Lecture 11: Why Is Development Spatially Concentrated? PDF

MODULE 4: WHY IS DEVELOPMENT SPATIALLY CONCENTRATED? PART 4-A: URBANIZATION PART 4-B: SCALE ECONOMIES PART 4-C: THE BENEFITS OF TRENDS AND TRANSPORT COSTS PROXIMITY IN COMMERCIAL AND INDUSTRIAL REAL ESTATE...

MODULE 4: WHY IS DEVELOPMENT SPATIALLY CONCENTRATED? PART 4-A: URBANIZATION PART 4-B: SCALE ECONOMIES PART 4-C: THE BENEFITS OF TRENDS AND TRANSPORT COSTS PROXIMITY IN COMMERCIAL AND INDUSTRIAL REAL ESTATE ❑ 1 Same question: Why is there so much variation in demand for real estate across cities and neighborhoods? PART 4-A: URBANIZATION PART 4-B: SCALE ECONOMIES PART 4-C: THE BENEFITS OF TRENDS AND TRANSPORT COSTS PROXIMITY IN COMMERCIAL AND INDUSTRIAL REAL ESTATE ❑ 2 The Paradox of Urbanization and the Future of Urban Real Estate Markets ◼ Despite dramatic declines in the cost of moving people and goods, and technology that has virtually eliminated communication costs from afar, cities have never played a larger role in the world economy than they do today. Why? ◼ Will cities continue to thrive now that risks from COVID-19 have receded? ◼ Or will widespread adoption of remote work among information-oriented workers persist? ◼ If yes, how will that affect cities and especially central city commercial real estate? ◼ Is this a good time to invest in downtown real estate? 3  The Decline of the Costs of Moving Goods.185063 Dollars per Ton Mile (Real).02323 1890 2000 year Railroad Revenue per Ton Mile 4 Source: Triumph of the City, Edward Glaeser, Penguin Press, 2011. Until Recently – About 100 Years Ago – Cities Were Unhealthy Places Survival in 19th Century Cities: The Larger the City, the Smaller Your Chances Source: Cain L, Hong SC. Survival in 19th Century Cities: The Larger the City, the Smaller Your Chances. Explorations in Economic History. 2009 Oct 1;46(4):450-463. doi: 10.1016/j.eeh.2009.05.001. PMID: 20161075; PMCID: PMC2743429. The Urban Mortality Transition in the United States, 1800-1940 “I the United States in the 19th and early 20th centuries, there was a substantial mortality “penalty” to living in urban places. This circumstance was shared with other nations. By around 1940, this penalty had been largely eliminated, and it was , c s s, sd c c s d.” Source: Haines, Michael R., 2001. The Urban Mortality Transition in the United States, 1800-1940, NBER Historical Working Papers 0134, National Bureau of Economic Research, Inc. 5  Sewer Systems and Clean Water Lower Urban Mortality Rates – Contributing to Urban Growth 6 Source: Triumph of the City, Edward Glaeser, Penguin Press, 2011.  1,500m Spatial Concentration ECN 745 of Business Activity, Productive Cities, and Tall Urban Areas are Booming! Commercial Buildings Stuart S Rosenthal Maxwell Advisory Board Professor Stuart S Rosenthal Syracuse University Maxwell Advisory Board Professor Syracuse University Sky Mile Tower Japan 1,700m 7 7  Ub sb c … Most economic activity now takes place in cities North America HUGE growth of cities in China World 61% of population in 2020 5 d’s 10 China tallest buildings are in China Source: World Bank https://data.worldbank.org/indicator/SP.URB.TOTL.IN.ZS 8  Urbanization is still increasing! Rapid growth of urbanization in Africa and Asia Still rising in Europe, North America, and Latin America Source: United Nations, Department of Economic and Social 9 Affairs, Population Division (2018). World Urbanization Prospects: The 2018 Revision, Highlights  Higher Income Countries Are More Heavily Urbanized 10 Source: United Nations, Department of Economic and Social Affairs, Population Division (2018). World Urbanization Prospects: The 2018 Revision, Highlights  The Urbanization Gap Between High- and Lower-Income Countries is Narrowing 11 Source: United Nations, Department of Economic and Social Affairs, Population Division (2018). World Urbanization Prospects: The 2018 Revision, Highlights Where are the Most Heavily Urbanized Areas? 12 Source: United Nations, Department of Economic and Social Affairs, Population Division (2018). World Urbanization Prospects: The 2018 Revision, Highlights  The World has been experiencing a boom in tall building construction ◼ From the Council on Tall Buildings and Urban Habitat (CTBUH)  2023 CTBUH YEAR IN REVIEW Year in Review 2023 - The Skyscraper Center  2021 CTBUH YEAR IN REVIEW Year in Review 2021 - Data - The Skyscraper Center 13 CTBUH YEAR IN REVIEW: TALL TRENDS OF 2021 The Council on Tall Buildings and Urban Habitat's (CTBUH) global study of 200-meter-and-taller building completions for 2021. (http://www.skyscrapercenter.com/year-in-review/2021/data) 14  200 Meter-plus Building Completions in 2021 by Country CTBUH YEAR IN REVIEW: TALL TRENDS OF 2021 The Council on Tall Buildings and Urban Habitat's (CTBUH) global study of 200-meter-and-taller building completions for 2020. 15 (http://www.skyscrapercenter.com/year-in-review/2021/data)  Among Tall Buildings Mixed Use Is Becoming More Common Huge Increase in Mixed-Use CTBUH YEAR IN REVIEW: TALL TRENDS OF 2021 The Council on Tall Buildings and Urban Habitat's (CTBUH) global study of 200-meter-and-taller building completions for 2021. 16 (http://www.skyscrapercenter.com/year-in-review/2021/data) Most people in the United State live in cities ◼ Over 80 percent of people in the U.S. live in urban areas. ◼ Cities occupy only roughly 2 percent of the land area of the U.S. ◼ Why is economic activity so concentrated into urban areas? ◼ A puzzle because cities are expensive! 17 Cities have high office rents! Rent USD sqft/yr (2022) Hong Kong: $259 New York, Midtown: $220 London: $182 Beijing Finance Street: $167 Silicon Valley: $149 Beijing, CBD: $140 New York, Midtown South: $132 Shanghai, Pudong: $128 Beijing, Zhongguancun: $115 San Francisco: $114 Singapore: $110 Hong Kong East: $109 Shenzhen: $108 Source: JLL Global Premium Office Rent Tracker 18  Cities also have high wages! Source: Glaeser and Mare (2001), “Cities and Skills” Journal of Labor Economics. 19 19 Cities are expensive places to do business! Why would for-profit firms locate where the cost of space and labor is so high? Cities must be productive (valuable) locations! Companies are drawn to cities and that increases demand for city center office space 20  Why are cities productive places? ◼ Understanding why cities are productive places ps s … ❑ Anticipate where entrepreneurs may want to locate their businesses ❑ Refine local government policies so as to attract (or at least not deter) new businesses ❑ Operate buildings in ways that enhance productivity and profit ❑ Identify opportunities and locations for successful development. 21  Why would households put up with the cost of living in cities? ◼ Cities have higher levels of congestion, crime, pollution and other attributes associated with crowding together. ◼ Cities must provide households with something to offset these things. ❑ Higher real wages must compensate for congestion ❑ Urban amenities must compensate for congestion 22 Where do industries and companies choose to locate? ◼ This affects local demand for real estate. ◼ It is the same question as asking where should you locate your business and why? ◼ For now, we’ll mostly focus on export industries (e.g. manufacturing) as opposed to local services like grocery stores ◼ Pictures tell a story! 1  Wine Production is Concentrated in Select Locations Why? California and the Finger Lakes region have good soil and climate for grapes. Grapes are fragile so produce wine close to where grapes are grown Source: “The Micro-Empirics of Agglomeration Economies” (Rosenthal and Strange) 2  Computer Software Development is Concentrated in Select Locations Why? Seattle (Microsoft) Silicon Valley Boston Route 128 Austin Texas Salt Lake Utah It must be something different from wine making Source: “The Micro-Empirics of Agglomeration Economies” (Rosenthal and Strange) 3  Carpet Production is Concentrated in Dalton Georgia Why Dalton? 