Monday, September 10, 2012

March of Folly

Speaking of the March of Folly, I just tweeted a link to the International Economic Development Council's website where bullet 2 blanketly suggests "highway building" as a universal form of economic development.  No context or qualifiers.  Just build those damn highways and expect the money to flow.  And we wonder why economies around the globe are cracking and faltering.

Similarly along these lines, I was asked to expand upon a recently posited theory about network form and the geographic effect of value within the network for the American Society of Civil Engineers' upcoming journal issue on sprawl.  Kudos to them for wanting to be part of the solution rather than part of the problem.

The most interesting point to understand that I found exploring this is that grids have no built-in hierarchy, but the higher order emergent pattern of grids DOES have a hierarchy.  Meanwhile, the dendritic network imposes a hierarchy, but the higher order system (land use intensity) DOES NOT have an emergent hierarchy.  In the dendritic pattern, the system branches every outward.  While in the grid the branching is nested internally.  This creates the divergent forces of disparate networks on value, inward vs outward.

Here is my first draft:

In 1984, historian Barbara Tuchman wrote a book entitled The March of Folly: From Troy to Vietnam. In it, she traced historical examples, which in her own words, paralleled “the pursuit of policy contrary to the self-interest of the constituency or state involved.” Though she exhaustively detailed four such examples, it could also be understood as how otherwise very smart people tend to do very stupid things even in spite of all evidence to the contrary. You won’t be surprised that the corrupting influences tend to be money and power. 

Though her subject matter chronicles historical diplomacy and militarism, I suspect we might be reliving a similar march of folly domestically with regards to our transportation: its planning, funding, and design, or at least its modern incarnation. I suspect this is due to a lack of a broader perspective, a set of metrics allowing us to directly compare how policy initiatives have long-reaching impacts into how well or poorly the city functions. However, those metrics do not exist or at least tend not to be put into practice in any serious manner because of their inherent indirect nature, in the butterfly flaps its wings sense. More specifically, it is the effect modern transportation planning, funding, and design has upon the real estate market, in need of questioning and further examination. 

In fact, it is the very metrics we use for transportation planning, particularly demand/capacity projections and modeling and level of service, that contribute directly to the inertia. The very attempt at solving a problem of our own creation ensures that the problem will always exist. And thus we march forth, creating evermore problems on a much larger scale. Whether it is traffic modeling or level of service, neither of which is concerned about anything of importance in the scheme of city purpose such as the real estate market and individual behavior, these metrics measure mere symptoms not the overall system. Instead, they tend to exacerbate the problem they aim to fix. 

 According to Jevons Paradox, the more efficient something is the greater its inevitable utilization. Pertaining to transportation, the irony here is that making vehicle flow more efficient, assuming that is something other than a theoretical impossibility, actually has an adverse effect on all other forms of transportation, and therefore upon necessary elements to a living system: choice, rationality, and behavior. The actors can’t be smart without choices. 

To ensure that our transportation serves our cities in a more robust and complete way than merely moving automobiles, we must understand cities for what they are: complex systems. That is not to say they are impossibly complicated, but rather their roots can be best understood through complexity science. Rather than being convoluted and chaotic, the complexity is defined by simple operating systems repeated ad infinitum. The result is actually order…if the city is based on a true network of information and infinite feedback loops. 

What is then formed can be deemed a second order system in that it has entirely different properties and behaviors than its individual components. An ant is different from its colony. A leukocyte only knows its task because that’s what its neighbors are up to, with no knowledge of the actual human being it serves. A singular person’s behavior doesn’t describe an entire city. 

Think of any fractal geometry like snowflakes, which are merely aggregated crystalline water (first order system) forming spectacular new shapes (second order system). Same pattern, same process, infinite possibilities. As with all complex systems, cities can be defined by four innate characteristics: individual elements, a generally defined purpose for those elements, interconnections between said elements, and a whole that is greater, and different, than the sum of its parts. 

With regards to cities, it is humans that are the individual elements or actors of the system. Our fundamental purpose, i.e. what drives and in turn governs the system in a bottom-up manner, is best explained by Abraham Maslow’s Hierarchy of Needs. The most basic need is safety and survival, then replication, and ultimately improved quality of life, and advancement of the possibilities for civilization. The city’s behavior and its physical patterns are the aggregation of the millions of daily actions of its actors, the citizenry. However, it is daily movement of the citizenry meeting their needs and the infrastructure governing that movement, which is the primary determinant of the resultant physical form of the city. 

