshi Hikino (University of Tokyo), Chisato Asahi (Ministry of Land, Infrastructure and Transport)

Consumption Side Agglomeration Economies in Japanese Cities

We estimate the magnitudes of the consumption side urban agglomeration economies for Japanese metropolitan areas. Following the pioneering work of Tabuchi and Yoshida (2000), our approach exploits the fact that consumers tolerate higher living costs in larger cities, in particular, higher housing costs, if they benefit from urban agglomeration. This living cost approach requires an appropriate measure of the average living cost in a metropolitan area, which is not easy to estimate because housing prices have extremely wide variation within a metropolitan area. Tabuchi and Yoshida (2000) chose the average land price for commercial use. Since the prices of residential land are typically much lower than those of commercial land, this might have resulted in biased estimates. We estimate bid rent functions for municipalities within metropolitan areas to obtain a better estimate of urban agglomeration economies.

Christian A. L. Hilber (London School of Economics), Frederic L. Robert-Nicoud* (London School of Economics)

Desirable Locations and Land Use Constraints

We model residential land use constraints as the outcome of a political economy game between owners of developed and owners of undeveloped land. Land use constraints are interpreted as shadow taxes that increase the land rent of already developed plots and reduce the amount of new housing developments. In general equilibrium, locations with nicer amenities are more developed and, as a consequence, more regulated. We test our model predictions by geographically matching amenity, land use, and historical Census data to metropolitan area level survey data on regulatory restrictiveness. Using amenities as instrumental variables, we demonstrate that metropolitan areas with better amenities are indeed more developed and more tightly regulated. Moreover, consistent with theory, metropolitan areas that were more regulated in the 1980s observed a greater slowdown in the growth rate of new housing construction from the 1980s to the 1990s.

Hideo Konishi* (Boston College)

Tiebout's Tale in Spatial Economies: Entrepreneurship, Self-Selection and Efficiency

This year is the fifty year anniversary of celebrated Tiebout's paper (Tiebout, 1956 JPE). This paper proposes an equilibrium concept based on Rothschild and Stiglitz (1977 QJE) in local public goods economy a la Tiebout (1956 JPE) with spatial elements. Land, location-specific production technology, and wage differences are introduced, and taxes are land taxes (property taxes). Assuming small group effectiveness in the manner of Wooders (1978), Kaneko and Wooders (1986 MASS), Ellickson et al (1999 Econometrica) and Conley and Wooders (1997 mimeo), we show the existence and efficiency of equilibrium. The key requirement is anonymity of land tax, which is attained by imposing Hamilton's (1975 Urban Studies) zoning constraints (otherwise, no equilibrium based on the logic of Rothschild and Stiglitz). Wage differential across locations are allowed, yet we can assure efficiency of equilibrium in our particular model with small group effectiveness, despite of intuitions by Tiebout (1956 JPE), Buchanan and Wagner (1970 essay) and Flatters, Henderson and Mieszkowski (1974 JPubE). Lastly, our theorem can be directly applicable to the existence and efficiency of a monocentric city equilibrium in urban economics with commuting time costs even if we allow existence of collective residences such as apartments.

Session 13. Honouring M. Fujita: Sprawl

Stephen Sheppard* (Williams College)

Infill Versus Outgrow: Estimating the Microstructure of Urban Expansion

The continuing pace of urbanization around the globe has generated an ongoing process of expanding urban land use. This is verified both in everyday experience and in empirical studies such as Burchfeld, Overman, Puga and Turner [2006] or Sheppard, Civco and Angel [2006]. Several factors contribute to the growth in urban land use: for example increasing population and income, falling transportation costs, tax and regulatory distortions and the spatial distribution of amenities. In this paper we investigate how the local structure of urban expansion is affected by these factors. Some of these forces encourage increasing urban land use at the periphery, and some encourage infill of vacant areas within the city. It is difficult to empirically test our understanding of these processes using data from a single city or even a single country because the data do not offer sufficient independent variation in the relevant factors to reliably distinguish their separate impacts. This paper makes use of newly collected data from a global sample of cities with land use data classified from satellite imagery in all cities. The analysis distinguishes increasing urban land use at the periphery from increasing urban land use through infill of built up areas. We obtain theoretical predictions for the expected impact of economic variables on each type of urban expansion, and test the predictions using the data.

Elena Irwin* (Ohio State University), Hyun Jin Cho (Ohio State University), Nancy Bockstael (University of Maryland)

Measuring and modeling urban sprawl: Data, scale and spatial dependencies

Although urban decentralization is a well documented trend, the extent to which this decentralization process is characterized by discontinuous development or “sprawl” is widely debated. Because the measurement of urban land use patterns depends critically on the data source, resolution, definitions and scale, it is likely that different conclusions will be drawn depending on the data and scale of analysis. First, we use spatial statistics to investigate a series of hypotheses regarding the data dependency of urban pattern analysis using three sources of land use data for the State of Maryland that differ in their definitions, resolution and scale. Our analysis reveals enormous differences across datasets in the basic measurement of urban land and its pattern. For example, we find that 82 percent of all land identified by the most accurate dataset as low density residential land in 2000 is classified as undeveloped by the remotely sensed data. Second, we use spatial statistics and the most accurate data to investigate a set of hypotheses regarding the spatial and scale dependencies of urban land use patterns and changes in these patterns over time. We examine how various dimensions of urban pattern, including density, patch size, spatial autocorrelation and fragmentation, change along an urbanization gradient and across different spatial scales of analysis ranging from land use patches to entire counties. Lastly, we estimate a series of reduced form models to examine spatial and scale dependencies in the associations between urban land use pattern and heterogeneous features of the landscape, including distance to urban centers, road access, public services, surrounding land uses and land suitability. For each spatial scale, we attempt to isolate the spatially heterogeneous features that matter with respect to urban pattern by comparing the expected urban land use pattern with a pattern generated using the mean values of these spatially heterogeneous variables. The results demonstrate a number of interesting findings regarding the spatial and scale dependence of urban land use patterns and the dynamic process of urbanization.