90% of wall-to-wall carpet worldwide is produced within 25 miles of Dalton. See the History of Carpet making for more details. The reason must differ from wine making Source: “The Micro-Empirics of Agglomeration Economies” (Rosenthal and Strange) 4  Furniture Production is Concentrated in High Point North Carolina Why High Point? Following World War II, an estimated 60% of all furniture made in America was produced within a 150-mile radius of High Point See Furniture History from the High Point Museum for more details Once again, something different is going on Source: “Evidence on the Nature and Sources of Agglomeration Economies” (Rosenthal and Strange) 5  Where Will Entrepreneurs Locate Their companies? ◼ This is really two questions, or maybe three …  Why do firms within an industry tend to cluster together?  Where and why do industry clusters locate? 6  Industry Clusters ◼ Two primary factors cause firms within an industry to cluster together  Transportation costs and natural advantages  Agglomeration economies (external economies of scale) ◼ Let’s chat about transportation costs and natural advantages first … 7  Industry Clusters ◼ We’ll group export industries into two categories:  Production cost oriented  Transportation cost oriented 8  Where do industry clusters locate? ◼ Production cost oriented ◼ Firms whose total costs are especially sensitive to spatial variation in the cost of factor inputs used in production ◼ Labor (wages) ◼ Land (rent/sq foot) ◼ Other local inputs (e.g. electricity; water) ◼ The cost of these factors varies across cities and also within cities ◼ Production cost oriented industries choose location to minimize production costs 9  Where do industry clusters locate? ◼ Transportation cost (materials) oriented ◼ Firms whose total costs are especially sensitive to shipping costs that differ with a company’s location  Cost of shipping inputs  Cost of shipping products ◼ These industries choose locations to minimize transportation costs 10  Transportation cost oriented industries ◼ Suppose there is one input source and one market ◼ Would a transport cost oriented firm locate …  At the input source?  At the market?  Somewhere between the input source and market? 11 Where does a transportation cost oriented firm locate its factory? Total transport costs = PC + DC Cost of Cost of shipping shipping product input (DC) (PC) Slope = cost of shipping input per unit distance (procurement cost, PC) Input Source Market Factory location Slope = cost of shipping product per (distance) unit distance (distribution cost, DC) 12  Transportation cost oriented industries ◼ Because unit shipping costs differ for inputs and outputs, transport oriented industries …  Locate at the market … ◼ If it is cheaper to ship the input than the product ◼ e.g. Broadway theaters, major hospitals ◼ These industries have a Market Orientation  Locate at the input source … ◼ If it is cheaper to ship the product than the input ◼ e.g. paper mills, wine production ◼ These industries have a Resource (Input) Orientation ◼ Industries would generally not locate between the input source and the market 13  Transportation cost oriented industries ◼ The picture a few slides back assumes transportation costs per unit distance are constant returns to scale (we drew straight lines) ◼ How would EOS in distance shipped affect the picture and our conclusions?  Suppose (to simplify) the cost of shipping inputs X miles = the cost of shipping the product X miles.  With EOS in distance shipped, would the firm be more or less likely to locate the factory at either the input source or the market? Why? 14  Transportation cost oriented industries Cost of Cost of shipping shipping product input With EOS in distance shipped, shipping costs increase at a declining rate with distance → concave curves Input Market Factory location Source (distance) With EOS in distance shipped, would the firm be more or less likely to locate the factory at either the input source or the market? Why? 15  Transportation cost oriented industries Total transport costs Cost of Cost of shipping shipping output input With EOS in distance shipped, shipping costs increase at a declining rate with distance → concave curves Input Market Factory location Source (distance) ANSWER: The firm is even more likely to locate at either the input source or the market because one long trip is less expensive than two short trips when there are EOS in distance shipped (e.g. less loading and unloading) 16  Transportation cost oriented industries ◼ What if there are multiple input sources and markets? ◼ Suppose five customers along a line, A through E  A-------B-------C-------D-----------E  One separate delivery to each customer each day.  Price and production costs are not affected by the factory’s location.  Where should you locate the factory to minimize total distance traveled? 17  Transportation cost oriented industries ◼ A-------B-------C-------D-----------E ◼ Are you better off at B than A. Why??  Yes! Moving 1 step from A towards B lowers travel distance 1 step to B, C, D, and E, while increasing distance by 1 step to A. ◼ Total distance traveled falls by 3 steps ◼ Same moving from B to C ◼ But, from C to D, total distance increases ◼ So, locate at the median site, C.  Moving away from the median increases distance to the majority of customers  This holds regardless of distance between customers! ◼ This is the Principle of Median Location 18  Transportation cost oriented industries ◼ What if your markets and input sources are not located along a line? ◼ Does the principle of median location help to explain where you might want to locate your company – i.e. why cities form? 19  Transportation cost oriented industries ◼ Suppose customers are in cities A, B, and C A: 100 Customers B: 100 Customers C: 500 Customers ◼ Locate factory in City C (NY, Chicago, LA) ◼ Largest city often contains  > 50 percent of the local market  median customer 20  Implications: Different types of cities ◼ Resource orientation ➔ Cities built around input sources  Fishing villages  Lumber towns  Mining towns ◼ Market orientation ➔ Regional and national market centers (often the median location for customers)  New York  Chicago  Los Angeles 21 Implications: Many small cities and few very large cities ◼ Median location cities draw in additional market- oriented companies causing the city to become especially large compared to other cities ◼ Other economic forces …  Reinforce the tendency for countries to have large numbers of small cities and relatively few very large cities  And ensure that large cities typically have diverse economies with many industries present 22 Large Cities Have Diverse Economies The role of … ◼ Per capita demand ◼ Internal economies of scale → A firm’s ability to reduce average cost when it produces at larger scale 23  Large cities have diverse economies Small EOS – High Per Few customers Capita needed to get to Avg. Demand the MES Low Per Large EOS – Many Cost Capita customers needed Demand to get to the MES Gasoline Stations Specialized Hospitals Minimum Efficient Minimum Efficient Quantity Scale for Gasoline Scale for Specialized Stations Hospitals 24  Large cities have diverse economies ◼ If EOS are exhausted by small populations (per capita demand high) such industries will be located in many cities. (e.g. gas stations) ◼ If EOS are exhausted only by a large population (PCD low), a region may only support 1 such enterprise. (e.g. certain types of hospitals)  This model is sometimes referred to as the Central Place Theory.  