I spoke last week at a conference in Cardiff on the future of work. It was organised by the law firm Dawson Gray. You can’t talk about the the future of work without thinking about the future city, since the shape and structure of work is bound up more or less completely with the shape and structure of cities.
Edward Glaeser’s book The Triumph of the City (2012) gets a lot of love in these conversations. It’s hard to find people who have a bad word to say about it. In the presentation, this was the quote that I pulled out to illustrate his version:
Cities are the absence of physical space between people and companies. They are proximity, density, closeness. They enable us to work and play together, and success depends on the demand for physical connection… And during the last thirty years… technological change has increased the returns to the knowledge that is best produced by people in close proximity to other people (p6).
The first half of this seems like a description for the ages—the city as experienced in Roman times, or Elizabethan England, would match this description. The second half is actually more contentious. The phrase “the last thirty years” is a bit of a giveaway. Because what he’s actually describing in his book is not a general model of the city but a particular one—the version of the city that has emerged through a combination of technology, globalisation, and the growth of services as the overwhelming share of modern economies.
In this model, competition between cities happens at a global level, and represents competition for high value added companies, mostly in the tech and finance space. There’s two elements to this, one of which is endlessly discussed, and another which is overlooked.
The endless discussion: that if companies can be based anywhere, no longer tied by raw materials or sunk industrial capital, then they will go somewhere—and the somewhere will be the places with big pools of labour and also some corporate prestige. (As we discovered when New York told Amazon it wouldn’t subsidise its new headquarters building, while other cities would, and it decided to go to New York anyway.)
(Google Ireland. Photo by Outreach Pete. CC BY 2.0 http://www.outreachpete.com/)
The overlooked: the returns to knowledge that Glaeser discusses are as likely to be the result of extraction and rent-seeking as from higher productivity generated by all that close proximity. (There’s a telling passage in Glaeser—p5, if you want to look it up—where he extols without irony or self-wareness the fact that 1980s New York produced financial innovations such as Michael Milken’s junk bonds, and Henry Kravis’s debt-funded leveraged buy-outs. That is the same Michael Milken who is now a model citizen having served 22 months in jail for violating US securities laws.)
So maybe this model is less impressive than it looks. And it may also be a feature of a particular moment of capitalism.
If you want to do this from first principles, you need to look at the Santa Fe physicist Geoffrey West, whose book Scale developed a theory of how cities work. His starting point was a finding from the data that says that a city in any given country that is twice as large as another city produces 15% more of everything—wealth, yes, but also crime, illness, etc. It also does it on 15% less infrastructure. These are interesting numbers.
The second startling fact in the book is the amount of energy humans get through in cities. In the pre-industrial age, humans would get through about 100 watts a day each. Now, they get through 11,000 watts. The ecologist Herbert Girardet has a striking image related to this: that Europeans have the equivalent of 60 full-time ‘energy slaves’ on tap, Americans, 80 ‘energy slaves’.
West’s model therefore starts from energy and infrastructure, which he says drives social connectivity and social capital. From the interaction between these comes economic opportunity. I developed a simple schematic that tried to link these together (West in not to blame for this diagram).
(Geoffrey West’s urban model, visualised by Andrew Curry. CC BY-SA-NC 4.0)
West was born in Britain, but perhaps has been in the US for too long, because he says that this economic layer is driven by “greed”. But the data he quotes don’t say this. Regardless of city size, a new business is created for every extra 22 people in the city, and these businesses employ, on average, eight people. That doesn’t sound like greed to me, more like people making a living.
The problem with Glaeser’s glittering economic version of the city is that it is almost completely dependent on fossil fuels to bring in the resources it needs to survive. All of these global cities have a vast global footprint. Girardet, again, calculates that London has an ecological footprint 125 times its size, pretty much the equivalent of all the productive land in Britain.
In their book, How Everything Can Collapse (2020), Pablo Servigne and Raphael Stevens say:
The paradox that characterises our era… is that the more powerful our civilization grows, the more vulnerable it becomes… To preserve ourselves from serious disruptions to the climate and the ecosystems… we need to turn off the engine (p88).
A failing model
These cities are no longer working at other levels, either. They are also failing to reproduce themselves, in that sense that they struggle to provide adequate care for their young, their sick, their elderly. Increasingly they are also unaffordable both for their essential services workers, and for the generation of young workers that historically would have represented the next generation of managers and entrepreneurs. All of this has been exacerbated by the increases in housing costs caused by deregulation of the financial system and asset inflation that has benefitted the wealthy at the expense of everyone else.
So it’s possible to imagine that the city that Glaeser described in his book was not the city of the future, but a city that was already sliding into the past. If this is true, it has significant implications for the future of work.
In summary then: for a range of reasons—environmental, economic, social—cities are no longer able to sustain themselves.
A sustainable system
When I was doing the research, I revisited West’s triple model, of energy and infrastructure, economic opportunity, and social connectivity, through the lens of what it would require to make this a sustainable system once again.
My version of this diagram looks like this:
(Geoffrey West/ Andrew Curry. CC SA-BY-NC 4.0)
In summary: The urban energy system is based on renewables, not fossil fuels, and this energy is increasingly produced locally. Time becomes more important in the way the economy works. Care—in all its aspects—becomes more important to social connection.