Henry Overman (London School of Economics), Diego Puga* (Universitat Pompeu Fabra), Matthew Turner (University of Toronto)

The relationship between the growth in residential land use and population

The amount of land built up for residential use in the United States is growing much faster than population and there is a common perception that this reflects almost entirely the fact that people are building larger houses. In this paper we show that this is not the case. Population Between 1976 and 1992 the amount of land built up for residential use in the United States grew by 47% while population only grew 17%. However, only 27.1% of the growth in US residential area can be attributed to individual households using a greater amount of residential land in each state. 36.1% is due to overall population growth, 21.2% to an increase in the number of households over this period as baby-boomers have left their parents' houses, 5.9% to the shift of population towards states with larger houses, and the rest to interactions between these changes.

Antonio Bento* (University of Maryland)

Urban Decline, Urban Sprawl, and the Double Dividend Hypothesis

This paper asks the following questions: Does the introduction of a metropolitan wide development tax produce a double dividend when revenues are earmarked to finance a revitalization program? What are the potential welfare gains of such policy and how does it compare with an urban growth boundary? To properly address these questions and guide policy, there is a need to develop a framework that simultaneously capture the following key aspects of the problem: the durability of the housing stock, including the fact that some housing exhibits a quality that is below the threshold accepted by building regulations; neighborhood urban decline that results from the spatial concentration of degraded housing, which in turn generates negative externalities to near by neighborhoods; adjustment costs in the decision to improve the existing housing stock; strategic behavior amongst developers when deciding on the level of improvements; production of new housing at the urban fringe; and spatially concentrated benefits from open space at the urban fringe

Session 14. Migration

Luisito Bertinelli* (University of Luxembourg), Salvador Barrios (European Commission, Institute for Prospective Technological Studies), Eric Strobl (Ecole Polytechnique , Paris)

Climatic Change and Rural-Urban Migration: The Case of Sub-Saharan Africa

We investigate the role that climatic change has played in the pattern of urbanization in sub-Saharan African countries compared to the rest of the developing world. To this end we assemble a cross-country panel data set that allows us to estimate the determinants of urbanization. The results of our econometric analysis suggest that climatic change, as proxied by rainfall, has acted to change urbanization in sub-Saharan Africa but not elsewhere in the developing world. Moreover, this link has become stronger since decolonization, which is likely due to the often simultaneous lifting of legislation prohibiting the free internal movement of native Africans.

Curtis Simon* (Clemson University), Robert Tamura (Clemson University)

Fertility and the City

This paper documents a negative relationship between fertility and the scarcity of living space across U.S. cities over the period 1940-1990 using data from the Integrated Public Use Microdata Samples (IPUMS). We use our estimates to try to (1) account for the U.S. baby boom, explained by Greenwood et. al. (2005) as resulting from household productivity shocks; (2) account for the U.S.-Europe fertility differential; and (3) estimate the marginal value of a child. Our preferred estimates can account for ¼-1/3 of U.S.-Europe fertility differential, and imply a marginal value of a child of $60,000. However, our estimates can account for only 12-16% of the baby boom.

Sukkoo Kim* (Washington University)

Immigrants and Cities

see attached

Raven Saks* (Federal Reserve Board of Governors), Abigail Wozniak (University of Notre Dame)

Is Migration Procyclical? New Evidence on Fluctuations in Internal Migration over the Business Cycle

This paper provides new evidence on the cyclicality of internal migration patterns in the United States. Using historical reports of the Current Population Survey, we examine gross migration patterns during the entire postwar era, a period that spans ten recessions over more than fifty years. We obtain additional evidence on state-to-state population flows during the past thirty years from statistics compiled by the Internal Revenue Service. We find that internal migration within the US is strongly procyclical in both sources. Even after accounting for variation in relative local economic conditions, migration is lower during downturns in the aggregate economy. To investigate explanations for this result, we use Current Population Survey microdata to identify the personal and demographic characteristics of individuals for whom the probability of migrating is most strongly tied to the business cycle. The procyclicality of migration is strongest for younger and less educated workers, but is not related to employment status. In closing, we relate our findings to the current literature on search models of the labor market. The new facts on worker flows that we document in this paper suggest that labor market matching models should be extended to incorporate aggregate macroeconomic conditions.

Session 15. Measurement of Agglomeration

Bentley Coffey* (Clemson University)

In Search of a Monocentric City

Despite the pervasiveness of strong assumptions about monocentricity and radial symmetry in theoretical work on cities, there is a dearth of empirical research on the shape of actual cities. This paper fills the literature’s gap by testing those city shape assumptions with a new semi-nonparametric methodology that is nearly as flexible as a purely nonparametric approach while nearly retaining the computational (and conceptual) simplicity to estimation and testing as a purely parametric approach. Applying this procedure to a cross-section of 62 cities with tract-level data collected in the 2000 Census, we find strong evidence that most cities are neither monocentric nor radially symmetric.

Andrea Lamorgese* (Banca d'Italia), Gianmarco Ottaviano (University of Bologna)

Intercity interaction

Using national level input-output matrices, we propose a strategy to identify pecuniary externalities operating through the markets for intermediate goods at the local level. Then, controlling for common shocks in a spatial econometric framework, (i) we estimate the effect of pecuniary externalities on productivity growth; (ii) we disentangle such effect from the one of other local interactions (i.e. knowledge or other face-to-face spillovers) and that of local characteristics; (iii) we evaluate the spatial scope of operating of all kind of externalities using different distance measures. Our estimates suggest that pecuniary externalities and other kinds of local interactions coexist, that their effect on productivity growth is decreasing with distance, thus inducing agglomeration, and that it depends on inter-city diversity and the pattern of local specialisation.

Giulio Bottazzi (Sant'Anna School for Advanced Studies), Giovanni Dosi (Sant'Anna School for Advanced Studies), Giorgio Fagiolo* (University of Verona), Angelo Secchi (Sant'Anna School for Advanced Studies)

Sectoral and Geographical Specificities in the Spatial Structure of Economic Activities

This work explores the spatial structure of location of production activities. We try to disentangle location- from sector-specific drivers in the dynamic process of spatial agglomeration. We argue that the former typically apply "horizontally" (i.e. across all industrial sectors), while the latter unfold in the form of non-decreasing dynamic returns to the current stock of installed business units. A stochastic model of location is developed and three different specifications are tested against Italian data on the location of manufacturing firms. Our results suggest that different locations exert different structural influences on the distribution of production activities. Moreover, a widespread horizontal power of "urbanization", which makes particular locations more attractive irrespectively of the sector, does emerge. However, after controlling for the latter, one is still left with sector-specific forms of dynamic increasing returns to agglomeration, which vary a lot across different manufacturing activities.