Central place theory implies that larger cities have more diverse economies 25  Large cities have diverse economies ◼ Economic diversity in large cities is self-reinforcing and can attract additional industry to the area, causing the city to grow further. ◼ Local economic diversity contributes to positive productivity spillovers (agglomeration economies)  Easier to hire workers with specific skills  Opportunities to gain knowledge from nearby companies  Opportunities to support specialized input providers, reducing the need to produce intermediate inputs in-house (e.g. manufactured buttons for fashion design or specialized legal services) 26 Large cities typically have diverse, productive, economies Local economic diversity contributes to agglomeration economies that make companies more productive. This makes large cities especially valuable for many companies. That in turn attracts additional industry to very large cities, causing them to grow further, and increasing large-city real estate values! 27  How does the declining importance of transport costs affect where industries locate and cities form? ◼ We’ve noted that the importance of transport costs has fallen dramatically in the last 100 years.  Reduced shipping costs (trucks, faster ships, container technology, planes …)  Efficient manufacturing reduces the quantity of materials needed ◼ As an example, it requires much less iron ore today to produce 1 ton of steel than it did in 1900.  Our economies have become very dependent on service and information-oriented industries → These are much less sensitive to transport costs compared to manufacturing. 1  How does the declining importance of transport costs affect where industries locate and cities form? ◼ With increasing mobility …  Firms → move to where labor is cheap  Workers → move to where wages are high ◼ Wage differences across regions should diminish as local supply and demand for labor shifts 2  How does the declining importance of transport costs affect where industries locate and cities form? ◼ Consider two regions that have identical attributes except different current supply and demand for labor. ◼ Suppose initially that workers in Region 1 earn higher wages than their counterparts in Region 2. ◼ As time passes, what happens to wage rates in the two regions? 3  Regional convergence in the cost of labor Region 1: Initially Region 2: Initially Wage Low Wage (at A) Wage High Wage (at C) SC WC C SA DC WA A DA Employment Employment Workers have incentive to migrate from Region 1 to 2 drawn by higher wage. 4 Firms have incentive to shift from Region 2 to 1 drawn by lower wage.  Regional convergence in the cost of labor Region 1: Initially Region 2: Initially Wage Low Wage (at A) Wage High Wage (at C) SC C SB SD B SA D DC A DD DB DA Employment Employment Labor demand shifts out in Region 1 while Labor supply shifts in. The reverse occurs in Region 2. Absent other differences between regions, this continues until wages are about equal at B and D ensuring a spatial 5 equilibrium.  Regional convergence in the cost of labor Regional per Capita Income, 1880 to 1985 (Source: Mills and Hamilton, Urban Economics, 4th Ed, pg. 43; Statistisal Abstract of the United States) 250 Percentage of U.S. Average Income 200 150 100 50 0 1880 1900 1920 1930 1940 1950 1960 1970 1980 1985 Year United States South Atlantic Northeast Pacific Wage convergence in the United States between 1880 and 1980 → As predicted by the model in the previous slides. 6  What will determine industry locations in the future? ◼ With reduced transport costs industry locations become …  Less sensitive to costs associated with shipping material and/or people.  AND less sensitive to regional differences in wages which diminish. ◼ Could future cities simply form anywhere? ◼ Or will some locations still be more prone to economic development? 7  What will determine industry locations in the future? ◼ Don’t count transport costs out yet! Access to transport hubs remains important (e.g. harbors; major airports). Just not as important as before. ◼ Location-specific agglomeration economies for a given industry (e.g. the financial district in Manhattan).  Can any town create its own tech-center like Silicon Valley?  Mostly no! We’ll talk about this soon. ◼ Local amenities that appeal to workers as with sunshine, warm weather in January, scenic views. ➔ Will firms increasingly move to where workers want to live? ➔ Will work from home and COVID-19 accelerate this trend? 8 How Close Does Your Company The Spatial Reach Need to Be to Benefit from of Agglomeration Economies Nearby Economic Activity? 1 The focus in these slides and the following slide deck … ◼ We will review evidence on how quickly agglomeration economies diminish with distance between companies ◼ This is especially relevant for production cost-oriented industries: Industries in which companies choose location primarily to reduce production cost (as opposed to shipping costs) ◼ Remember, agglomeration economies refer to productivity gains from operating close to other companies. 2 The focus in these slides and the following slide deck … ◼ Much of the material in this slide deck is from: How Close is Close? The Spatial Reach of Agglomeration Economies ◼ This will include images of spatial patterns of employment for the northeast of the US in 2018 based on 8.9 million establishments and 56 million workers (from Dun & Bradstreet). 3 Attenuation of agglomeration economies affects spatial patterns of demand for real estate ◼ If companies are more productive (profitable) when other establishments are nearby, that will contribute to spatial concentration of economic activity. ◼ Valuable proximity to other companies will increase demand for nearby real estate (e.g. commercial office space) ◼ That in turn will increase local real estate values and lease rates The rate at which agglomeration economies attenuate with distance between companies will affect spatial patterns of development and real estate values We’ll consider agglomeration at different levels ◼ Increasingly narrow levels of geography ❑ Regional ❑ Metropolitan ❑ Neighborhood ❑ Building ❑ Within building ◼ Evidence suggests that productivity spillovers (agglomeration economies) occur at all of these levels but other mechanisms (e.g. transport costs) also contribute to agglomeration. ◼ The list of studies on the next few slides provide evidence. Focus on the patterns highlighted in brown and blue … 6 Geographic scope and attenuation of spillovers ◼ Region and Metropolitan Level ❑ “Determinants of Agglomeration” Rosenthal-Strange, JUE, 2001 → Up to state level ❑ “The Influence of State Policy and Proximity to Medical Services on Health Outcomes,” Li, JUE, 2014→ 25 miles ❑ “Geographic Localization of Knowledge Spillovers as Evidenced by Patent Citations”, Jaffe, Trajtenberg, Henderson, QJE, 1993 → MSA level in patent citations ❑ “The Effect of High-Tech Clusters on the Productivity of Top Inventors.” Moretti, AER, 2021 → MSA level in patent production ❑ “The Agglomeration of R&D Labs,” Buzard et al, JUE, 2017 → Up to MSA level ❑ “Novel Ideas: The Effects of Carnegie Libraries on Innovative Activities”, Berkes and Nencka, 2019 → About 20 miles Geographic scope and attenuation of spillovers ◼ Region and Metropolitan Level ❑ “Urbanization, productivity, and innovation: Evidence from investment in higher education.” Andersson, Quigley, Wilhelmson, JUE, 2008 → Roughly 10 miles ❑ “The Attenuation of Human Capital Spillovers” Rosenthal-Strange, JUE, 2008 → 5 miles Geographic scope and attenuation of spillovers ◼ Neighborhood