Global Metropolitan Areas: the Natural Geographic Unit for Regional Economic Analysis
Total Page:16
File Type:pdf, Size:1020Kb
ECONOMIC & CONSUMER CREDIT ANALYTICS June 2012 MOODY’S ANALYTICS Global Metropolitan Areas: The Natural Geographic Unit for Regional Economic Analysis Prepared by Steven G. Cochrane Megan McGee Karl Zandi Managing Director Assistant Director Managing Director +610.235.5000 +610.235.5000 +610.235.5000 Global Metropolitan Areas: The Natural Geographic Unit for Regional Economic Analysis BY STEVE COCHRANE, MEGAN MCGEE AND KARL ZANDI n understanding of subnational economies must begin with a definition of a functional economic region for which data can be collected and models of economic activity can be appropriately con- Astructed. Metropolitan areas, as opposed to fixed administrative units, represent individual and uni- fied pools of labor that form cohesive economic units. Because they are defined in economic terms, metro- politan areas are better suited than other sub-national geographic units for regional economic forecasting and global comparisons. Metropolitan areas are the key geo- space and time of similar functional eco- The Natural Unit for Regional Economics graphic unit for regional economic analysis nomic units—metropolitan areas. Second, Regional economists have a wide spectrum within nations and states. Economic activity it creates a spatial definition of metropoli- of subnational functional economic regions to concentrates in metropolitan areas through tan areas that can be modeled. consider for analysis, such as states, counties interactions among businesses, people and and municipalities in the U.S. or NUTS 1,2, and governments where investors may benefit Defining Metropolitan Areas 3 and LAU 1 and 2 in Europe. Although there from large labor markets, public infrastruc- A metropolitan area is defined as a core are reasons to study subnational economics ture, and deep pools of consumers1. The prox- urbanized area and the adjacent areas tied to- at each level of geography, metropolitan areas imity of such elements further creates pro- gether by strong economic connections. They are the natural choice for modeling, forecast- ductivity that leads to economic growth and form unified labor pools that are linked by ing and global comparisons for several reasons: ultimately drives the broader national and infrastructure for daily commuting and form » They represent complete labor markets, global economy. Indeed, firms benefit by lo- service regions for consumers and businesses. enabling equilibrium-based modeling. cating in metro areas that maximize produc- A metropolitan area typically spans a number » Considering complete labor markets tivity growth and that are likely to contribute of local government authorities. leads to more robust modeling and most to global economic growth2. The Organization for Economic Coop- more accurate regional forecasting. Moody’s Analytics has developed a eration and Development (OECD) defines » Their boundaries, defined in economic unique global system to define metropoli- such an economic area as a geographical terms, mitigate the influence of local tan areas for practical regional economic space within which a number of economic political boundaries on regional eco- modeling, forecasting and analysis. This links are concentrated, most obviously nomic analysis. system creates, for the first time, a set of labor markets but also networks of firms, Historically, Christaller and Lösch’s de- consistent metropolitan area definitions important parts of supply chains, and rela- velopment of Central Place Theory nearly with global coverage for which time series tions between firms and local authorities3. a century ago first identified ways to cre- economic data is available. Thus, Moody’s Commuting is at the heart of a metropoli- ate a hierarchical structure of functional Analytics accomplishes two objectives. tan region, as it brings together firms and economic units to identify the optimal First, it accommodates comparison across workers through transport and telecommu- location of industries5. These functional nications infrastructure4. 5 See W. Christaller (1933) Die zentralen orte in Südde- 1 See Emilia Istrate, Alan Berube, and Carey Anne Nadeau, utschland. Translation: C.W. Baskin (1966) Central places in Brookings Institution Metropolitan Policy Program: Global Me- southern Germany. Prentice-Hall, Englewood Cliffs; A. Lösch troMonitor 2011, Volatility, Growth and Recovery, 2012, p. 4. 3 OECD Territorial Reviews: Competitive Cities in the Global (1941) Die raumliche ordnung der wirtshaft. Translation: 2 McKinsey Global Institute: Urban World, Mapping the Eco- Economy, p. 31, ISBN 92-64-02708-4, 2006, p.31. W.H. Woglom and W.P. Solper (1954) The economics of nomic Power of Cities, p. 1, March 2011. 