Arthur O'Sullivan* (Lewis & Clark College)

Using GIS to Map and Measure the Spatial Distribution of Employment within Cities

Geographic Information Systems (GIS) provide powerful tools to describe the spatial distribution of economic activity within cities. I will use GIS software to generate maps of employment density by census tract and block group, for total employment and for employment in individual sectors, e.g., manufacturing, FIRE, services, and retail. I will generate maps for several metropolitan areas. GIS software also computes spatial statistics that describe the degree of centralization, concentration, and clustering. I will compute the following for total employment and for employment in individual sectors. • Standard deviation distance is the radius of a circle containing one standard deviation of observations. The • Moran’s I measures the degree of spatial autocorrelation for different levels of geography, providing a general measure of concentration. • The Getis-Ord Gi* identifies clusters or “hot spots” of economic activity. I will compare the results of GIS analysis to the results of earlier studies of the intraurban distribution of employment. Of particular interest is how the hot-spots identified by GIS compare to the employment centers and subcenters identified in earlier studies.

Session 16. Rental Markets

Pablo Casas-Arce (University of Oxford), Albert Saiz* (The Wharton School)

Do Courts Matter? Rental Markets and the Law

We argue that the allocation of ownership rights will minimize enforcement costs when the legal system is inefficient. In particular, when legal enforcement is costly, there will be a shift from contractual arrangements that rely on such enforcement (such as a rental agreement) towards other forms that do not (such as direct ownership). We then test this prediction on data on the rental housing market, and show that costly enforcement of rental contracts hampers the development of such a market in a cross-section of countries. We argue that this association is not the result of reverse causation from a developed rental market to more investor protective enforcement. The results provide supportive evidence on the importance of contract enforcement for the development of financial and other markets.

David Genesove* (Hebrew University of Jerusalem)

Real Effects of Exchange Rate Movements in U.S. dollar denominated Isareli Rents

Apartment rents in Israel are typically set in U.S. dollar terms. This paper asks whether that convention (left over from the hyper and high inflationary times of twenty years ago) has real effects. In particular, it considers whether movements in the shekel-U.S. exchange rate, instrumented by movements in the Euro-U.S. exchange rate, affect real housing rents at the time of contract setting for new tentants.

Etienne Wasmer* (UQAM et IEP Paris), Francis Kramarz (CREST)

Tenant’s Eviction and Discrimination in the French Housing Market: An Empirical Strategy”

Housing and labor markets exhibit many similarities. First, information is imperfect. Tenant quality, like worker quality, is unobserved. Second, separation is costly and time consuming. The laws and regulation typically complicate or slow down the termination process of the contractual relationship and make it more costly for firms and landlords to fire an employee /evict a tenant. And finally, there are rigidities in nominal wages and rents. Adapting tools from labor theory, we attempt to understand how landlords wish to screen and possibly statistically discriminate against potential tenants. They do so when housing regulations are more stringent. An empirical strategy, based on administrative data on tenants' eviction in France is provided.

Donald Haurin* (Ohio State University), Hazel Morrow-Jones (Ohio State University)

The Differences in Financial, Mortgage, and Real Estate Market Knowledge between African American and White Renters and Homeowners

We hypothesize that a household’s level of knowledge about the real estate market and financial information is determined simultaneously with its tenure decision. Undergoing the process of searching for and obtaining a home and mortgage should increase a household’s amount and quality of financial information. On the other hand, we hypothesize that the lack of information about real estate and financial markets is a barrier to home buying. Our empirical results support both of these views. The most important finding is that lack of financial and real estate knowledge is a significant barrier to becoming a homeowner. This result, derived from a simultaneous equations estimation model, is consistent with descriptive data reported in the survey where many renters indicate that the lack of knowledge about how to buy a home, how to get a real estate agent, and how to get a mortgage are barriers to their becoming homeowners. However, we find that the racial differences in the knowledge variable used in the estimation are small and thus it explains only a small part of the racial gap in homeownership rates.

Session 17. Local Taxation and Public Goods

Yannis Ioannides (Tufts University), Kurt Schmidheiny* (Tufts University)

Estimating Equilibrium Models of Local Jurisdictions: A Discrete Choice Approach with Individual and Community-Level Data

The paper develops a discrete choice model of community choice by combining features of the approach by Berry, Levinsohn and Pakes (1995; 2004) with the approach of Epple and Sieg (1999) and of Epple et al. (2001; 2006). The paper reports estimation results that involve an iterative procedure consisting of two stages. At a first stage, we use information on the joint distribution of household characteristics in a metropolitan area, in our case obtained from the American Housing Survey Boston metropolitan area micro sample, to predict population shares and moments of household characteristics for each community. We then match them with population shares and moments from the US Census by means of a generalized method of moments method. This stage estimates coefficients reflecting interactions between individual and community characteristics and community-specific intercepts. These intercepts serve as sufficient statistics for the estimation, at a second stage, of coefficients expressing the effects of community-specific characteristics by the different communities in Boston metropolitan area in 1980. We use the same data set from the US Census that Epple et al. have used, but augment it by community-specific housing prices, which are obtained from the record of housing transactions. Our results demonstrate that use of two public sources of data circumvents the need of confidential data that other research has relied on.

Carlo Cambini (Polytechnic of Torino ), Massimiliano Piacenza (Ceris-CNR National Council of Research), Davide Vannoni* (University of Torino)

Restructuring Public Transit Systems: Evidence on Cost Properties and Optimal Network Configuration from Medium and Large-Sized Companies

This paper analyses the cost structure of a sample of Italian local public transport (LPT) companies operating in medium and large urban centres. The main focus is to identify the proper network configuration for the LPT service, by verifying the presence and the extent of scale and density economies. Technological characteristics of public transit systems are analysed by estimating both variable and total cost function models, which consider three alternative supply-oriented output measures, include firm-specific fixed effects, and allow for X-inefficiency to play a role through the estimation of a stochastic cost frontier. The evidence is remarkably robust across the different specifications which have been tested and shows the presence of short-run and long-run economies of scale, as well as of economies of network density, for both the average sample firm and for operators belonging to the highest percentile (large-sized companies). This suggests that, from a technological point of view, a proper LPT network design should at least include a large urban centre and should be extended so as to embrace the intercity service too, while a regulatory policy aimed at fragmenting the served area in various sub-networks would imply an efficiency loss.