Level ❑ “Geography, Industrial Organization and Agglomeration” Rosenthal- Strange, ReStat, 2003 → Within 1 mile ❑ “Attenuation of Agglomeration Economies: Evidence From the Universe of Chinese Manufacturing Firms” Li, Li, Liu, JUE, 2022 → Within 1 mile ❑ “The economics of density: Evidence from the Berlin Wall.” Ahlfeldt, Redding, Sturm, Wolf, Econometrica, 2015 → Within 10 minutes travel time ❑ “The Geography of Entrepreneurship in the NY Metro Area”, Rosenthal- Strange, NY Fed Journal, 2005 → a few city blocks ❑ “Networking off Madison Avenue” Arzaghi and Henderson, ReStud, 2008 → a few city blocks Geographic scope and attenuation of spillovers ◼ Building Level ❑ “Agglomeration Economies and the Built Environment: Evidence from Specialized Buildings and Anchor Tenants” Liu, Rosenthal and Strange, JUE 2024 → Within individual commercial buildings ❑ “Employment Density and Agglomeration Economies in Tall Buildings” Liu, Rosenthal and Strange, RSUE, 2020 → Within three or four floors in commercial buildings ◼ Office and Team Level (peer effects) ❑ “Sorting and Agglomeration Economies in French Economics Departments,” Bosquet and Combes, JUE, 2017 → Within academic departments ❑ “Peers at Work,” Mas and Moretti, AER, 2009 → Within shifts in grocery stores Consider some stylized facts … It is well known that regional employment is highly concentrated in select urban areas … 11  Figure 1: Aggregate Employment Within 2 Miles (All values are smoothed out to 10 miles with inverse exponential distance weighting) Syracuse Albany Boston New York Buffalo Philadelphia Washington DC 12 Employment is also more concentrated than productivity. 13 Employment is more concentrated than productivity ◼ New York MSA (20,300,000) is 23 times larger than Albany MSA (880,091) ◼ Doubling city size increases labor productivity and related wage rates by 2 to 5 percent ❑ Rosenthal and Strange (2004) ❑ Combes and Gobillon (2015) ◼ Many rural areas have high productivity 14  Figure 1: Single-Site Average Sale/Worker Within 2 Miles (All values are smoothed out to 10 miles with inverse exponential distance weighting) Albany Buffalo Syracuse Coastal Maine Boston 15 Why is employment more spatially concentrated than productivity? ◼ With competitive markets, even relatively modest local advantages will tend to draw companies to more advantageous locations ◼ That tendency amplifies and reinforces tendency for spatial variation in demand for real estate and related patterns of urbanization 16 Don’t forget, cities are expensive places … ◼ Doubling zipcode employment increases commercial rent by roughly 10.5% (Liu, Rosenthal and Strange, 2018) ◼ Doubling nearby employment (at roughly the city level) increases wage by 2 to 5% 17 Cities have high office rents! Rent USD sqft/yr (2022) Hong Kong: $259 New York, Midtown: $220 London: $182 Beijing Finance Street: $167 Silicon Valley: $149 Beijing, CBD: $140 New York, Midtown South: $132 Shanghai, Pudong: $128 Beijing, Zhongguancun: $115 San Francisco: $114 Singapore: $110 Hong Kong East: $109 Shenzhen: $108 Source: JLL Global Premium Office Rent Tracker 18 Why would firms locate where the cost of space and labor is so high? 19 Cities must be unusually productive places! Otherwise entrepreneurs would locate in less densely developed, less expensive areas. 20 Why are cities unusually productive? ◼ Local natural advantages may enhance productivity → This is related to why transport-cost oriented companies cluster together (Ellison, Glaeser and Kerr, 2011). ◼ Talented individuals may also migrate to cities because they prefer urban amenities or because they benefit more from the large labor market. ◼ Density (agglomeration Economies) may cause workers to be more productive (Rosenthal and Strange (2004) and Combes and Gobillon (2015) reviews) 21 Policy options for how to make a city more productive differ depending on WHY cities are productive ◼ Examples … ❑ If intrinsically talented people want to live in a city → Mayor has incentives to enact policies that encourage high-skill migration to the city ❑ If density causes workers to be more productive → Mayors have incentives to enact local zoning laws that encourage advantageous concentrations of business activity (e.g. research parks, industrial parks). ◼ Because policy implications differ depending on why cities tend to be productive, it is helpful to confirm why cities are productive. But that is not so easy to do. 22 Hard to measure causal productivity spillovers from cities ◼ Cities can be productive because productive/skilled workers may migrate into cities because … ❑ They prefer urban amenities (e.g. restaurants, theater) ❑ They have more to gain from higher urban wages. ◼ There is a potential two-directional relationship. ❑ Large cities cause workers to be more productive. ❑ But productive workers cause cities to grow and become large. ❑ This bi-directional relationship also makes it difficult to confirm the causal effect of density on labor productivity. 23 Hard to measure productivity spillovers from cities ◼ Several different types of statistical methods have been used to determine the causal effects of density and cities on an individual’s productivity. ◼ We will skip over such details although I’m happy to chat with you about various methods for those who are interested (see the appendix to this slide deck for references). ◼ In the pictures that follow we will adopt an increasingly narrow geographic focus. ◼ At a sufficiently narrow level of geography … ❑ Differences in natural advantages or appeal of residential amenities would go away. ❑ Localized benefits from operating near other valued companies remain and provide evidence of agglomeration economies. 24 Implications of agglomeration economies for geographic attenuation patterns ◼ Spatial concentration enhances productivity → This implies that proximity is important → Which implies that productivity spillovers attenuate with distance. → And that the strongest spillover effects tend be those generated by nearby activity 25 The attenuation and spatial reach of productivity spillovers matters. 26 Spatial reach of productivity spillovers affects … ◼ Where entrepreneurs may want to locate their businesses ◼ The potential for local government to attract new business ❑ Could subsidies attract highly productive anchor companies whose presence prompts nearby unsubsidized growth That is why shopping mall owners charge anchor stores lower rent (Brueckner, 1993; Pashigian and Gould, 1998). ❑ Can zoning enhance productivity by ensuring advantageous spatial clusters of activity? ◼ Can managers of commercial buildings enhance local real estate values by ensuring complementary tenants on a street or in individual buildings? 27 How close must companies be to benefit from nearby activity? ◼ The answer depends on why companies benefit from operating close to their neighbors 28 How close must companies be to benefit from their neighbors? ◼ Input sharing for physical products ❑ Requires shipping by truck/train and may extend throughout a region (see Remco Button Company) ◼ Drawing on a common pool of skilled workers ❑ Commuting extends over MSA-level distances while job market networks operate down to the neighborhood level (e.g. Baer et al, 2008; Hellerstein et al, 2011, 2014) ◼ Knowledge spillovers and learning from nearby workers ❑ Sensitive to face-to-face interactions. These benefits attenuate quickly and may operate even within buildings 29 Industries differ in how close they need to be to benefit from their neighbors ◼ Manufacturing may be especially sensitive to input sharing → Clearly relevant up to region-level concentration ◼ Industries that rely on specific skills benefit from nearby companies that require similar skills → Neigh-level labor networks up to metro-level commuting. ◼ The service and technology sectors are sensitive to knowledge exchange → Neigh/building-level interactions. 