Barnabé Monnot

Year 4 PhD candidate in Singapore, research in algorithmic game theory and large systems with a data-driven approach.






Efficient markets for urban design

Taipei, in an effort to “green” the city after decades of industrialisation and develop alternative modes of transportation, has steadily built a large network of bicycle lanes. The more awkward ones run on parts of sidewalk with two simple parallel lines delineating the space, but the more scenic paths are to be found along its rivers. Though flanked on every side by turbulent expressways and large interchanges feeding into long bridges, once an opening below the motorised hell is passed, you finally access the riverbeds. There, a sinewy bicycle path, following the curves of the stream, connects you from the city to mangroves and picturesque green spaces where herons and fishermen coexist.

The riverside bicycle lane is over 50km long and on a hot day, you will undoubtedly become thirsty and dream of a refreshing drink and a pause after some time there. What are the odds that you will find a place offering you just that along the way?

The hypothetical is a simple question about our expectations: can I expect to find such a space? If I can expect to (and I do find it), the urban design has not failed me. What are the conditions for it to happen?

Surprisingly, this line of inquiry is there in a completely separate field, that of pricing financial assets. Under the assumption of efficient markets, I can derive that my (rational) expectations must respect the true nature of reality. For the issue at hand, if I can expect that there should be a break area with drinks, then, under the assumption of efficient markets for urban planning, there must be one.

Efficiency opens up another way to reason about any system, based on my own rational expectations. How does this work? Let’s look at what is an efficient market first, and see if we can link back this concept to urban design.

The no-gap theorem

Suppose you can buy an asset A for $20 and collect in a year $30. Now suppose that if you buy an asset B for $40 and sell an asset C for $30 (call it B minus C), for a total price of $10 right now, you can also collect $30 in a year. What will you do? Certainly buy B and sell C instead of buying A.

This is called an arbitrage opportunity: the market was somehow offering the same outcome at two different prices: there was a gap. If the market is liquid enough and I can buy a ton of B minus C, which is cheap, I can profit by using the price difference with A, which is expensive. But while doing that, I make the combination B minus C more attractive and soon other buyers want in too. The market will respond by saying “Wait a minute, B minus C should be more expensive than $10 if so many people want some.” Soon enough, prices update such that B minus C is valued exactly at $20, and the arbitrage opportunity, or the gap, is closed.

How can we have such markets? For one, they must be liquid enough, i.e. provide me with the opportunity to buy and sell quantities of the assets themselves. They also must not be monopolies, where a single seller controls the production of A and effectively blocks me from buying B minus C. Under these assumptions, the market is called efficient. All information is available to all the agents (e.g., that they can buy B minus C at the stated price) and thus, if there was a gap in the market, it would soon enough be closed by the agents profiting from that gap. Equivalently: it is not possible to beat the market. Equivalently again: there is no gap.

Cities with no gaps

Let us work on making the connection with urban design clearer. Around the world, cities have developed on a spectrum between more and less centralisation of their planning, expansion and design. To a large extent, urban choices are mostly guided by policies that allow greater or lower control by the residents over the spaces they inhabit. Where that control is lesser, and the city has in effect a monopoly over its organisation, it is more likely to see gaps appear throughout the fabric of the city. Where control is greater, and decisions about aspects of the city’s design are decentralised to its residents, the gaps may close naturally by direct intervention of the citizens. In some cases, it can also be private interests such as owners of an office building that impact public spaces themselves, with the blessing of the city’s construction authority via some clever policy.

But what do we mean by gap? Is it simply economic incentive, as hinted by the no-gap theorem? At first, it can be, when a single citizen decides to open an independent coffee shop out of a little cart along the riverside bicycle lane, looking for a way to profit off the cyclists’ inevitable thirst. But it need not stop there: a gap can be found in the infrastructure of the city itself. New buildings may wish to connect to existing pedestrian or public infrastructure, aided by a policy allowing such private/public links. Or found in the aesthetic realm of the city, where a particularly boring residential area is improved with a little more greenery or improvised art.