Sylvie Charlot* (INRA), Sonia Paty (INRA and University Lille 1)

Taxable agglomeration rent: evidence from a panel data

The main purpose of this paper is to test the existence of a taxable agglomeration rent in the French local tax setting by taking account the tax interactions among the urban jurisdictions. After presenting a simple economic geography model with fiscal interactions, we estimate a model of tax setting for the local business tax using the econometrics techniques on panel data for 1993 to 2002. We observe that the relationship between tax rate and fiscal base gives presumption of the existence of a ``taxable agglomeration rent''.

Patricia Beeson (University of Pittsburgh), Lara Shore-Sheppard (Williams College), Tara Watson* (Williams College)

The Effect of Local Fiscal Policies on Urban Wage Structure

While it has long been recognized that average wages vary strikingly across regions and urban areas, differences in the variance of wages remain relatively unexplored. In this paper we empirically examine differences in the extent and persistence of wage dispersion across urban areas. Using data from the 1980 and 1990 Censuses, we show that metropolitan area wage distributions vary, that the variation is substantial, and that it is not entirely accounted for by differences in the supply of workers with different skills or the size or geographic region of the city. We find that the differences in wage distributions across cities are highly persistent. We investigate whether there is a link between local fiscal policy and the degree of dispersion in the wage structure, and find evidence that such a relationship exists. Cities with higher overall taxes, fewer transfers from state and federal governments, and a greater share of spending on public health and community development appear to have higher levels of overall dispersion. In addition, we find that cities that rely more heavily on property taxes have greater dispersion in the lower half of the wage distribution, and cities with higher expenditures on education have more dispersion in the upper half.

Session 18. New Economic Geography

Masahisa Fujita* (Kyoto University)

Economic Development Capitalizing on Brand Agriculture: Turning Development Strategy on Its Head

Challenging the common literature of economic growth and development that tends to assign the agrarian sector the backseat, this paper seeks to move agriculture and rural development to the forefront through a community-based strategy designed to cause major innovation dynamics and human capital accumulation to occur. Specifically, this paper explores the possibilities of two unique Japanese concepts advocating community-based rural development -- the One Village One Product Movement (OVOP) and Michino Eki (or Roadside Stations) -- initiated in the peripheral regions of Japan in the early 1960s and the mid-1990s, respectively. Both OVOP and Michino Eki have attracted widespread attention in many developing countries as potential tools for bridging the gap between cities and rural areas through community-driven development, and are being implemented in many countries. From the viewpoint of spatial economics and endogenous growth theory, this paper considers both OVOP and Michino Eki as rural development strategies of a broader nature based on “brand agriculture." Here, brand agriculture represents a general strategy for community-based rural development that identifies, cultivates and fully utilizes local resources (including natural, historical, cultural and human resources) for the development of products or services unique to a certain "village" or geographical area. This process feeds the engine of sustained development of a greater variety of these products and gradually establishing these local brands in increasingly larger markets. In the context of brand agriculture, selected examples of OVOP and Michino Eki from Japan and developing countries are introduced, with special attention given to the role of various types of infrastructure in the effective promotion of OVOP and Michino Eki. The paper concludes with a discussion of policy implications for successful promotion and implementation of brand agriculture programs in developing countries.

Yasusada Murata* (Advanced Research Institute for the Sciences and Humanities, Nihon University)

Intergenerational Linkages and Economic Geography

This paper presents a model of economic geography with overlapping generations and shows that intergenerational linkages lead to the symmetric equilibrium in which young and old people are equally mixed in each region. Once the discount factor is taken into account, however, symmetry breaks and ex ante identical agents in the same generation are divided into those with higher lifetime income and greater product diversity when old and those with lower lifetime income and greater product diversity when young, thus yielding spatial sorting by age. This is because when making a location decision, each individual cares about how many varieties are available at which stage of a life cycle in the destination.

Antonella Nocco* (University of Lecce)

Preference heterogeneity and economic geography

New economic geography models analyze agglomeration and dispersion forces, whose interactions determine the spatial distribution of economic activity. In this paper we investigate the effect of preference heterogeneity between skilled and unskilled workers on agglomeration. We introduce consumers’ taste differences in the model by Ottaviano et al. [2002], and we argue that this allows us to represent an additional source of dependence of equilibrium prices on the demand properties shaped by the interregional distribution of workers. In particular, we identify a new preference effect, and we show that when the intensity of skilled workers’ preference for the modern good and its variety is strong enough, prices charged by firms, either local or foreign, may increase when the mass of local firms increases therefore acting as a dispersion force.

Kristian Behrens* (CORE Université catholique de Louvain), Pierre Picard (CORE Université catholique de Louvain)

Tax competition, location, and horizontal foreign direct investment

We develop a model of capital tax/subsidy competition in which imperfectly competitive firms choose both the number and the location of the plants they operate. The endogenous presence of horizontal multinationals is shown to attenuate the ``race to the bottom'' and yields some results that are opposite to traditional findings in the tax competition literature. First, in the presence of horizonal multinationals, increasing subsidies decrease firms' profits by exacerbating price competition due to more firms `going multinational'. Second, instead of being always subsidized, capital may actually be taxed in equilibrium. Third, taxes/subsidies become strategically independent policy instruments, instead of being strategic complements. Last, there may exist multiple equilibria with either low or high subsidies.

Session 20. Urban Sprawl

Matthew Turner* (University of Toronto)

A Simple Theory of Smart Growth and Sprawl

This paper considers the simultaneous determination of residential density and the supply of local versus remote retail services. Possible equilibrium development patterns either correspond closely to what anti-sprawl activists describe as smart growth, or to its opposite. Equilibrium and optimal patterns of development do not always coincide, and when they diverge, optimal density is always higher than equilibrium density, and, equilibrium development is discretely rather than marginally different from the optimum. This occurs in the absence of congestion externalities, and is due to a free-rider problem and a coordination problem. The analysis indicates that a tax on large lots or a subsidy for small lots may be welfare improving under certain conditions.