30 Additional plots indicate i. Different spatial patterns between industries ii. Region-level spatial concentration is mirrored at lower levels of geography 31  Figure 2: Employment Within 2 Miles For Select Industries (All values are smoothed out to 10 miles with inverse exponential distance weighting) Panel A: All Industries Panel B: Manufacturing Panel C: Finance Panel D: R&D 32  Figure 3: Employment Within 0.05 Miles in the Five Boroughs of New York City (All values are smoothed out to 0.1 miles with inverse exponential distance weighting) Panel A: All Industries Panel B: Services Panel C: Manufacturing Panel D: Finance 33 These images display spatial concentration down to the neighborhood level ◼ Close proximity appears to be valuable. ◼ This is consistent with the idea that learning from neighbors (and possibly sharing of skilled labor) contributes to productivity. ◼ Close proximity has likely become even more important for cities today than ever before (but will that remain true with work-from home?). 34 The office sector is large and growing in importance ◼ City centers are now dominated by business services, retail and other information sensitive activities (www.bls.gov) ◼ 1950 U.S. non-farm employment share … ❑ Manufacturing 30% ❑ Business services 6.5% ◼ 2016 U.S. non-farm employment share … ❑ Manufacturing 8.5% ❑ Business services 14.1% 35 The office sector is large and growing in importance ◼ This has increased the importance of information exchange and face-to-face interactions. ◼ This has also increased the role and importance of tall commercial buildings in which business services often operate. 36 Tall buildings are rapidly growing in numbers ◼ World Trade Center’s twin towers in total had roughly 50,000 workers in 10 million square feet of space ◼ Tall buildings are equivalent in scale to small cities. 37 Buildings are important spatial units ◼ Suppose that productivity spillovers are stronger within as opposed to across buildings. ◼ Why? 38 Why would buildings affect interactions? ◼ Convenience ❑ Meeting with individuals inside one’s own building entails walking down the hall or at most an elevator ride. ❑ Meeting elsewhere may require dressing for weather and navigating city traffic. ◼ Spontaneous (valuable) interactions ❑ Enter and exit through the same doorway ❑ Ride the same elevators ❑ Have lunch in the same ground-floor restaurants. 39 The tendency to interact more with others within versus outside of your building is … ◼ Well-known in the commercial sector ◼ Familiar to many of us in information-oriented industries like academia ◼ Central to architectural design (see The Architecture of Social Interaction in ArchDaily) that seeks to enhance the quality of social interactions 40 Buildings as important spatial units ◼ Recall also that buildings are managed for profit. ❑ Building managers have the ability and incentive to shape the mix of tenants in the building. ❑ This should make the building more desirable and should increase rent ◼ Example: ❑ The financial district in Manhattan is famous as a neighborhood that has a high concentration of finance. ❑ But most of the financial sector is concentrated in a small number of buildings → This suggest that many tenants care about who else is in the building → Your building matters! ❑ See the next slide for a picture. 41 Figure 4: Finance Share of Employment (SIC 62, 67) In Manhattan’s Financial District Most buildings in New Yorks’s financial district contain no companies in securities trading and investment → Buildings are specialized! 42 Appendix: Hard to measure productivity spillovers Various statistical approaches have been used to measure the causal effect of agglomeration on labor productivity. ◼ Lagged controls (e.g. Glaeser et al, 1992; Henderson et al 1995) ◼ Instrumental variables (e.g. Rosenthal and Strange, 2008) ◼ Person fixed effects (Glaeser & Mare, 2001; De la Roca and Puga, 2017; Eckert et al, 2022) ◼ Structural methods (e.g. Baum-Snow and Pavan, 2011; Gaubert, 2018) ◼ Natural experiments (e.g. Greenstone et al, 2010; Ahlfeldt et al, 2015) ◼ Narrow geographic focus (e.g. Arzaghi and Henderson, 2008; Rosenthal and Strange, 2003; Liu, Rosenthal and Strange, 2020 & 2022) ◼ Shape of the factor return distribution (e.g. Duranton et al, 2012; Jales, Jiang and Rosenthal, 2022) 43 Two questions that are linked. How close must companies be to benefit from their neighbors? Will work from home threaten office markets? 1 How close? ◼ Input sharing of manufactured products ❑ Often requires shipping products by truck ❑ Extends over distances throughout a region ◼ Drawing on a common pool of skilled workers ❑ Commuting extends over MSA-level distances ❑ Job market networks operate down to the neighborhood level ◼ Learning from nearby workers & networking (office/info sector) in addition to shopping externalities (retail) are ❑ Sensitive to face-to-face interactions. ❑ May operate even within buildings 2 Our focus in this slide deck will be on industries that rely heavily on in-person interactions For such industries, the building you are in can make a difference because of opportunities to interact with other companies in the same building. This suggests that buildings can matter. It is also important to remember that some commercial buildings are BIG! 3 Tall Buildings are BIG! ◼ World Trade Center’s twin towers in total had roughly 50,000 workers in 10 million square feet of space ◼ That is equivalent in space to 5,000 homes, each of which is 2,000 square feet ◼ Approximately equal to one-quarter of housing units in Ithaca, NY! 4 Tall Buildings are BIG! A “typical” 1.3 million sqft mixed- use building occupies about one- half block in New York City. The same amount of usable space would typically occupy 21 square blocks in the suburbs! Source: Kate Ascher (2011), The Heights: Anatomy of a skyscraper. New York: Penguin Books. http://thesocietypages.org/graphicsociology/2011/10/25/efficiency-of-a-skyscraper-kate-ascher/ 5  Spatial sorting and equilibrium in tall buildings ◼ Along with co-authors, I have been working on a series of papers that consider the role of commercial buildings as distinct spatial units. See references in the appendix to this slide deck. 6 Buildings are important spatial units ◼ Suppose that productivity spillovers are stronger within as opposed to across buildings. ◼ Why? ❑ Individuals are more likely to have unplanned encounters and meetings when they work in the same building ❑ Meeting with someone outside of your building may require navigating bad weather and/or city traffic. 7 Buildings as important spatial units ◼ Recall also that buildings are managed for profit. ◼ Building managers have the ability and incentive to shape the mix of tenants in the building. ◼ This should make the building more desirable and should increase rent 8 Evidence of building specialization and productivity spillovers ◼ “Agglomeration Economies and the Built Environment: Evidence from Specialized Buildings and Anchor Tenants” (Liu, Rosenthal and Strange; 2024) ❑ Millions of establishment-level records from Dun and Bradstreet that provide the level and mix of employment in each building in five major cities (NY,, Chicago, LA, SF, DC). ❑ We examine the extent and causes of building specialization ❑ We also infer evidence of the impact of building specialization on productivity. 