I’m going to expand a little on each of these points here.
In this new model, energy moves from being globally produced and distributed to being more locally managed, and possibly more locally owned as well. Other current imports become more local as well (to reduce the vast urban footprint) such as food. I’ve heard that supermarkets, for example, are investigating how to use hyroponics in the UK to replace imports of crops currently grown in Spain. There are new types of infrastructure and investment to manage climate mitigation and adaptation. And there are social and political changes designed to reduce the city’s energy consumption—such as the ‘15 minute city’.
(The 15 minute city. Concept by Carlos Moreno/ illustration by (c) Micael)
Time has been a battleground ever since the industrial revolution—‘not a minute on the day, not a penny off the pay’, as the old miners’ slogan put it. From memory, many of the gains in productivity in Europe are being translated into gains in leisure, not more income, and this was true in Britain when we still had productivity growth.
Ideas about the flexibility of the working week and working days have been accelerated by the pandemic, and there are trials of the four day week all over the place, even in Britain.
Time becomes an important element of care—and also enables people to volunteer and engage in other local activities. As mark Fisher wrote before he died,
Most political struggles at the present moment amount to a war over time.
If cities want to be able to function, they need to make sure that care works. This means care for schools and children, health, transport, eldercare, improved mental wellbeing, and so on. Care was defined in a 2020 book by The Care Collective as
a social capacity and activity involving the nurturing of all that is necessary for the welfare and flourishing of life.
It’s striking if perhaps unsurprising how much of this literature is gendered—most of the innovative thinking is being done by women writers and researchers.
But if it is going to be effective, care—in all of its forms—needs to be disconnected from the market: the needs of one don’t align well with the incentives of the other. Things that should be done by ‘guardians’, in Jane Jacobs model, shouldn’t be put in the hands of ‘traders’ instead. In her book on work the sociologist Lynne Pettinger asks:
What would happen if care work, not industrial production, was the start of understanding work?
And it’s worth noting that almost all—perhaps all—of Britain’s current wave of strikes are actually fights about the importance of delivering care, or not. Perhaps that’s one reason why public opinion has largely remained supportive despite attempts by the government and the media to replay the 1970s.
Wot? No AI
I didn’t talk much, or at all about AI in my talk, and since much of the discourse on the future of work is obsessed with this, it’s worth explaining why. One of the issues here is that this conversation is still overshadowed by Frey and Osborne’s 2013 forecastthat AI would have an effect on 47% of American jobs, which both missed new jobs that AI might create (as other technology changes have created new jobs), and had a simplistic binary model of how jobs are constructed—as bundles of tasks that evolve over time.
Generally estimates of impact have been falling ever since, while Frey and Osborne have been adding layers of nuance on this ever since, and Osborne and his model was involved in a Pearson/IPPR project which did take account of new jobs found an adverse impact in the range of 5-10%. In a world where time is a new element in the economy, this is a modest impact.
The impact of technology on work is also a function of market power. Digital technology has already wrought destruction on labour markets in the restructuring at the end of the 20th century, when the rate of global population growth and globalisation meant that unions were at their weakest and most workers had little bargaining power.
From this perspective, augmented work—people plus AI—is the mostly likely future.
After the ICT wave
But there’s also a bigger picture story. Information and communications technologies, in Carlota Perez’ model (which also draws lightly on Kondratiev), is the fifth wave of economic transformation since the Industrial Revolution, and it’s coming to an end.
But there’s always a messy overhang, when the innovation that lays the ground for the next technology wave has happened, but it hasn’t yet had market effects. In that phase a couple of things happen. The first is that the technology gets ‘normalised’ in everyday life—think about the equivalent period in the ‘70s and ‘80s around the car, when we saw the evolution of business parks and edge of town stores, a final significant wave of motorway building, and the development of ‘logistics’ as a critical business sector.
The evolution of AI should be read, I’d say, as the equivalent of this.
The other thing that happens is that regulators catch up with the technology, and are finally alert to the risks of adverse externalities.
It’s also worth noting that each of these new waves tends to be distinctively different from the previous one, and that it generates a huge amount of money for the businesses that are in the right place at the right time. There are real questions about where and how AI will make money, which I’m planning to come back to here at some point. But without that potential for rapid sectoral growth, you don’t get a technology wave. And AI is operating in a saturated market.
Living in this city
The city that emerges from this transition is more resilient to shocks. It’s more liveable. It uses data, but as a way to improve its services. It has a richer social and cultural life. It’s probably more decentralised and more distinctive, partly because people who spend more time at home are more invested in their local area. And it probably marks the end of the noisy hype about the ‘global city’.
There’s a fine quote from the 20th century urbanist Lewis Mumford, quoted by Geoffrey West in Scale, that talks about what cities are supposed to be for:
The chief function of the city is to convert power into form, energy into culture, dead matter into the living symbols of art, biological reproduction into social creativity (p373).
This is a much richer version of the idea of the city than the one we have got used to talking about in the past thirty years or so. The focus on GVA (gross value added) and all the other narrow markers of urban success is killing us all.
The slides from the talk can be seen here.
A version of this article was also published on my Just Two Things Newsletter.