Jean Cavailhès (INRA-CESAER, Dijon), Dominique Peeters* (Université catholique de Louvain)

Residential Equilibria in a Green Urban Area

‘Green’ amenities attract households into the countryside around cities where a mixed residential and agricultural land use prevails. We model a land market where two rents co-exist for those two land uses. The purpose is to define a zoning system that maximizes landowners' income by providing the optimal quantity of agricultural amenities enjoyed by households close to farmland. Urban configurations and analytical properties of the equilibrium are presented. Simulations with actual data from the Dijon urban area (France) supplement the analysis.

Yasushi Asami* (University of Tokyo), Yukari Niwa (University of Tokyo)


A method for identifying typical lots in a residential block is proposed based on the shape distance matrix among all the lots in the block. The method is applied to 20 blocks in Setagaya ward in Tokyo. The typical lots tend to be rectangular-like lots even in irregularly shaped blocks, suggesting the preference for rectangular shape in the choice of residential lots. Moreover, rectangular like blocks tend to contain rectangular typical lots, such that length of their depth is approximately a half of the shorter edge of the block, and two variation of width, which may imply that part of earlier lots were subdivided into halves.

Christopher Wheeler* (Federal Reserve Bank of St. Louis)

Urban Decentralization and Income Inequality: Is Sprawl Associated with Rising Income Segregation Across Neighborhoods?

Existing research has found an inverse relationship between urban density and the degree of income inequality within metropolitan areas, suggesting that, as cities spread out, they become increasingly segregated by income. This paper examines this hypothesis using data covering more than 160000 block groups within 359 US metropolitan areas over the years 1980, 1990, and 2000. The findings indicate that income inequality - defined by the variance of the log household income distribution - does indeed rise significantly as urban density declines. This increase, however, is associated with rising inequality within block groups as cities spread out. The extent of income variation exhibited between different block groups, by contrast, shows virtually no association with population density. There is, accordingly, little evidence that sprawl is systematically associated with greater residential segregation of households by income.

Session 21. Interactions Between Housing and Labour Markets

Edward Coulson* (Penn State University), Lynn Fisher (Massachusetts Institute of Technology)

Housing Tenure and Labor Market Impacts: The Search Goes On

We develop a search-theoretic model of the Oswald hypothesis, the idea that homeownership is linked to inferior labor market outcomes, and compare its predictions to two extant theories. Beyond their prediction that individual homeowners will have a higher probability of unemployment, the three models have surprisingly different predictions about the labor market at both the aggregate and micro levels. We estimate micro and US state-level regression models of wages and unemployment and compare the estimates to those predictions.

Laurent Gobillon* (Institut National d'Etudes Démographiques (INED)), Thierry Magnac (Université des Sciences Sociale in Toulouse - IDEI), Harris Selod (Institut National de la Recherche Agronomique (INRA) - LEA)

How does Residence Matter When Looking for a Job in Greater Paris?

There are large spatial disparities in unemployment durations across the 1,300 towns in the Ile-de-France region (Paris Greater Area). In order to characterize these imbalances, we estimate a proportional hazard model stratified by town on an exhaustive dataset of all unemployment spells starting in the first semester of 1996. For each town, this model allows us to recover a survival function purged of individual effects. Analyzing these survival functions, we show that nearly 80% of the gross disparities remain, especially for long-term unemployment. However, we show that 60% of the remaining differences can be explained by aggregate explanatory variables, especially indices of residential segregation.

Giorgio Topa* (Federal Reserve Bank of New York), Stephen Ross (University of Connecticut), Patrick Bayer (Yale University)

Place of Work and Place of Residence: Informal Hiring Networks and Labor Market Outcomes

We use a novel research design to empirically detect the effect of social interactions among neighbors on labor market outcomes. Specifically, using Census data that characterize residential and employment locations down to the city block, we examine whether individuals residing in the same block are more likely to work together than those in nearby blocks. We find evidence of significant social interactions operating at the block level: residing on the same versus nearby blocks increases the probability of working together by over 33 percent. The results also indicate that this referral effect is stronger when individuals are similar in sociodemographic characteristics (e.g., both have children of similar ages) and when at least one individual is well attached to the labor market. These findings are robust across various specifications intended to address concerns related to sorting and reverse causation. Further, having determined the characteristics of a pair of individuals that lead to an especially strong referral effect, we provide evidence that the increased availability of neighborhood referrals has a significant impact on a wide range of labor market outcomes including employment and wages.

Janet Kohlhase* (University of Houston), Jia-Huey Lin (University of Houston)

Spatial Mismatch and Urban Labor Markets in the United States: Evidence for Blacks, Immigrants and Hispanics

(see extended abstract below)

Session 23. Commuting

André de Palma* (University of Cergy-Pontoise and Ecole Nationale de Ponts et Chaussées,), Kiarash Motamedi (University of Cergy-Pontoise), Nathalie Picard (University of Cergy-Pontoise and INED), Paul Waddell (University of Washington)

A Model of Residential Location Choice with Endogenous Housing Prices and Traffic for the Paris Region

There is a growing interest in the development and the use of large-scale planning models. In this paper, we describe the first step of a project to integrate UrbanSim, a dynamic microsimulation land use model, and METROPOLIS, a dynamic traffic model. This is the first attempt, to our knowledge, to integrate a dynamic land use model and a dynamic traffic model. We briefly describe the two models and propose a unified framework for their integration. Within this integrated framework we develop a model of residential location choice, with endogenous housing prices and traffic. The study area for this research is the Ile-de-France (Paris region), for which we provide empirical results.

Jos van Ommeren* (Free University), Wouter Vermeulen (Free University)

Commuting and spatial structure

Urban economic theory and public policy are based on the assumption that the spatial structure of the economy, so the spatial distribution of jobs and residences, determines the length of the commute. Empirical support for this assumption is weak, but based on cross-section observations. We examine this assumption employing panel data for the Netherlands between 1990 and 2000 which reduces measurement problems. Hence, we examine to what extent the change in the length of the commute is due to changes in spatial structure. We show for the Netherlands that about 65% of the average increase in the length of the commute can be attributed to changes in spatial structure.