9 Why mix anchor and non-anchor tenants in a building? ◼ Anchor establishments attract smaller companies in their own industry. ❑ Smaller companies are willing to pay higher rent to be in a building with an anchor ❑ Anchors receive a rent discount given the spillovers they create ❑ Retail anchors often have restrictive “go-dark” clauses in their leases that impose penalties if they close their stores. See also Retail Anchor Tenant Lease Agreement. 10 Why mix anchor and non-anchor tenants in a building? ◼ Building managers balance advantages of having a large anchor present against the foregone rent from tenants they displace ❑ If anchor is too large → Not enough high-paying smaller tenants ❑ If anchor is too small → Not enough attraction for smaller tenants 11 Why mix anchor and non-anchor tenants in a building? ◼ Adjust the tenant mix to equalize marginal revenue of space allocated to … ❑ Anchor ❑ Small companies in anchor’s industry ❑ Small companies outside of the anchor’s industry ◼ Why equalize marginal revenue across tenant-types? ◼ Quick answer: Because that will maximize returns and profit. ◼ Longer answer: See the next few slides 12 Adjusting tenant mix to maximize returns ◼ Suppose all tenants occupy the same amount of space (to simplify) and you currently lease to two types of tenants in the building, Type-H and Type-L. ❑ Type-H tenants benefit from having other Type-H tenants in the building but at a declining rate The 1st other Type-H tenant in the building is worth more than the 2nd, which is worth more than the 3rd … ❑ Type-L tenants benefit from having other Type-L tenants in the building and also at a declining rate. 13 Adjusting tenant mix to maximize returns ◼ The return from an additional Type-H tenant includes … ✓ The rent that the next Type-H tenant is willing to pay plus … ✓ The impact of an additional Type-H tenant in the building on the rent that other types of tenants in the building are willing to pay. ◼ The same is true for an additional Type-L tenant. 14 Adjusting tenant mix to maximize returns ◼ Suppose now that Type-H tenants currently provide a higher return per square foot of space. ❑ When an office becomes vacant, try to fill that space with the higher yielding tenant, Type-H. ❑ Continue to fill vacancies with Type-H tenants until there is no advantage to any further change in the mix of tenants in the building. ◼ At that point the return on the next (marginal) Type-H lease will be similar to the next (marginal) Type-L lease → Returns will be maximized. 15 To confirm that anchors attract other companies we … ◼ Evaluate whether the presence of an anchor tenant skews the composition of a building’s other tenants towards the anchor’s industry ◼ We conduct separate analyses for eight different industries ❑ Retail Finance ❑ Advertising Law ❑ Software/Data Eng & Mngmt ❑ Manufacturing Other 16 To measure the degree to which anchors attract other companies we consider several questions ◼ Primary questions: By how much does the presence of an anchor skew the industrial mix of other tenants towards the anchor’s industry … ❑ In the building in which the anchor is located? ❑ In other buildings on the same side of the street on the same city block (the same “blockface”)? ❑ In other buildings across the street on the same city block? 17 To answer these questions, we use various statistical methods ◼ We run many regressions (don’t worry about details if you are not familiar with regression methods) ◼ This allows us to determine the effect of the presence of an anchor on tenants in … ❑ The anchor’s building ❑ Buildings elsewhere on the blockface ❑ Buildings across the street ◼ We are also able to determine the effect of employment in a company’s building, elsewhere on the blockface, across the street, within 0.1 mile, and the composition of employment within 0.1 mile 18 No single way to define an anchor tenant and so … ◼ An establishment is defined as an anchor if ALL three of the following conditions are met … ❑ The building must have 50 or more workers ❑ The establishment must be the largest in the building ❑ The establishment must have 20% or more of the workers in the building ◼ We tested many other definitions of anchors. Results are robust to reasonable alternatives 19 Main results: Anchors attract smaller companies primarily in their own industry, not across industries ◼ The presence of an anchor is associated with a 5% to 25% higher share of employment from smaller companies in the anchor’s industry within the building. ◼ Anchors mostly attract smaller companies only in their own industries. 20  Small retail firms are attracted to buildings with a retail anchor Main diagonal (blue) is the own- Retail industry anchor Small finance firms are attracted to effect for the target buildings with a Finance finance anchor building. Off-diagonal (gold) Advertising are the other- industry anchor effect. That have Law Offices mostly negative and often small effects Software/Data Anaylsis Own-industry Engineering & Mngmt anchor skews composition of smaller tenants Manufacturing towards the anchor’s industry All Other Industries 21 Main diagonal (blue) is the own- Retail Small retail and industry anchor finance firms steer away from effect for the target buildings with a Finance manufacturing building. anchor Off-diagonal (gold) Advertising are the other- industry anchor effect → mostly Law Offices negative and often small effects Software/Data Anaylsis Own-industry anchor skews Engineering & Mngmt composition of smaller tenants towards the Manufacturing anchor’s industry All Other Industries 22 Main results: Attraction to an anchor in your company’s industry attenuates rapidly Attraction to own-industry anchor is much smaller when the anchor is in a building elsewhere on the blockface and less still if it is across the street. Source: Liu, Rosenthal and Strange (2024), “Agglomeration and the Built Environment: Evidence from Specialized Buildings and Anchor Tenants” 23 Main results: Attraction to an anchor in your company’s industry attenuates rapidly Attraction to own-industry anchor is much smaller when the anchor is in a building elsewhere on the blockface and less still if it is across the street. Anchor effect is especially strong in retail where effects also attenuate rapidly outside of the anchor’s building → Suggests that shopping externalities are large and attenuate quickly Source: Liu, Rosenthal and Strange (2024), “Agglomeration and the Built Environment: Evidence from Specialized Buildings and Anchor Tenants” 24 Main results: Attraction to an anchor in your company’s industry attenuates rapidly Attraction to own-industry anchor is much smaller when the anchor is in a building elsewhere on the blockface and less still if it is across the street. Anchor effect is a bit smaller and still attenuates quickly in information-oriented office industries → Suggests that benefits of other nearby companies attenuate quickly Source: Liu, Rosenthal and Strange (2024), “Agglomeration and the Built Environment: Evidence from Specialized Buildings and Anchor Tenants” 25 Main results: Attraction to an anchor in your company’s industry attenuates rapidly Attraction to own-industry anchor is much smaller when the anchor is in a building elsewhere on the blockface and less still if it is across the street. Manufacturing is different! Absence of attenuation suggests that small manufacturers don’t care whether a manufacturing anchor is in their building as opposed to elsewhere on the street → Suggests that frequent in-person interactions are not so important. Source: Liu, Rosenthal and Strange (2024), “Agglomeration and the Built Environment: Evidence from Specialized Buildings and Anchor Tenants” 26 Main results: Robust to alternate specifications ◼ Results are robust to different sets of controls and different reasonable definitions of anchors ❑ Controls for other buildings on the street and neighborhood ❑ Controls for the physical attributes of the building ❑ Analogous results based on building-level average sale/worker for retail and law firms 27 Implications? ◼ In the commercial sector, the mix of other tenants in your building is important! ◼ Within-industry productivity spillovers occur within individual buildings. ◼ Building managers have incentive to adjust their tenant mix to enhance within building spillovers → This makes the building more valuable and allows the building manager to charge higher rent. This also contributes to urban productivity. 