Function Follows Form – This Version: October 9th, 2015 – Kristof Dascher1 Abstract: Urban policy visibly molds urban form. This paper’s interest is in how (if less visibly) urban form molds urban policy. The paper presents two concepts of urban form – the skewness of the city’s shape as well as the convexity of the urban skyline – that frame the political economy of the decision to end the city center’s function as commercial heart of the city. This theme – of how urban form pins down city functions – rivals the proverbial form follows function (Sullivan (1896)) prominent in architectural theory. Function also follows form. This theory the paper then applies to the effects of building height limitations, city center zoning, green belts, central city housing destruction, and to the origins of jurisdictional merger and sprawl. Keywords: Urban Form, Shape Skewness, Skyline Convexity, Population Density, Sprawl JEL-Classifications: R52, L81, N90 Kristof Dascher Regensburg University Department of Business, Economics, Information Systems Management and Real Estate 93 040 Regensburg Germany Email: [email protected] 1For comments on an earlier version I am indebted to Rainald Borck, Alexander Haupt, Yarema Okhrin and Arthur Silve. I am also grateful for comments I enjoyed at the “Bavarian Micro Day”, the statistics seminar at Augsburg University, and the IIPF conference in Dublin. Any remaining errors are mine. 1 Introduction There can be no doubt that urban policy molds city shape. When commenting on Paris’ urban development, Alexandre Gady, a historian of that city, appears to suggest the reverse though: “Paris – it’s beautiful. But it’s a doll’s house! And that’s one reason the Parisian élite is so conservative. They live in the doll’s house. The blindness of the élites is to reproduce a model of returning to the center, always back to the center . ” (as quoted by Gopnik (2014)) Gady argues that city form or shape molds urban policy. This paper’s interest is in the urban political economy of precisely this shape-to-policy connection. We will replace the example of Paris with a monocentric city model, the notion of “doll’s house” with the city’s shape, and the claim of “élites’ blindness” with an analysis of landlords’ pursuit of their interests. Then we will credit the above quote with containing more than just a grain of truth. Here is a three-paragraph sketch of the paper. Consider the secular decision of whether to decentralize city center jobs and shops out to the city periphery – a decision many cities have actually taken (e.g. Lampugnani (1985), Garreau (1991)). Intuitively any landlord’s preference regarding this decision might easily be read off the “average position” of her or his properties’ locations. For example, a landlord with an average property near the city center will not easily give up on that center; the opposite must be true for a landlord whose average property is close to the periphery. If we knew landlords’ property locations we could easily map these into landlords’ policy preferences. We could even go on to try to predict the city’s decentralization decision. Unfortunately, of course, landlords’ property locations are not observable. But the city form is. And this observable form puts constraints on the locations of prop- erties resident landlords can possibly own. After all, these properties must nest into the city’s given form. Detailed “nesting constraints” emerge once we inspect the city’s form closer. The most binding of those happens to vary, as we will see, with a simple index of the city’s physical form, i.e. the city shape’s skewness. This skewness literally puts bounds around resident interests. If the city shape’s skew is strong enough then a majority willing to hold on to the traditional center is inevitable. Conversely, if the city shape’s skew is sufficiently weak then decentralization sets in. The skewness of the city’s shape is our first graphic concept revealing the city form’s hidden grip on urban politics. Our second graphic, and even visual, concept uncovering this hidden grip is the convex- ity of the city’s skyline. Shape skewness bounds urban voter shares in equilibrium; but skyline convexity bounds changes in these shares should shocks to the equilibrium arise. The shocks we consider are as diverse as an Athens charter type functional zoning driv- ing housing away from the center, the wartime destruction of central city housing or its deliberate demolition, or the introduction of a green belt. The urban political economy 1 response these shocks induce tracks, as we will also see, the city skyline’s convexity. If the city’s skyline is convex then a central housing loss reinforces opposition to decentraliza- tion; while the opposite is true if the skyline is concave. What is more, the urban skyline is a variant of the city’s