Elena Safirova* (Resources for the Future), Sébastien Houde (Resources for the Future), D. Abram Lipman (Resources for the Future), Winston Harrington (Resources for the Future), Andrew Baglino (Resources for the Future)

Congestion Pricing and Land Use Change

In their quest for solving the problem of ever-increasing traffic congestion, cities around the globe are facing more and more difficulties building new roads. Problems with financing, public opposition, environmental concerns all contribute to an infrastructure gridlock. One of the modern solutions to alleviate infrastructure problems is to introduce road pricing. On the one hand, road pricing promises to improve efficiency of existing roads and therefore effectively postpone the need for building new expensive infrastructure. On the other hand, revenue collection can at least partially finance infrastructure investments. Of course, road pricing comes with its own package of problems, such as concerns about its effectiveness and distributional effects. One of the effects of congestion pricing that is largely unknown is its long-term economic impact that can result from a series of adjustments in wages, rents, and spatial distribution of jobs and residents in the city. Relying on the short-run projections alone may be inaccurate, especially when estimating future revenue collection. In this paper, we evaluate economic effects of a congestion pricing scheme in the long run when land use and economic changes are taken into account and compare them with short-run effects when the land use and economic changes are ignored. In order to conduct this analysis we employ a spatially disaggregated general equilibrium model of regional economy that incorporates decisions of residents, firms and developers integrated with a spatially disaggregated strategic transportation planning (START) model that features mode, time period and route choice. First, we look at the impacts of a road pricing policy evaluated based on the integrated model and compare them with short-term effects obtained from the START model alone. Then we analyze how optimal short-run policy differs from the optimal long-run policy. Finally, we look at distributional impacts of the policy in question and point out their differences and similarities the short and long run.

Jan Rouwendal* (Free University), Thomas de Graaff (Free University)

Unemployment and Commuting in a Spatial Labor Market with Search

In this paper we develop a model for a spatial labor market in which employment is spread over a finite number of locations and workers are located around employment centers. Information frictions on the labor market induce workers to accept job offers from outside their city of residence so as to reduce unemployment spells. This generates commuting flows between cities. We investigate the optimal search strategy of unemployed workers and the labor market equilibrium with commuting. Unemployment rates are inversely related to the accessibility to jobs in the worker’s city of residence as well as in other cities, which agrees with the spatial mismatch hypothesis. We take a closer look at the special case in which all cities have the same numbers of jobs and workers and identical positions with respect to other cities, and in which the matching function has constant returns to scale. We study the welfare losses caused by information friction and distance friction and the effect of changes in the wage and in intercity commuting costs on them. The cost benefit analysis of transportation improvements suggested by our model differs from the conventional one.

Session 24. Urban evolution

Jens Suedekum* (University of Konstanz), Kristian Behrens (CORE - Université catolique de Louvain)

Convergence of human capital shares across German cities

In this paper we analyze the impact of human capital on local employment growth for the case of West Germany (1977-2002). We find robust evidence that a large initial share of high-skilled workers in a city raises subsequent total employment growth, but reduces subsequent local growth of high-skilled jobs. In other words, there has been convergence of local human capital shares across German cities over time. This stylized fact sharply contrasts evidence from the US. We then present a theoretical model that is consistent with the empirical observations from Germany. Our main point is to argue that the observed convergence trend is consistent with the presence of localized human capital externalities, provided they are not too strong. Finally, we present some highly tentative suggestions what might explain the differences between Germany and the US concerning the spatial evolution of human capital.

Maarten Bosker (Utrecht School of Economics, Utrecht University), Steven Brakman (Faculty of Economics, Univ of Groningen), Harry Garrretsen* (Utrecht School of Economics, Utrecht University), Herman de Jong (Faculty of Economics, Univ of Groningen), Marc Schramm (Utrecht School of Economics, Utrecht University)

The Development of Italian Cities 1300-1861

The evolution of city growth is usually studied for relative short time periods. The rise and decline of cities is, however, typically a process that takes many decades or even centuries. In this paper we study the evolution of a large sample of Italian cities for the period 1300-1861. The first contribution of our paper is that we use various descriptive statistics on individual city sizes and the city-size distribution as a whole to highlight the main characteristics of Italy’s urban system such as the differences between Northern and Southern Italy. Our second contribution is that our data allow for a panel estimation where city-size is regressed on various geography, political and other determinants of city size for the period 1300-1861. We show that, although large shocks are visible in the data, the main determinants of Italy’s city growth invariably are physical geography, political importance and, over time, the productivity in the agricultural sector and institutional constraints. In our estimations the North-South difference turns out to be important too. All in all, the historical development of the Italian cities is an intriguing mixture of stability and change.

Yuming Fu* (National University of Singapore), Sheng Zhou (National University of Singapore)

Urban Growth across Chinese Cities—Estimates of Structural Equations

We investigate the urban growth behavior across China cities between 1990 and 2004, when China’s urbanization level rose from about 28% to almost 40%. We examine urban growth in terms of population, wage rate, and per capita GDP. Unlike in developed economies like US where these three urban growth indicators would be highly correlated, they have little correlation in China. Urban population was influenced by government policies that limited population growth in large cities. Wage rate, which reflects urban productivity, would be influenced by economic reform, fixed investment, and such variables as average education level and location advantages that were found to influence urban growth in US. GDP, which reflects investment profits in addition to labor productivity, would be influenced also by openness to foreign direct investment. It is, therefore, import to analyze not only how the exogenous variables affect the alternative urban growth measures but also how these urban growth measures interact with each other during China’s recent rapid urbanization. We expand our set of exogenous instruments in our structural equations by including such variables as the distance-weighted average of GDP, infrastructure investment, human capital and manufacturing employment of all the cities other than the city in each case. These variables provide detailed measures of location advantages and disadvantages of the individual cities. Our preliminary three-stage OLS estimates show the following results: 1. The city demand for labor was downward sloping—a positive shock to population growth slowed wage rate growth and, to a lesser extent, the growth in GDP per capita. 2. Wage rate grew faster in cities with a higher average education level, a higher per capita fixed investment, and a lower share of employment in state-owned enterprises (SOEs). Other things being equal, wage rate grew more slowly for larger cities and cities with a higher initial wage rate. 3. Per capita GDP growth was driven by fixed investment and the share of FDI in total investment. It appears independent of the initial population size but considerably smaller where the initial per capita GDP and the wage rate growth were high. 4. Population growth was highly responsive to per-capita GDP growth but not responsive to wage rate growth, suggesting a highly elastic supply of unskilled labor to manufacturing job growth but an inelastic supply of skilled labor. Population growth was considerably slower for large cities, reflecting a policy bias; it was, however, significantly higher for cities with an initially high wage rate. An implication of these results is the importance of investment in skills for China’s continued productive urbanization. Apparently, receiving more unskilled workers is not going to improve the productivity of Chinese cities and the growth of the skilled jobs appears to be constrained by skill shortages. Large cities in China are yet to become productive places for unskilled migrants to acquire new skills.