The interconnections of the city as a system are critical in that they are first invisible. We are a social species by nature in that all of those aforementioned needs and the ability to achieve them necessitating the interaction with others. Human need is the fundamental driver of demand. It is the infrastructure of movement and city form that acts as the release valve of demand, dictating where and how it takes shape; facilitating the ability of people to meet their needs. If you have ever seen a path worn into grass, you are witnessing the invisible city made manifest where the physical infrastructure does not match nor accommodate, the interconnections between people, places, and things. 

 Complex systems are based on networks. The shape and form of the network and how well it matches the “invisible city” determines a city’s functional capacity as a facilitator of human need and a platform of possibility. Whereas zoning codes are often described, quite literally, as the genome to the built form’s phenotype, it is transportation networks that have the most profound impact, as the bones and circuitry the body adheres to, forming its ultimate shape. The remainder of this paper will explore the shape of the network and its impact on city form and functionality. 

Networks are the agent for multiplying value while diminishing costs through shared resources. They are economic engines. The primary element of a network is the hub, for cities that is an intersection, the root of all city formations be it the intersection of navigable waterways with verdant, arable land, trade routes, or intercontinental railroads. Essentially, the city sprouts from the seed where opportunity meets availability. It is that opportunity combined with desirability which drives demand. When crowds overwhelmed cities at their current scale, new neighborhoods spawned like cellular mitosis, aggregating and assembling, first as dependent to the initial cell, then interdependent. Grids expanded. Each new addition was still connected as it achieved self-sufficiency. The majority of resident needs there in one place; an outgrowth of the very presence of those residents. 

Transportation networks have always prioritized the preeminent movement technology of the day. It is only in modern times with the advent and ultimate subordination of other modes to the automobile and its infrastructure that the very foundation of cities, the network, has been undermined and replaced by an entirely different form of network. The dendritic network of street hierarchies: local roads feed into larger collectors, which in turn feed larger arterials, which feed freeways, the trunk of the branching structure. It is often wrongly assumed that automobile use and the resultant physical form of the city, sprawl, are a product of market choice. And this is correct. However, the assumption also ignores the role that public agencies and governmental policy towards transportation infrastructure has upon that “market force.” The invisible arm guides the invisible hand. 

Historically, as cities aggregated, there was still a prioritization on the multiplicity of the network, a multitude of routes, thus creating a “reticulated,” or intermeshed, network. Without a central organizing power, and generally defined by foot travel and topography, reticulated networks often meant irregular grids that we typically associate with medieval cities. However, there is also plenty of evidence of more organized, rectilinear grids both before and after, particularly as modern cartography and increasingly sophisticated and efficient real estate markets demanded the regularity that rectilinear grids provide. 

Other than reticulated and irregular (“organic”) grids there are also radial grids exemplified by strong diagonals emanating from important hubs of the baroque and city beautiful movements in urban planning, as evidenced to varying degrees by Paris, L’Enfant’s Washington, D.C., center city Detroit, and Broadway, the diagonal cut across the rigid orthogonal grid of Manhattan. Barcelona is an example of a city possessing each form, the irregularity of the gothic quarter, the dominant radial angles of Avignudas Diagonal and Parallel, and the rigid regularity of the gridded L’Eixample or “expansion” district. 

However, these were simply variations on the same basic principle of the reticulated network, multitudinous interconnectivity of a meshwork of feedback loops and the availability of route choice appropriate to each individual trip where the dendritic network of modern transportation planning funnels routes towards exceedingly larger corridors. At root of the decay of many American downtowns has been the application of dendritic networks to more complex reticulated ones creating a cannibalistic effect wherein value was extracted from one place, downtowns, and relocated towards the edge. I suspect Detroit’s decline has less to do with unions or foreign competition as much as the overlay of the arterial and highway network upon its grid. While Detroit’s population has plummeted, losing over one million people since 1950, it is important to note the metropolitan area has gained nearly two million people in the same timeframe. 

