The Great Stink, 2020s-style
Sketching a future state: Digital infrastructure
End State: Blog 2/52
This is the second blog in a series exploring how we govern the future.
In last week’s blog, I argued that policymakers should spend more time getting under the skin of the most advanced practice of production. This was all pretty abstract, so this week I thought I’d course-correct by trying to be more concrete.
Remember the exam question for this blog series is:
What institutional settlement do we need if we’re going to make a success of digital capitalism?
So let’s see if we can sketch out some parts of a new state. Even if we don’t yet know the specifics, can we at least name some key components of a future institutional settlement?
Guesswork and intuition
Before I jump into the specifics, let me start with an excuse. Sketching out the components of a future state is a hard thing to do.
After all, the whole point of this series (and my book End State) is that a new state, fit for the 21st century, will be unrecognisably different to the one we’re living with today. At best, the form of this new state is only just starting to emerge.
To give an analogy: imagine asking someone in the 1890s or 1920s to describe what the state would look like in the 1950/60s, once the social democratic settlement had more fully taken shape.
They might have got some things right — the direction of travel on free education, for example, was probably fairly clear. But the intricacies of the modern state, and its organising principles, paradigms, and mindsets — from the goal of full employment and its associated policy infrastructure, to the modern benefits system, to the way we regulate markets — would have seemed like unorthodox idealism.
If we try to sketch the state of 2050, sitting here in the early 2020s, we’re in a similar position. It’s a job that requires a mix of educated guesswork and intuition. We need to ask some quite open questions like:
- What are the areas of our current policy settlement where, if we’re honest with ourselves, something seems to be missing, even if we’re not yet sure what that something will turn out to be?
- Are there ideas out there today that remind us of those big ideas of the past? i.e. ideas like free education or social security that sounded radical at first but that later came to be accepted as common sense?
There are, I think, some viable and non-arbitrary answers to these questions, which I’ll start to unpack in this post and will develop more in coming weeks.
Anyway, that’s the excuse. But with that in mind, let’s start by describing one plank of a future state that feels fairly obvious: the state’s role in digital infrastructure.
A prediction
I don’t like making predictions but this one feels pretty safe: the state of 2050 will play a far bigger and more sophisticated role in digital infrastructure than the state does today.
A quick definition. What I mean by digital infrastructure in this context is, roughly speaking, the basic underpinning software that enables the operation of a well-functioning society. I don’t mean the physical infrastructure of the internet, like mobile phone masts, undersea cables, and fibre optic networks. I mean code that enables a healthy society and economy.
When it comes to the state’s role in digital infrastructure, there are a whole range of questions still to be answered.
Some of these are abstract questions about the nature of the relationship between the state and digital infrastructure. What particular mix of building, facilitating, funding, enabling, mandating, and regulating digital infrastructure should we expect the future state to do?
And then there are more concrete questions about how this work will be done. What specific new laws will we need to pass and what institutions will we need to create?
A good mental warm-up before tackling these questions is to remind ourselves how we answered similar questions in the past. And the obvious analogy is with the role the state developed in relation to physical infrastructure over the course of the 1800s and early 1900s.
Today the state’s role in physical infrastructure is fairly well-established and stable. It involves a mix, varying in different countries, of building, funding, and regulating. It’s one of the least controversial functions of the modern state. Indeed, supporting the development of physical infrastructure is almost unique as a role for government that is accepted across left and right, even in polarised America.
It took a long time for this position to develop and settle. The modern policy settlement for physical infrastructure took shape over the 19th and early 20th centuries as physical infrastructure — canals, roads, railways, telegraph networks, sewers, sanitation systems, electricity networks — emerged as a critical part of a modern economy.
Although each of these types of infrastructure has its own story to tell, when it came to the social and economic issues that took hold, and the way we solved these issues, there were some recurring themes.
Excessive/unfair prices
When a new type of physical infrastructure emerged, it was common for one owner to secure a monopoly and then overcharge people for access. Sometimes this meant prices were simply too high and sometimes it meant prices varied unfairly, meaning low income people couldn’t access services like the telegraph or post. In other cases, exploitation played out not as excess fees but as poor service. On Britain’s mid-19th century railways, for example, third class passengers often had to travel on trains with no seats or roofs. It was a moment to celebrate when the inferior low-sided carriage was finally upgraded; as one sanguine commentator wrote at the time, the new carriage was “an obvious improvement in that passengers could not fall out”.
Exploitative prices and poor services (typically coupled with high profits) fuelled public anger, and this encouraged the state to intervene, although invariably hindered by vested interests. Sometimes the state started to regulate prices, for example by mandating companies to provide a cheap option for poor people, as with the low cost Parliamentary Trains that railways were forced to provide; or by requiring a flat pricing structure, as with postage stamps and a mandated flat fee for telegrams. Sometimes it meant imposing order, for example by regulating/licensing a system of tolls. And sometimes it meant funding infrastructure directly, such as by raising levies/taxes to build sewers or roads. In general, this all made things fairer without retarding progress; if anything, by widening access, cheaper and flatter prices often sped up the adoption of new infrastructure.