Mohammad Arzaghi* (American University of Sharjah), James Davies (U.S. Census Bureau)

Urban Resurgence

We explore empirically the mobility of industry agglomerations. Within a city, some industries fade and are continuously replaced by others. For the period over 1987 to 1997, the data show of 313 PMSA cities in the U.S., 152 (49%) lost more than 15% of their city employment due to employment losses in locally declining industries. However, only 24 cities have a decline in total city employment over the same period, and for only 11 cities is the decline more than 5%. City employment is generally quite stable, yet the composition of industries within the city can change quite rapidly. This paper investigates this industry replacement process within cities.

Session Amenities

Lynn Fisher (MIT), Henry Pollakowski (MIT), Jeffrey Zabel* (Tufts University)

Amenity-Based Housing Affordability Indexes

The most common notion of affordable housing implies that households that spend more than 30 percent of their income on housing have an affordability problem. This fails to address the actual supply of “affordable” units in a geographic area or to anticipate the spatial implications of where the supply of housing is located. In contrast, we develop a new measure of area affordability that characterizes the supply of housing that is affordable to different households in different areas of a metropolitan region. By focusing on area affordability, we recognize that the price of a house is affected by its location since this price includes the value of the services provided by the local amenities. Thus, in generating an affordability index, it is important to account for locational amenities when comparing house prices across towns in a metropolitan area. We adjust our index for accessibility, school quality, and open space. We apply our affordability index to the Boston metropolitan area for 2005. The flexibility of our index allows us to present the indices for a set of household incomes (as a percent of area median income) and family sizes. The main source of data is the 2000 Census; we have updated prices and the housing stock to 2005 levels. The results show that the ranking of towns by affordability can change substantially when we adjust for local amenities. Finally, we discuss the policy uses and implications of our affordability index.

Yong Chen (Syracuse University), Stuart Rosenthal* (Syracuse University)

Local Amenities and Life Cycle Migration: Do People Move for Jobs or Fun?

Between 1995 and 2000, roughly 21 percent of U.S. residents moved to a different county. Migration on this scale has important implications for the local supply of labor and land markets. Using data from the 1970-2000 Censes, we investigate factors driving U.S. migration patterns. As a first step, we construct a unique panel of MSA and state-specific non-MSA quality of life and business environment measures. Based on this panel, it is clear that households tend to prefer MSAs in warm coastal areas and non-metropolitan locations, while firms prefer large, growing cities. In addition, cities with improving business environments acquire increasing shares of workers, especially workers with high levels of human capital; cities with improving consumer amenities acquire increasing shares of retirees. We next examine migration decisions at the individual level. Two dominant patterns emerge: (i) between age 20 to 35, regardless of marital status, highly-educated households move to places with high quality business environments; (ii) after age 55, regardless of education, married couples move away from places with favorable business environments and towards places with highly valued bundles of consumer amenities. Evidence also supports the idea that young “Power Couples” seek locations attractive to business in part to solve their labor market co-location problem.

Jordan Rappaport* (Federal Reserve Bank of Kansas City)

Moving to Nice Weather

U.S. residents have been moving en masse to places with nice weather. Well known is the migration towards places with warm winters, which is often attributed to the introduction of air conditioning. But people have also been moving to places with cooler, less-humid summers, which is the opposite of what is expected from the introduction of air conditioning. Nor can the movement to nice weather be primarily explained by shifting industrial composition or by elderly migration. Instead, a large portion of weather-related moves appear to be the result of an increased valuation of nice weather as a consumption amenity, probably due to broad-based rising per capita income.

Kenneth Small* (University of California at Irvine), Seiji Steimetz (California State University at Long Beach)

Spatial Hedonics and the Willingness to Pay for Residential Amenities

Housing rents and prices may be influenced not only by the characteristics of the house in question, but by those of nearby houses. Recent work has shown how this effect can be included in a hedonic housing-price equation by using a spatial autoregression model that includes “spatial lags” (prices of nearby properties) in the specification. But if there is a change that influences all prices simultaneously, such as a uniform pollution reduction, what role do spatial lags play in measuring the welfare effects arising from this change? One suggestion in the literature is that the full marginal value of an improvement in air quality is given by the reduced form equation of the autoregressive model, effectively applying a “spatial multiplier” to the directly-measured implicit price of air pollution. We show that this suggestion is correct only if the spatial price interdependence arises from technological spillovers, such that my utility depends on actions my neighbor takes as a result of that neighbor’s property value (e.g. better maintenance). In this case simply estimating a model without spatial lags may also provide a reasonably accurate welfare measure. On the other hand, the spatial inter-dependence could be from pecuniary spillovers, e.g. when real-estate agents use prices of nearby houses to estimate the true equilibrium price of a house being sold. In that case, correct benefit estimation requires the use of spatial lags to separate these pecuniary effects from the implicit prices of pollution. And because those implicit prices alone carry all the required information for benefit estimates, applying a spatial multiplier to them would overestimate benefits.

Session Urban Wages

Matthias Wrede* (RWTH Aachen University)

Distortive wage tax and commuting subsidies

Within an urban economics framework, this paper studies the implications of a wage tax which distorts workplace choices on account of commuting costs. Commuting subsidies increase welfare if and only if they shift labor supply from less to more productive business districts.