The no-gap theorem, under sufficient decentralisation of design choices, would hold for me to expect a pleasant (if green and visually exciting), short (if featuring well-connected infrastructure) or refreshing (with an ice latte by 30 degrees) excursion in the city. Such an efficient market —- though less a market in the sense of trade than in the sense of participation —- for urban design can induce reasonings of the type “If I expect this thing to be there, it should be there”.

For instance, one of the reasons I have enjoyed walking around Hong Kong is not only the myriad exciting storefronts and street activity, but also that I always could very intuitively reach my destination. Do I need to be a few meters higher than I am now and behind that tall building? Let me look for a public escalator somewhere in its lobby to take me up there. As emphasised in Cities Without Ground, a book on Hong Kong’s layered cityspace:

Urbanism in Hong Kong is a result of a combination of top-down planning and bottom-up solutions. […] Footbridge networks throughout the city that grew piecemeal, built by different parties at different times to serve different immediate needs, eventually formed an extensive network and became a prevailing development model for the city’s large-scale urban projects. (emphasis mine)

View this post on Instagram

A post shared by Kerri Spiteri (@necessary_madness) on

The artist Jan Vormann uses Lego bricks to literally fill the gaps.

Filling the gaps

As hinted in the previous section, a lot more gaps than simply infrastructure can be filled to produce a compelling urban environment, one where expecting something cool to happen is equivalent to knowing it will indeed happen, by the forces of decentralisation.

On the other end of the spectrum, centralisation and public policies that discourage such public interventions will have the effect of stifling civic expressions and sapping a sense of autonomy and control over one’s own spaces. One example that comes to mind is the recent installation of golden sheets by Priyageetha Dia on a staircase of her HDB, Singapore’s public housing buildings, in an effort to “upgrade” and beautify the concrete block that is ubiquitous to Singapore’s landscape (over 80% of the population lives in one such block, whose designs are heavily standardised), as well as to prompt a conversation on public and private spaces. Unceremoniously, the town council removed the sheets from the staircase, citing the lack of permission for the installation. She reiterated months later by hanging golden sheets reminding of the shape of flags usually hung at HDB’s balconies in the days leading up to August’s national day, which were also removed. The town council later told that it reached out to the artist to work with her, but it remains that the balance between “top-down organisation and bottom-up solutions” is a difficult one to strike.

Street art or spontaneous installations are one way to fill the gap. Another potent avenue is simply to add a bit of greenery. The benefits are multiple, and in recent years, many cities have developed proactive policies to enroll their residents in these efforts. Paris launched in 2015 the “right to vegetalise”, a permit to obtain allowing you to take care of a little patch of land, while many others have made it possible to install plants on the streets, in public spaces. In the aforementioned Singapore HDBs, some community gardens are tended by the residents and foster social interactions between them.

From the transportation side, new ways of commuting have recently appeared and further decentralised flows in the city. The new ride-hailing and bike-sharing operators offer to fill the gap between your place and a close point of interest — maybe a train station — , that would otherwise be difficult to reach. Decentralisation can induce inefficiency (see the Price of Anarchy), just as many cities have grappled with the problems of oversupply of bicycles or traffic congestion. Now the same cities are looking at ways, more or less restrictive, more or less clever, to introduce a bit of centralisation, perhaps by regulating operators to use a single database of all bicycles and cars and optimise from there. What is clear however is that these innovations, brought by intelligent systems and smartphones, are tipping the scale in unprecedented ways towards decentralised decision-making by citizens and profoundly remodelling the cityscape.

I have argued two things: One, that applying the arguments of efficient markets to urban design yields a novel way to think of well-designed spaces as ones where expectations are self-enforcing. Two, that cities which strike an appropriate balance between public participation and central planning effectively close the gaps that are the precondition to the former arguments. It is worth noting that few economists believe markets are anywhere near the state of efficiency and in a similar vein, that ideal is most likely unattainable in an urban setting. But the intuition it gives carries over and may be useful to interpret why some policies are working and why others aren’t.

Back to posts