28 Summarizing … ◼ These results support the view that the built environment and buildings in particular affect the quality of interactions in the commercial sector ◼ The building you are in can enhance productive in-person interaction ❑ Amplify shopping externalities, benefiting retail ❑ Increase information exchange, benefiting some office industries ❑ This is not as important for manufacturing ◼ Spillovers are weaker outside of the own (target) industry → Contributes to building specialization because small companies are often drawn to anchors in their industry ◼ Spillover effects attenuate quickly upon leaving your building → Contributes to spatial concentration of individual industries 29 Is the office sector at risk? ◼ With work from home (WFH), will companies in the commercial sector continue to care about who else is … ◼ In their building? ◼ Elsewhere on the street ◼ In their neighborhood? 30 Is the office sector at risk? ◼ Does work from home (WFH) reduce valuable interactions between companies, weakening productivity in the office/information sector? ❑ Will WFH undermine demand for central city commercial space? ❑ If so, can we reshape underulitized commercial buildings for alternate uses? ◼ Recent data suggests that many office markets are at risk! 31 Commercial cap rates have risen sharply! Commercial real estate capitalization rates in the United States from 2012 to 2023 with a forecast until 2026, by property type Commercial real estate cap rates in the U.S. 2012-2023 with a forecast until 2026 High cap rate! Retail Office Industrial Multifamily 8.0% 7.5% 7.0% 6.5% 6.0% Cap rates 5.5% 5.0% 4.5% 4.0% 3.5% 3.0% 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024* 2025* 2026* Note(s): United States; 2012 to 2023; as of Q4 of each year Further information regarding this statistic can be found on page 8. 2 Source(s): CBRE Group; Real Capital Analytics; ID 245008 Office vacancy rates have also jumped up dramatically! Quarterly office vacancy rates in the United States from 4th quarter 2017 to 4th quarter 2023 Office vacancy rates in the U.S. 2017-2023, per quarter 17.5% 16.5% 15.5% Vacancy rate 14.5% 13.5% 12.5% 11.5% 10.5% Note(s): United States; Q4 2017 to Q4 2023 Further information regarding this statistic can be found on page 8. 6 Source(s): Colliers International; ID 194054 Vacancy rates are especially high in some cities! Vacancy rate in the largest office markets in the United States in 4th quarter 2023 Vacancy rate of office space in the U.S. 2023, by market 40% 35% 30.4% 27.7% 29% 30% 25.4% Vacancy rate 25% 20.5% 21.4% 21.6% 21.8% 22.1% 18.1% 18.4% 18.8% 20% 15.4% 16.7% 15% 12.7% 10% 5% 0% Phoenix Seattle Manhattan Boston Austin South Florida Detroit Hartford Houston Baltimore Philadelphia Denver East Bay Portland Chicago Silicon Valley Kansas City Charlotte Nashville Minneapolis-St. Paul Atlanta Pittsburgh Milwaukee Fort Lauderdale Los Angeles Dallas-Fort Worth San Francisco Bay Area Cleveland/Akron Metro Washington, DC NYC Northern Suburbs Note(s): United States; Q4 2023; Downtown: class A. 30 markets with the largest Class A office inventory. Further information regarding this statistic can be found on page 8. 4 Source(s): Colliers International; ID 605467 Data suggests that many office markets are at risk! ◼ These plots suggests that WFH and related decline in demand for office space has increased risk associated with investment in the office sector (e.g. high cap rates). ◼ We will talk about this further later in the semester. 35 Appendix: References For Material in this Slide Deck ◼ The Vertical City: Rent Gradients and Spatial Structure (JUE, 2018; Liu, Rosenthal & Strange) ◼ Employment Density and Agglomeration Economies in Tall Buildings, (RSUE, 2020; Liu, Rosenthal & Strange) ◼ Agglomeration and the Built Environment: Evidence from Specialized Buildings and Anchor Tenants (JUE, 2024; Liu, Rosenthal & Strange) ◼ Filtering in Commercial Real Estate (Working paper; Baum-Snow, Heblich, & Rosenthal) 36 MODULE 5: SPATIAL PATTERNS OF REAL ESTATE USE AND VALUE WITHIN CITIES PART 5-A: PART 5-B: RESIDENTIAL – PART 5-C: LOCAL COMMERCIAL – GOVERNMENT HOUSEHOLD SPECIALIZATION BY POLICY SORTING BY INCOME NEIGHBORHOOD AND AND DEMOGRAPHICS BUILDINGS ❑ 1 The Chicago Skyline! 2  Why Do Building Heights and the Related Intensity of Development Decline So Rapidly? 3  How are urban land rents (cleared of any building) determined and why is rent higher in the downtown than in the suburbs? Key Ideas What drives spatial variation in building heights and density? in this What drives horizontal spatial patterns of urban land rents and prices? We will chat about vertical patterns soon! Packet How is urban land allocated between competing uses... Industrial Commercial Residential 4  What determines urban land rents? ❑ We’ll first chat about where firms decide to locate within a city → This determines land prices and rents ❑ Then we will talk about residential developers, a type of firm that uses land to build housing → This links housing and land markets and their prices 5  What determines urban land rents? ❑ Lastly, we will consider the effect of substitution o Firms have the ability to substitute capital for land in production. o Households are willing to substitute non-housing for housing consumption ❑ Substitution affects local rents and patterns of development 6  Some terms we’ll be using ❑ Market value (price) of land: amount paid to take ownership of a parcel of land or a building ❑ Land rent: periodic payment from user of land to owner of land ❑ The cost of land has a direct effect on the cost of housing (for households) and office/industrial space (for firms) 7  Some terms we’ll be using ◼ Land rent and land price usually vary across locations in a similar manner. Why? ◼ Land is an asset. With efficient asset markets recall …  Pt = Rt + Rt+1/(1+r) + Rt+1/(1+r)2 +…+ Rn/(1+r)n  Suppose Rt = R for all t (in real terms). Then … P = R/r and rearranging gives R = r∙P ◼ Because R and P are approximately proportional to each other, they vary across locations in a similar manner ◼ At times we will use land rent and price interchangeably when the spatial patterns of each measure are similar. 8  What determines land rent? ◼ Consider two farms …  Lots of rocks – low yield  No Rocks – high yield ◼ Which farm would you be willing to pay more for? 9  What determines land rent? ◼ Consider two farms …  Lots of rocks – low yield  No Rocks – high yield ◼ Which farm would you be willing to pay more for? ◼ Suppose nonland costs are the same for each farm  $15 on both farms 10  What determines land rent? ◼ What is the maximum willingness to pay for each farm? ◼ Willingness to pay for farm  $5 = $20 - $15 on low fertility farm  $25 = $40 - $15 on high fertility farm ◼ We’re willing to bid more for high fertility land because it is easier and more productive to farm 11  What determines land rent? ◼ Suppose land markets are competitive. ◼ With competitive markets, business owners earn a fair market wage but do not earn unusual profits. ◼ And so, with competitive markets, we will treat profit as being driven to zero while recognizing that entrepreneurs earn wages like other works. 