On ‘Smart Cities’ and ‘Smart Growth’

A reader writes:

“Not sure why every idiot now is talking smart cities. It might worth a blog from you to say what you guys name smart cities are things we were doing in transport and working on from around 16 years ago.”

The short answer is many people (‘idiots’ in the parlance of the reader) are buzzword compliant. The buzzword standard today is ‘smart cities’. Now this is some strange fusion of ‘smart growth’ ‘smart transport’ and ‘information technology’.

What is a smart city? The begetter of all knowledge, wikipedia says:

smart city is an urban development vision to integrate multiple information and communication technology (ICT) and Internet of Things (IoT) solutions in a secure fashion to manage a city’s assets – the city’s assets include, but are not limited to, local departments’ information systems, schools, libraries, transportation systems, hospitals, power plants, water supply networks, waste management, law enforcement, and other community services. The goal of building a smart city is to improve quality of life by using urban informatics and technology to improve the efficiency of services and meet residents’ needs. ICT allows city officials to interact directly with the community and the city infrastructure and to monitor what is happening in the city, how the city is evolving, and how to enable a better quality of life. Through the use of sensors integrated with real-time monitoring systems, data are collected from citizens and devices – then processed and analyzed. The information and knowledge gathered are keys to tackling inefficiency.[1]

A smart city is to urban operations (i.e. real-time) what smart growth is to urban planning (long-term, planning horizon). Wikipedia says:

Smart growth is an urban planning and transportation theory that concentrates growth in compact walkable urban centers to avoid sprawl. It also advocates compact, transit-orientedwalkablebicycle-friendly land use, including neighborhood schools, complete streets, and mixed-use development with a range of housing choices.[1] The term ‘smart growth’ is particularly used in North America. In Europe and particularly the UK, the terms ‘Compact City‘ or ‘urban intensification’ have often been used to describe similar concepts, which have influenced government planning policies in the UK, the Netherlands and several other European countries.

Smart growth was in many ways a rebranding and extension of the ‘growth management’ of the 1970s-1990s, which according to wikipedia is

Growth management, in the United States, is a set of techniques used by government to ensure that as the population grows that there are services available to meet their demands. These are not necessarily only government services. Other demands such as the protection of natural spaces, sufficient and affordable housing, delivery of utilities, preservation of buildings and places of historical value, and sufficient places for the conduct of business are also considered.

‘Smart growth’ is more outcome-oriented (and thus more proactive and more governmentally invasive) than ‘growth management’.

I worked on growth management in my youth in Montgomery County, Maryland. I wrote a paper about it. One main point was that growth management, as it was conceived was more or less impossible to do optimally because of the combinatorics of the number of solutions. The objectives were also flawed. Certainly the system restricted growth in the county. This likely drove growth to other jurisdictions, leading to a net increase in the amount of social costs generated by new development. Now of course, were growth management implemented everywhere evenly, this might not occur. The solution to inadequate regulation is always more regulation.

The problems of ‘smart growth’ (and undoubtedly of ‘smart cities’) (and few admit to be in favor of ‘dumb growth’ or ‘dumb cities’) is the assumption that the central planner (manager, operator) knows what is best.  Hayek called this The ‘Fatal Conceit‘ in a book of the same name and suggested that knowledge is widely distributed, and can never be adequately centralized (see his The Use of Knowledge in Society for a somewhat more academic treatment).  It is related to the classic ‘planning fallacy’ of Kahneman and Tversky:

The planning fallacy, first proposed by Daniel Kahneman and Amos Tversky in 1979,[1][2] is a phenomenon in which predictions about how much time will be needed to complete a future task display an optimism bias and underestimate the time needed.

Kahneman and Tversky were being psychological, while Hayek was being sociological. Individuals cannot plan for themselves very well, and others observe planners cannot forecast well (see work by Flyvbjerg, Bain, and our group for examples), why should we think planners (who are after all, individual people) can plan accurately for others, when they are even less motivated to be accurate?

Smart growth, in its extreme, was about a centrally-directed outcome that was in contravention of what the market would have produced in the absence of such direction (hence the need for the direction). This was generally regulatory (police powers), though on occasion supplemented by subsidy or taxation (purse powers).

It would be much simpler to try to get the prices right in terms of the (a) infrastructure costs, and (b) externalities, impose those as taxes, and then let individual actors decide what to do. If the real costs are what the planner thinks they are, if those costs are as important as the planner thinks they are, real estate developers may produce what the planner thinks they should. But we don’t actually know that. We don’t know the benefits from whatever the developers propose. Now arguably, the developers don’t either, but that is their risk to shoulder, and the planners certainly don’t know better then developers on average.

Smart cities are perhaps not quite as presumptuous as smart growth. The idea deals primarily with making city governmental decisions better informed with real-time information. But it still has the mental model of a set of people at a centralized control center dispatching services, rather than decentralized agents within the city making localized decision. It’s selling to the bosses not the workers. It is the same vision transport people had when promoting many of the components of Intelligent Transportation Systems, except moreso. [And people in the transport field have been using sensors (loop detectors, cameras, and the like for decades now, only recently is it called ‘Big Data’.]

It may be that city services would improve if they were more centralized and better informed. The struggle between centralization and decentralization is a common theme, and there is no universal answer. But I have the feeling the benefits are overblown. What specific information are cities lacking that this provides? Certainly digitizing paper records in standardized formats is a step forward. But does the control center add value? How much, and how does that compare with costs.

Traffic control can certainly benefit from better algorithms, but it is not clear they need to be centralized. This research from Carnegie-Mellon argues for decentralized approaches. I suspect the same is true in many other areas considered for the Smart Cities treatment. Now it is certainly smart to use information better and reduce travel time. But that is not the centralized model that is being pitched by the Smart Cities crowd.

Think about smart cars. These might be connected or they might be autonomous. As I note in a recent post, it is more important they are autonomous than they are connected.

In 2009 I posted Dumb search engines, dumb roads, simple cars. It was short, so I quote in full:

From the blog Unqualified Reservations (via Daring Fireball) Wolfram Alpha and hubristic user interfaces

“… control interfaces must not be intelligent. Briefly, intelligent user interfaces should be limited to applications in which the user does not expect to control the behavior of the product. If the product is used as a tool, its interface should be as unintelligent as possible. Stupid is predictable; predictable is learnable; learnable is usable.”

This applies as much to search engines as it does to transportation. Are you listening designers of ITS applications?

The key is for everyone to be as modest as possible, price what needs to be priced, and let actors be as autonomous and decentralized as possible in making decisions given the information about the relevant prices and their own missions and objectives. This is smart, though I would frame it as smart prices, dumb agents. It is not about more information being centralized so the putative decision maker can rationally allocate resources. It is about people optimizing locally with appropriate taxes for externalities and public services, as prices ensure they consider the costs they otherwise would not bear.

Now this might not answer the original question at the top of the post, which basically wanted to know why planners weren’t getting credit for the things planners have been advocating for decades. The answer to that is planners don’t have the marketing capacity of the White HouseIBM, Xerox, and other players entering this space.