Marigee Bacolod (University of California - Irvine), Bernardo Blum* (University of Toronto), William Strange (University of Toronto)

Hard Skills, Soft Skills and Agglomeration: A Hedonic Approach to the Urban Wage Premium

This paper considers urban labor markets. In Marshall’s (1890) discussion of industrial localization, he argues that skills are fundamental. Workers can acquire better skills in a localized setting (“the secrets of the trade”), and workers with unusual skills can expect to make better use of them in such a setting as well. The same sorts of effects have been ascribed to an urbanized setting by Jacobs (1969). Knowledge spillovers and matching in urban labor markets have been modeled by Glaeser (1999) and Helsley and Strange (1990) respectively. See Fujita and Thisse (2002) and Duranton and Puga (2004) for surveys. Arguably, the most striking conclusion in the literature on urban labor markets is the existence of an urban wage premium, where labor in large cities is rewarded more highly than labor in small cities. For instance, Glaeser and Mare (2001) show that workers in larger cities are paid more, and that the premium persists when various approaches are taken to deal with unobserved worker quality. Wheaton and Lewis (2002) show that workers in cities with large numbers of workers in the same industry also receive higher wages. Other papers that have considered urban wage premiums includes Combes et al (2003), which employs panel methods to control for unobserved heterogeneity, and Rosenthal and Strange (2005), which employs geological instruments to estimate the geographic pattern of the urban wage premium. Taken as a group, these papers all find a positive relationship between agglomeration and wage. There is no consensus, however, on which sort of worker benefits most from agglomeration. It is theoretically possible that more skilled workers benefit from the many opportunities to employ their specialized abilities. It is also theoretically possible that less skilled workers benefit more from the presence of complementary skills. Wheeler (2001) finds that more educated workers enjoy a larger wage premium. Adamson et al (2004) find a “negative skill bias,” with more educated workers enjoying a smaller urban wage premium. Lee (2005) focuses on the health care sector, also finding a smaller effect for more educated workers. Rosenthal and Strange (2005), in contrast, find slightly larger urban wage and human capital effects for more educated workers. In all of these papers, the skills considered are “hard” cognitive skills, and the only sort of worker skill differentiation considered is vertical differentiation in education. In this paper, in contrast, we allow for horizontal differentiation as well, and focus on the impact of agglomeration on the hedonic prices of fundamental worker skills. To do this, we make use of the Dictionary of Occupational titles and data from the U.S. Census. We employ the indices of occupational skill requirements developed by Bacolod-Blum (2005) to characterize an occupation’s requirements for cognitive, motor, and people skills. The latter are a “soft” skill, in contrast to the hard skills that have been considered in prior work on the urban wage premium. Using these skill measures, we estimate a range of wage models, also including as regressors standard controls for worker education, family status, race, gender. The analysis reaches three principal conclusions. First, contrary to our expectations, worker skills are not very different across cities. The differences are smaller than are the differences in worker education across cities, which Berry and Glaeser (2005)) argue are themselves not very large. The differences are much smaller than are the well-known differences across cities in industrial composition. This uniformity characterizes both traded and non-traded goods sectors. It is not a consequence of a uniform distribution of occupations across city sizes; occupations are not present in approximately constant shares in cities of various sizes. The paper’s second conclusion is that worker skills are differentially impacted by urbanization. Specifically, there exists a cognitive premium that rises with city size. Similarly, there exists a similar but slightly weaker pattern for the people skills premium. In contrast, motor skills have a lower price in larger cities. Third, there are important interactions between hard cognitive skills and soft people skills. More precisely, when skill complementarities are considered, the returns to cognitive skills interacted with people skills are positive and significant, while the returns to cognitive and people skills by themselves both become insignificant. Taken as a group, these results are consistent with models of agglomeration where interaction is fundamental. It is well known that a wage premium associated with agglomeration can reflect either agglomeration economies that make workers more productive or a selection effect where the more productive workers choose agglomerations. We address this issue by making use of the National Longitudinal Survey of Youth in two ways. First, we employ the Armed Forces Qualification Test and the Rotter Index to control further for unobserved worker ability. The pattern described above continues to hold: hedonic prices of cognitive and people skills that rise with city size and significant interactions between these hard and soft skills. Second, we exploit the panel nature of the NLSY to estimate worker fixed effects and so control for the entire range of unobserved worker skills. As above, cognitive skills and the interaction between cognitive and people skills are valued to a greater degree in large cities.

Susana Iranzo* (University of Sydney), Giovanni Peri (University of California, Davis)

Social Returns to High School Education and College Education: Theory and Evidence from US states

While there is no doubt that college education has granted large and increasing private returns to workers, it is still an unresolved issue whether social returns are higher than private returns for the highly educated workers. Recently the literature on appropriate technological adoption has emphasized that the supply of highly educated workers may affect the adoption of skill-biased technologies. If new skill-biased technologies are more productive, in the sense that they are associated to higher total factor productivity (TFP), the existence of highly educated workers may have a positive externality on the other workers. The current literature on externalities of schooling in the US, however, is rather mixed and does not find, in general, a positive external effect of schooling. We propose a simple model and an empirical strategy that reconcile the conflicting finding of the previous empirical studies with the role of education in technological choice. Our model assumes that modern technologies require an investment and provide a comparative advantage to highly educated workers over less educated ones; hence they are adopted only by workers above a certain level of schooling. This implies that the private as well as the social returns to schooling are affected by the technology adopted: the adoption of more productive (modern) technology generates higher private returns as well as higher TFP. We also adopt a strategy that identifies the effect o shifts in any part of the schooling distribution on TFP. We simulate those external effects using the model and we estimate those using U.S. states data for 1960-2000. Relying on compulsory state laws and international “brain” migration we have, for US states, some exogenous shifters of average schooling and of the share of college born. Consistently in the model and in the data, we measure positive and large externalities associated to an increase in the share of college graduates but very small externalities associated to an increase in high-school graduation.

Sabrina Di Addario* (Bank of Italy), Eleonora Patacchini (La Sapienza University of Rome)

Wages and the City. Evidence from Italy

We analyze empirically the impact of urban agglomeration on Italian wages. Using micro-data from the Bank of Italy's Survey of Household Income and Wealth for the years 1995, 1998, 2000 and 2002 on more than 22,000 employees distributed in 242 randomly drawn local labor markets (30 percent of the total), we test whether the structure of wages varies with urban scale. We find that every additional 100 employees per square kilometer (100,000 inhabitants) in the local labor market raises earnings by 0.4-0.6 percent (0.1 percent) and that employees working in large cities earn, on average, 2-3 percent higher wages than those in the rest of the economy. The application of spatial data analysis techniques enables us to state that this effect is present only in the large cities surrounded by low-populated areas. We also find that urbanization does not affect returns to experience and that it reduces returns to education and to tenure with current firm, while providing a premium to managers, worker supervisors, and office workers.

* Presenter