Those of you with the leisure and patience to read some of my blogs may have noted a recurring structure to their argument. I briefly describe a popular idea or position (e.g. 15-minute cities, cycling, electric-cars, heritage protection…); I then acknowledge the benefits of the idea but discuss some complicating factors. My point of contention is often that the benefits of the idea are over-stated, and that its ramifications and unintended consequences are often downplayed.
Cities as complex systems: urban planning is not rocket science
This points to a wider reality, one that explains why the general argument outlined above is applicable to almost all ‘simple’ or ‘one-factor’ solutions to urban issues: cities are complex systems.
This does not mean that individual elements of cities are complicated: one does not require mathematics or special relativity to think about cities! It means that cities are unpredictable entities, and that causal relationships within cities are always tentative and subject to change.
A complex system has two key properties. First, it is an entity made up of inter-related parts. Watches, computers and rockets also possess that property. Second, the behaviour of the entity as a whole cannot be predicted from detailed understanding of how each part works and connects.
It is this second property that makes cities far more complex than rockets. A rocket is highly complicated, but functions in a predictable environment governed by the laws of physics: with sufficient engineering knowledge and mathematics, a rocket’s behaviour can be foreseen, modelled, and directed. That is why rockets can land on Mars and deliver delicate payloads to its surface – indeed, when things become unpredictable, rockets crash.
A city is complex. It is made up of many parts. Each part, by itself, follows the basic laws of physics: but each part can also behave in ways that are unpredictable (many of these parts – such as people, organisations and companies – exercise free choice!), whilst displaying statistical regularities once aggregated.
It is because statistical regularities exist that cities usually reach some sort of equilibrium. Despite each element of a city being unpredictable, we can establish broad expectations about how its elements behave in aggregate. At equilibrium, all elements of a city will have adjusted to all others, in ways that are often poorly understood, and that urban planners try to make sense of. This (dynamic) equilibrium will slowly evolve through feedback, aging, immigration, technology, politics, etc… Indeed, one role of planning is to stabilise and manage this evolution.
Why one-factor ‘solutions’ are often problematic
Along come advocates of seemingly excellent ‘solutions’, say solutions to make the city more manageable , more sustainable or more creative. Such solutions will typically be narrow, focussing on a small facet of urban life – simply because cities are too vast and complex to be apprehended in their entirety.
But any change to one element in a system will have ripple effects across the whole system. It will have unexpected consequences, and may trigger deleterious effects in areas that advocates have not foreseen. In the longer term, cities usually return to equilibrium after such disruptions, sometimes, but not always, to a ‘better’ equilibrium (whether it is ‘better’ will depend upon who you ask, something that rocket scientists typically don’t need to worry about).
However, a return to equilibrium is not always possible. A key property of complex systems is that, on occasion, a change has so many unforeseen consequences that it destabilizes the system and leads to rapid and uncontrolled change. Climate scientists have for many years been telling us about threshold effects: beyond a certain level of global warming, its effects may become self-reinforcing and may lead to a host of completely unknowable outcomes. They are wrongly criticized for not being able to predict the threshold: but that is the nature of complex systems – they are inherently unpredictable. Thresholds can only be identified in hindsight because of the nature of complex systems.
Cities, like climate, are complex in this way: they tend to be stable and self-correcting, but certain changes may have unforeseeable system-wide effects, and should be approached cautiously. When cars were introduced, it is improbable that the myriad changes they led to could have been foreseen. Likewise, the rush towards ‘urban technology’ and AI systems (that manage cities as if they were rockets, conceptualising them as predictable and manageable) may lead to unforeseen consequences, and should be approached with caution.
How we conceptualize cities is important for planning practice
As urban planners, how we conceptualise cities is critical to the way we think about intervening in them.
Modernist planners saw the city as a rocket: complicated, but with clear cause and effect sequences. From that perspective, urban issues could be ‘solved’ with knowledge, modelling, and technically correct choices. Urban poverty? Demolish the slums. Congestion? Build highways. Economic woes? Attract creatives. Too many cars? On your bikes!
More recently planners have (implicitly at least) approached the city as a complex system: this requires consultation, not only as a tool for democracy, but as a tool for revealing unsuspected inter-dependencies. It leads to plans that are more tentative and flexible. Process becomes as important, if not more so, than expert predictions, technique and causal models.
Is urban planning reverting to modernism?
One of the issues that planning seems to be facing today is the return of modernism – in the guise not only of California-style technological determinism (e.g. Google will ‘solve’ the city with AI) but also of single-issue advocacy and single-factor solutions (be it the creative class, bikes, or building houses).
Each of these approaches views cause and effect as mechanical, displaying some of the arrogance of modernist planners who, as experts, thought they understood how cities work, and could manipulate them in a straightforward way.
But urban planning is not rocket science: it’s far more complex, and calls for more humility.
Richard Shearmur 