The cannibalistic nature of dendritic networks derives from the first and second levels of order in the various systems. Grid systems lack a hierarchy in their underlying form.  However, the hierarchy emerges due to the higher degree of interconnectivity of certain places within the grid. It centralizes demand towards the most interconnected places. The hierarchy is emergent in that it is only apparent at the higher order system, once the real estate market adapts to it. Therefore it can be said that reticulated, grid networks, due to the multiplicity and density of intersections instills a centripetal or convergent force upon the real estate market. 

By nature, dendritic networks branch out further and further and instill an outward or centrifugal force on the market due to the dynamics detailed above. The dendritic network tries to match the existing hierarchy with a presupposed one, however it is largely destructive. The dendritic network subjugates necessary local connections, which for living systems to exist must have the strongest relationship, in favor of regional and pervasively similar degree of connection. Therefore the resultant second order system lacks a hierarchy as all places are equally, and once again poorly, interconnected. The reticulated grid is centripetal and the modern dendritic system is centrifugal, or sociopetal vs. sociofugal. 

Moving from the theory to simple mathematics, if we take a basic grid and apply a semi-transparent (10%) circle upon every intersection (call each circle a neighborhood), we get a diagram that looks something like this: 



Even though each neighborhood is equal, the intermeshing and overlapping of each singular neighborhood produces a broader pattern for the entire city, a second level of order. A gradient emerges with the great density or hierarchy at the center. 

However, this also implies that each circle or neighborhood is equal rather than the second level order than re-informing the first. If there is the greatest interconnectivity and greatest value at the center that would theoretically impress a new dynamic upon each individual circle. Those that have the greatest overlapping interconnectivity would densify. Below, are two individual circles, one in the center and one at the periphery of the “street” network. Each has a number associated based on the amount of intersections it overlaps. 



 Here is the same chart with the numbers (degree of interconnection) applied across the board: 



 Factoring each intersection’s number as its intensity, which likely correlates to intensity of use, activity, and density, the transparencies of each circle have been altered to reflect its new “degree of interconnection.” Instead of 10% for every neighborhood, the transparencies now range from 8% to 25%. 

 As you can see the graphic is now even darker (more intense) in the center and lighter at the edge than in the previous iteration. The centralizing nature of this exercise is indicative of an emergent centripetal force where the greatest value exists at the most interconnected places, all else being equal. In theory, this correlates quite strongly with Jane Jacobs’ theory of border vacuums, that intensity of activity, and therefore value, comfort, and potential for crime increase where there are less “eyes on the street” due to the degree, or lack thereof, of interconnection. Disconnected networks (often via the application of the new dendritic networks) beget disintegration, which begets disinvestment and decay. The circle is either virtuous or vicious. 

 The growth of Sun Belt cities in particular has been illusory and will prove impossible to retain without the reapplication of reticulated, highly interconnected, networks. In other words, if we were to ascribe a range of hypothetical real estate values from 0 to 10, a place that once had a value of 10, for example (downtown), the dendritic network by nature extracts value from that place and moves it outwards, changing the value of places that once had a value of 0 (nature – not to say nature has no value, but in terms of real estate monetization it is limited. Until…) or 1 (agricultural uses or low intensity and/or undesirable, i.e. noxious industrial uses) and makes them a 2, suitable for low density development like single family homes. The single family homes, a market fueled by indescribable pollution, disease, and poverty from industrialization, invariably demanded retail and businesses closer to them. However, because the land was of low value (poorly interconnected), the resultant commercial development was also low density. The only profitable land is further outward. 

This process also reverted downtowns from value of 10 to a similar value of 2. The reason is that dendritic networks seek to make all places equally connected, albeit poorly. Many downtowns, such as Dallas where I live, have relics of a past where reticulated, gridded networks created the demand for high-rise buildings, many of which are empty vestiges of vacated demand spilled outward due to the centrifugal nature of the new network. Buildings are merely containers of demand, the liquid to fill or evacuate. It is the market’s job to match demand with availability of usable space. At the very moment developers and investors were meeting the market of the 1950’s, providing increased density in downtown, the new transportation network of intra-city highways in and around downtown sapped the demand for those buildings moving out to the suburbs and eventually the exurbs. 