population gradient, and so assigning a political role to the sky- line’s convexity assigns extra weight to the literature on population density gradients (e.g. McDonald (1989), Kim (2007)). Putting fundamental urban decisions down to city form, rather than tracing city form to these urban decisions, is this paper’s theme. Based on this theme it is tempting to suggest that to “see a city” may actually help “understand that city”. While the field of urban forms represents an interesting field in its own right (Baranow (1980), Roeck et al. (2013)), city morphology may have uses that go beyond the descriptive. From an urban externality perspective this seems intuitive enough. Residents care about their city’s form, much as the citizens of New York or Florence care about their cities’ silhouettes (Lynch (1960)). Only, once we stop taking these external benefits for granted, the role of city form for individual well-being is much less obvious. This paper’s theory also is an attempt to uncover the city form’s more immediate, non-external, benefits. Incidentally, the idea of reading restrictions on a city’s various political interests off its physical form complements a prominent view originally put forward by Louis Sullivan. According to Sullivan (1896), “. it is the pervading law of all things organic and inorganic, . that the life is recognizable in its expression”. Among architectural theorists Sullivan’s view has become the proverbial form follows function. According to this view, buildings’ uses (function) predetermine building contours (form). This paper’s suggesting that the built environment impact on the polity’s decision on where to locate a city’s commerce provides an explanation of how instead building contours (form) determine buildings’ uses (function), or more briefly of how function follows form. It is this reversal of ideas that motivates the paper’s title. Analyzing the causality extending from urban policy to urban form has always been a concern to the urban economics literature (e.g. Bertaud/Brueckner (2005), Brueckner (2005), Bento/Franco/Kaffine (2006), Baum-Snow (2006), de Lara et al. (2013), or more recently Ushchev/Thisse/Sloev (2015)). Yet a general analysis of the causality from urban form to policy appears to have attracted much less, or even no, attention – even as specific aspects of urban form, such as the role of urban compactness for the viability of urban public transport, have been of interest (e.g. Bertaud (2003)). In part this is for good reason. A young city’s policy is unlikely to be informed by its form. But for any mature city this seems more difficult to justify. Surely policy adapts much quicker to its surrounding physical structures than these structures do to policy? More specifically, consider a ring road. A ring road creates a rival to the traditional Central Business District (CBD). Locations along it connect almost as well to one another as locations in the traditional core do. A ring road attracts business away from the CBD and remakes the city’s form. At the same time the city’s form just as clearly feeds back into the decision on the ring road. A city characterized by substantial sprawl is more likely to build a ring road. Vienna provides an early prominent example of a ring road, 2 so much so that its ring road is actually referred to as the “Ring”. Victor Gruen, inventor of the modern US style shopping mall, appears to have modeled his malls on Vienna’s ring. Yet while Gruen conceived his malls as the “centerpiece of a tidy, dense, multi- use development”, most developers simply “were building them on the fringes of cities” (Gladwell (2014)). Urban form also plays a role for jurisdictional mergers and sprawl, even as those applica- tions we will not fully spell out. Central cities must fear annexing their suburbs if this, by reducing skew, will make decentralization more probable. These cities will abstain from suburb annexation even if this means foregoing the benefits of merger pointed out by the literature on fiscal federalism (starting with Oates (2011)). Likewise, also allowing for in- tercity migration should strengthen, rather than diminish, urban form’s role. Allowing for immigration could even elevate the analysis of the decision to decentralize jobs and shops to one of the origins of sprawl. A form-driven decision to instal the ring road attracts immigrants to the ring road’s vicinity. The original city form becomes another of those determinants of sprawl (e.g., Glaeser/Kahn (2004)), rather than just sprawl’s consequence. The paper has seven sections. Section 2 outlines our assumptions. Section 3 extracts the ring road’s political economy from the city’s shape, by introducing computable and useful bounds on urban voter shares (Proposition 1). Section 4 puts these voter shares down to a graphic aspect of the city’s shape.