12  What determines land rent? ◼ Solve for the market land rent assuming competition and zero profit  Step 1: Set profit () to zero because of competition ◼  = pq – cnonland – R = 0  Step 2: Solve for R ◼ R = pq – cnonland ◼ Your maximum willingness to pay for land is what is left over from total revenue after the cost of non-land inputs! 13  What determines land rent? ◼ We will refer to a firm’s maximum willingness to pay (WTP) for a parcel of land as its bid rent (a term used in academic literature) ◼ A firm’s bid-rent for a parcel of land is equal to the rent that drives its normal profits to zero (allowing for a fair market wage for the business owner) 14  Manufacturing bid-rent with shipping cost ◼ Suppose that carpet manufacturers in a city have inputs shipped to their factories and then ship their finished carpets to market ◼ Inputs and carpets are shipped by truck using a highway that runs through the city ◼ What is a carpet manufacturer’s bid-rent for land and how does bid-rent vary with distance to the highway? 15  Manufacturing bid-rent with shipping cost ◼ To simplify, suppose that all factories …  Occupy L units of land  Produce q carpets  Incur fixed non-land costs equal to cnonland  Incur higher shipping costs cFreightCost as distance to the highway x increases  Sell product at price p per unit 16  Solve for the bid-rent per unit of land R(x) ◼ Step 1: Set  to zero (because of competition) ❑  = pq – cnonland – cFreightCost(x) – L∙R(x) = 0 ◼ Step 2: Solve for R ❑ R(x) = [pq – cnonland – cFreightCost(x)]/L ◼ Numerical example ❑ Suppose that pq = 250, cnonland = 130, L = 2 and ∂cFreightCost(x)/∂x = 20. ❑ Then R(x) = [250 – 130 – 20x]/2 = 60 – 10x 17 Solve for the bid-rent per unit of land R(x) $ 60 Bid-rent per unit of land. Slope = -10 = freight cost per mile 30 3 6 x = distance to train station 18  Manufacturing bid-rent: Summary ◼ Bid-rent declines with distance from the highway as freight costs increase so that  = 0 ◼ Slope of the bid-rent function per unit land o R/x = – cFreightCost = -$10 o The slope of the bid-rent function ensures a spatial equilibrium. o Why? 19  Manufacturing bid-rent: Summary ◼ Answer: The bid-rent function traces out a set of rents such that a company would earn equal profit at all locations. ◼ With competitive land markets, companies should pay rent roughly equal to their bid-rent and have no incentive to relocate ◼ The slope of the bid-rent function governs the rate at which bid-rent changes so as to ensure that is the case, ensuring a spatial equilibrium. 20  Office Sector Bid-Rent ◼ Information oriented office companies have limited shipping costs. ◼ Would bid-rent vary spatially for these companies? ◼ Yes! But why? 21  Office Sector Bid-Rent ◼ Recall that for various reasons, office companies likely benefit from agglomeration economies ❑ Knowledge spillovers (opportunities to learn from other nearby companies) ❑ Labor pooling (opportunities to improve matching of worker skills with what employers most need) ❑ Support intermediate input providers (e.g. specialized lawyers, accountants and consultants) ❑ For the reasons above, locating near city centers is valuable! 22  Office Sector Bid-Rent ◼ Evidence discussed in earlier lectures suggest that office sector agglomeration economies attenuate very quickly with distance. ◼ In most modern cities, office workers are heavily concentrated in the downtown. ◼ What might the shape of a typical office company bid-rent function look like? 23  Office Sector Bid-Rent ◼ Moving away from the city center, the number of office workers in close proximity will tend to decline very quickly. ◼ Further out, additional movement away from the city center has more limited effect on proximity to potentially valuable neighboring companies. ◼ This suggests that benefits from proximity to the city center likely decrease at a decreasing rate as one moves away from the center. 24  Office Sector Bid-Rent, R(x) R(x) Office sector bid-rent is likely non- linear because proximity to other office workers declines rapidly at first with distance from the center City Center x = distance to city center 25  Substitution Affects Rent and Density ◼ The bid-rents described above made no mention of building heights. ◼ Implicitly, we were assuming that all buildings were of the same height. ◼ But most cities would look very odd if we took away their tall buildings in the center and their iconic skylines. 26  Substitution Affects Rent and Density Imagine Chicago with a flat skyline! 27  Substitution Affects Rent and Density ◼ Why do most cities have tall buildings in the center? ◼ In what way is this related to the shape of bid-rent functions? ◼ Does the presence of tall buildings in the center affect spatial patterns of commercial rent and home prices? ◼ Substitution is the key! 28  Substitution in the Office Sector ◼ Let’s focus on office industries for the moment. ◼ From a technological point of view, most companies are able to operate in low-rise buildings or in tall buildings. ◼ For that reason, companies will respond to higher land costs by operating in taller buildings in order to reduce total costs 29  Substitution in the Office Sector ◼ Suppose the price of capital for building tall structures does not vary spatially. o But we know that land is more expensive in city centers. o Companies will have incentive to substitute capital for land as land prices increase. ◼ This contributes to tall buildings in city centers. But there is more … 30  Substitution affects bid-rent ◼ How would bid-rent be affected by factor substitution? ◼ Ability to substitute lowers total costs ❑ Shift away from land as land becomes more expensive ❑ Shift towards land as land becomes less expensive ◼ Choose the building design that allows you to operate at lowest cost! ◼ Lower total costs increase the maximum willingness to pay (WTP) for land and the bid-rent 31  Suppose No Substitution All Buildings are Four Floors in Height $ Example: Moving from 5 to 1 unit distance closer to the city center causes bid-rent to increase from 400 200 to 360 360 200 1 5 Distance from city center (x) Bid-rent without factor substitution. All buildings are 4 floors in height. Slope = - 40 32  Now Allow Substitution Buildings Can Be of Different Heights Bid-rent with factor $ substitution With no factor substitution, moving from x = 5 to x 400 = 1 (closer to the city center) bid-rent increases from 200 to 360 360 But when substitution between land & capital is possible, moving closer in from x = 5 causes bid-rent to increase by more than when all buildings can only be 4 floors in height. 200 Moving further out from x = 5 also causes bid-rent to increase by more than when all buildings are 4 floors in height. 1 5 Distance from city center (x) Bid-rent without factor substitution. All buildings are 4 floors in height. Slope = - 40 33  Substitution affects bid-rent ◼ Moving out from the city center, land rents decline ➔ Developers respond by substituting land for capital and build lower-rise buildings. ◼ Only where developers prefer to build a four-floor building will the bid-rent be equal to the initial example in which developers only knew how to build 4-floor buildings. ◼ In the previous slide, that occurs at distance 5 units from the city center 34  Substitution affects bid-rent ◼ At all locations other than x = 5 developers choose to build buildings that are not four floors in height: ❑ Taller when closer to the city center (x < 5) ❑ Less than 4 floors further away from the city center (x > 5) ◼ Why? Because it reduces cost. ?