In lieu of an abstract, here is a brief excerpt of the content:

. Hodgson N O T E S Chapter 2. The Growth of Metropolitan Areas 1. The few summary statistics given in this chapter all refer to metropolitan areas. This is also the geography used by most of the studies surveyed below. For policy purposes , cities might be viewed as more relevant given their greater autonomy. However, municipalities seldom form self-contained units of analysis and are extremely heterogeneous . Some metropolitan areas are clearly dominated by one city, whereas others are highly fragmented. 2. For very different reasons (e.g., better access to markets), one may also expect roads external to cities, and more generally market access, to affect urban growth. Redding and Sturm (2008) use the division of Germany after World War II as a natural experiment to show that cities close to the Iron Curtain declined after it was imposed in Western Germany. 3. Obtaining experimental evidence would of course be better. Unfortunately, true experimental evidence on this kind of topic would be hard to obtain. It is difficult to imagine that politicians would accept a random allocation of highways. “Natural experiments” (real-life events that would recreate experimental evidence) are also extremely scarce (the reference mentioned in the previous note notwithstanding). As a result, using the type of approach described here and leveraging sources of exogenous variation in the data to mimic indirectly experimental conditions is often the best that is available to the analysts. 4. See Angrist and Pischke (2008) for a more detailed exposition of this type of approach. 5. Jacobs (1969) also forcefully defended very similar ideas insisting on the importance of cross-fertilization between sectors. 6. This “Chinitz” finding about the importance of many small establishments is confirmed by Rosenthal and Strange (2010), who suggest that small establishments matter because they provide a greater diversity of specialized suppliers to local firms. This would be consistent with the explanation developed above that highlights the dynamic implications of static externalities. 262 Notes to Pages 46–65 Chapter 3. The Relationship Between City Center Density and Urban Growth or Decline 1. For recent reviews of this literature, see Duranton and Puga (2004); Rosenthal and Strange (2004); and Puga (2010). 2. Rosenthal and Strange (2003); van Soest, Gerking, and van Oort (2006); Fu (2007); and Arzaghi and Henderson (2008) are examples of this work. Elvery and Sveikauskas (2010) find the strongest agglomeration effects at longer distances (ten, twenty, or twenty-five miles) but also show that short distance effects (within two-and-a-half miles) tend to be stronger when the workforce is more educated and belonging to similar occupational categories, suggesting the importance of the exchange of ideas for shortdistance agglomerative effects. 3. Available at http://www.census.gov/geo/tiger/cbdct.pdf. 4. Available at http://www.s4.brown.edu/us2010/Researcher/Bridging.htm. 5. See Rosenthal and Strange (2004). 6. See Glaeser and Saiz (2004). 7. LeRoy and Sonstelie (1983) show how a pattern of high-income people moving back to the CBD from the suburbs could occur when modes of transportation such as the car are adopted first by high-income and then by low-income people. Lin (2002) provides empirical support for this hypothesis. Brueckner, Thisse, and Zenou (1999) posit that variation in amenity levels may explain variation across cities in the degree to which high-income households tend to be concentrated in the suburbs versus near the CBD. 8. For the curious reader, two-stage least squares results from the specification noted above with an F of 10.37 yield a coefficient on change in population density near the CBD of 0.527 and a standard error of 0.182. Chapter 4. Central Cities and Metropolitan Areas: Manufacturing and Nonmanufacturing Employment as Drivers of Growth 1. The U.S. economy has long been concentrated in metropolitan areas, and that concentration is increasing. In 1970, metro areas accounted for 80 percent of the U.S. population. This share grew to 84 percent in 2001, with the rate of growth accelerating over the past decade. Similarly, the metro area share of total nonfarm payroll employment in the U.S. has risen from 83 to 86 percent. 2. Metropolitan Statistical Areas (MSA) comprise counties or county-equivalent areas and have at least one urbanized area of 50,000 or more population, plus adjacent territory, also defined by counties, that have a high degree of social and economic integration with the core as measured by commuting ties (OMB 2009). Some metropolitan areas have more than one principal...

Share