 Here it is important to note the distinction between INTRA-CITY highways and INTER-CITY highways. Inter-city highways link regions, macro-destinations to macro-destinations and were an integral part of President Eisenhower’s vision for an interconnected country. However, as spelled out in 1960 Presidential meeting notes, Eisenhower was furious with his staff for allowing the creation of highways within cities, as an expedient (and reckless) form of slum clearance. It is these intra-city highways that severed local neighborhoods and fragmented cities. 

The aforementioned values derive from the previous conversation of interconnectivity, the greater the accessibility to a variety of people, places, and things, creates value, which instills demand, and thus density. Network integration is the release valve of demand, instilling opportunity and access to markets. Integration begets accommodation. It is the trunk-like nature of the dendritic system which severs more connections than it creates decreasing the overall degree of network integration and therefore value. By nature the trunk of a dendritic network, the freeway, to which all smaller tributaries flow, is undesirable. There are too many vehicles moving too quickly for pedestrians, aka people, to feel safe, instilling a tension in a market which wants to assign value to traffic counts via accessibility and visibility (imagine if the most visited website, Google’s landing page assaulted you with an assortment of flashing pop-up ads). 

Few places manage to balance traffic, desirability, and value like Champs Elysees. However, it is the centerpiece of a highly reticulated network where the degree of integration is the highest demanded a proportional measure of accommodation. It is also important to note when counting pedestrians it moves twice the amount of traffic as the busiest of highways in half the right-of-way width, with far greater real estate values. Why? Because it connects more people to more places because of desirability and proximity. 

Lewis Mumford understood all of this explicitly when he wrote in The City in History in 1961, the primary purpose of transportation is not to move cars, but instead, “to concentrate the greatest variety of goods and people within that limited area, in order to widen the possibility of choice without making it necessary to travel. A good transportation system minimizes unnecessary transportation; and in any event, it offers change of speed and mode to fit a diversity of human purposes.” He goes on later to write, “an effective transportation network requires the largest number of alternative modes of transportation, at varying speeds and volumes, for different functions and purposes. The fastest way to move a hundred thousand people within a limited area is on foot. The slowest is to put them all in a car.” 

While Mumford focused on the efficiency of the system, in Death and Life of Great American Cities, Jane Jacobs gave birth to complexity science. She emphasized the importance of the sidewalk and highly interconnected, complex street networks, not for any particularly physical reason, but rather the value of communication and interaction. Via body language, eye contact, and whatever possible forms of communication, a bustling sidewalk will transmit an infinitely greater amount of information between pedestrians within the system than would drivers on a busy road. 

The key distinction in qualitatively assessing which network form is better comes from the cost of the infrastructure. As sociofugal networks drive people further and further apart, thus forcing those to drive more and more (as evidenced by the Brown University study citing an 18% population drop for every new inner-city freeway or the University of Toronto study illustrating a 1:1 relationship between increased highway capacity and increased vehicle miles travelled). The result is more and more expensive infrastructure to be maintained by an increasingly diluted tax base. Sustaining this system is impossible. 

To combat this and the overwhelming of core cities with heavy traffic, cities around the globe have and continue to shift policy towards strengthening alternative transportation networks, largely based on the economics. For example the Dutch spend $30 per year per capita on bicycle infrastructure. One reason why they do this is so they don’t have to spend $141 per year per capita on freeway infrastructure as the United States does. Furthermore, it empowers the citizenry, as they don’t have to spend $9,000 per year per car in operations and maintenance of that vehicle as is the approximate US average. 

In conclusion, building cities has been largely consistent throughout time except for minor exceptions, variations, and adaptions until the shift in transportation towards the applied hierarchy at the lower order of the system, the infrastructural network. This has since minimized the natural hierarchy prevalent in true cities and thus created sprawl where all places are relatively equally and lowly valued. By doing so, it diminishes the value inherent in a highly interconnected system, which is the very purpose of the city, why it exists, and why it has persisted throughout the history of civilization. In fact, it is possible that the reticulated network is the framework of civilization, aiding and maximizing feedback loops, the platform of interactions and transactions, an idea combustion and competition engine, where skills, labor, ideas, currency, and genes trade, nudging civilization incrementally forward. Only through reversing the process and policies of modern transportation planning, can the march of folly reverse course to the march of progress.