4 OECD (2006) p.33. location. Yale University, New Haven. MOODY’S ANALYTICS / Copyright© 2012 1 ANALYSIS �� Global Metropolitan Areas: The Natural Geographic Unit for Regional Economic Analysis economic units were based on popula- economic statements difficult, if not impos- Eurostat’s Urban Audit8, and the OECD9 tion centers with a fixed geographic shape sible, to interpret. Analysis using metropolitan have resulted in partial definition sets that and range to describe where goods were areas enables comparison of regional econo- are instructive but lack full global cover- produced within the hierarchical system. mies because their boundaries are constructed age and do not guarantee that time-series Further, their definition depended upon in economic terms rather than political terms. data are available. ESPON and Urban input-output linkages between and among Thus the often-arbitrary influence of local ad- Audit cover only metropolitan areas in Eu- firms and households. ministrative boundaries can be mitigated from rope. Furthermore, data are not available However, contemporary regional linkages regional economic analysis. for the Large Urban Zones (LUZ) defined are arguably more commonly established via by the Urban Audit on a regular basis. commuting patterns rather than strictly by The Need for a Practical Global System OECD does have a more global footprint, industry linkages, a trend that has been accel- Achieving the benefits of analysis of met- but only for the largest metropolitan ar- erated by improved transportation systems ropolitan areas relies on the capability to sys- eas, not covering many of the secondary and rural-to-urban migration due to labor tematically define them and obtain sufficient cities that businesses and investors need productivity growth6. Commuting patterns time-series economic data. Indeed, they usu- to analyze and understand. reflect a common labor pool for a region, ally comprise a number of officially recognized as well as a common pool of consumers of units of local government such as counties, Global Metropolitan Area goods and services. A geography based on districts and cities. Thus, their construction Definition System fixed boundaries and input-output linkages may vary from country to country, their loca- Moody’s Analytics has developed a prac- does not capture these dynamics. tion and spatial boundaries will change over tical system to define metropolitan areas This combination of demographic trends time as they grow, and hence their recogni- globally for regional economic analysis by and industrial structure within a metropolitan tion as critical economic units can be ques- aggregating subnational administrative area defined by commuting patterns ac- tioned if not defined in a uniform way. units for which economic data exist (see commodates the estimation of an economic A practical system to specify global met- Charts 1 and 2). These definitions specify model that can simulate their interaction and ropolitan area definitions for regional eco- that geographic areas contain, to the extent growth. Within such a complete labor mar- nomic analysis must satisfy several criteria: possible, all economic activity associated ket, a model can be developed to endogenize » The defined metropolitan areas must with each functional metropolitan area and employment, income and demographic be, or closely approximate, cohesive no more. Constructed in this manner, these trends in the context of local agglomeration economic units representing com- definitions guarantee data availability and economies within the area and trade patterns plete labor markets. mitigate biases that could be created by the with external economies. These more robust » Time series economic data for the peculiarities of a country’s subnational ad- models, in turn, facilitate more accurate fore- areas constituting each metropolitan ministrative structure. casting of subnational economies. area must be available. The process to specify the subnational Contemporary functional economic » The system must be capable of be- areas constituting each global metropolitan units—metropolitan areas—are defined in a ing applied globally to countries for area begins with a review of the geographic bottom-up way based on population centers which sub-national economic data and administrative structure of each na- and commuting patterns and can change is available. tion for which metropolitan areas are to over time, staying current as commuting » The system must be dynamic to ac- be defined. Chart 3 illustrates the decision patterns change. In contrast, Christaller and commodate changes in metropolitan process by which Moody’s Analytics con- Lösch’s earlier fixed geographic definition boundaries and subnational geo- structs each metropolitan area based on of functional economic units reflected the graphic