Instability
The diffusion of infrastructural technologies was often marked by bouts of manic speculation. This partly reflected the scramble to secure a monopoly, or to snap up valuable routes or territory; early investors could sniff the rewards that would be on offer to anyone who could establish a dominant position. Instability also flowed from the shape of the economic payoffs involved, with infrastructure typically requiring a splurge of upfront capital investment that paid back only later. This payback was also uncertain, not least because it depended on policy choices (for example, would the state crack down on monopolies?) All of this gave infrastructure a kind of soapy quality — perfect for creating bubbles.
Fragmentation
Privately-developed infrastructure, and the institutions that governed that infrastructure, often turned out to be too fragmented for the job. A good example was the sewage system in London, which wasn’t really a system at all, since it was controlled by lots of local sewage commissioners, which made it hard to join things up. London’s early road network was similar; Dickens captured the fury of one coach driver who made a rainy journey through seven tolls on a trip from Bishopsgate to Kilburn, each time pulling off his sodden gloves as he held the reins of his horse in an ever-rising fury. In the case of sewage, fragmentation was more than frustrating; it was lethal. Since sewage tends to run across administrative borders, only a joined up sewage system could deal with the problem. Without one, London’s rivers filled with so much human waste that thousands died of waterborne diseases like cholera. When a heatwave struck the city in 1858 the smell asphyxiated residents so badly that the incident became known as The Great Stink.
The policy response to fragmentation was to professionalise, standardise, and centralise the state’s approach to infrastructure. This didn’t always mean centralisation in the sense of central government delivering everything; sometimes it meant central government mandating rules for local government, or requiring that things (like inspections) be done, backed by guidance and/or funding.
Incompatibility
As well as being institutionally fragmented, early infrastructure was also often irregular and inconsistent thanks to an absence of standards.
My favourite example of these inconsistencies is the gauge wars, a wonderfully daft episode in British history that ran from the mid-to-late 1800s. The country’s two greatest engineers, Brunel and Stephenson, couldn’t agree how far apart to lay railway lines, and since there were no rules they both just cracked on, laying thousands of miles of track, each at a different gauge. The obvious implication was that wherever the tracks met they didn’t line up, so at so-called gauge breaks passengers had to change from one train to another. By the mid 1860s there were over thirty breaks of gauge on the network.
Shambles like the gauge wars were ultimately resolved by rules mandating what we would today call interoperability, but this often took decades to happen. Although the British parliament had in theory mandated a standard railway gauge as early as 1846, it took until the 1890s to close loopholes and reverse engineer a single network by re-laying wide gauge tracks. Telegraph networks faced a similar challenge when they crossed international borders, and again an interoperable standard had to be agreed.
Digital echoes
So what does this all teach us about the policy challenges that flow from digital infrastructure?
If you listen to today’s debate, it’s easy to hear digital echoes of these earlier dynamics.
There is growing concern, for example, about exploitative pricing by digital platforms, from the eye-watering commission that is charged by platforms like the App Store, Google Play, Deliveroo, or Airbnb, to the the way Google and Facebook dominate online advertising (charging fees that the CMA estimates equate to a staggering £500 a year for every household in Britain).
Instability and speculation are also once again the story of the day, with tech stocks feared to be reaching another frothy high and frenzied excitement about new underpinning technologies like blockchain and NFTs, not to mention nebulous concepts like the metaverse.
Digital infrastructure is also problematically fragmented. The digital infrastructure that sits behind public sector data, for example, is so piecemeal that it’s hard not to think back to those early sewer systems (or non-systems). With governments operating from hundreds or thousands of disconnected data systems, if data smelled we’d have another Great Stink on our hands. And this all has real practical effects. It makes it hard to help people through situations like Britain’s current spike in energy prices, because systems can’t be linked up in order to target support; it means that one part of the state can owe a citizen money while another part tries aggressively to collect a debt; and it means that even now, after decades of effort from reformers, doing something as simple as registering a death can take months of duplicative effort as system after system is updated.
When it comes to fragmentation and a sense of disorder, I’d also point to an even more direct analogy in the digital infrastructure that’s needed for modern cities. Whenever I see a bike from a shared bike scheme lying sprawled on a London street — another symptom of the frantic scrap between brands like Jump, Lime, Freebike, and Beryl — it makes me think of mid-19th century pre-Joseph Chamberlain Birmingham with its mess of private transport operators. Or look at the proliferation of platforms/technologies that underpin the charging of electric cars. Or the parcel delivery vans that shuttle duplicatively up and down our streets, pumping our unnecessary carbon. In all these cases the market has certain features of a natural monopoly but one that’s still in its early days, and a coherent policy position and role for the state hasn’t yet been worked out.
Finally, there’s the issue of interoperability and standards, where the big questions concern the globe-spanning ecosystems run by companies like Facebook and Google.
It turns out that much of our essential digital infrastructure — just like early physical infrastructure — is being built by private companies, and in similar fashion it’s not until these firms have got pretty far along that the significance of what they’re building has started to dawn on us. What the big tech platforms are building is in essence an underpinning digital infrastructure for modern society, an internet-era equivalent to the public square or town centre marketplace. Yet rather than working to shared and collaboratively-developed open standards, the digital ecosystems of Bezos, Dorsey, and co are, like the train tracks of Stephenson and Brunel, not interoperable, so you can’t move between them (or, in this case, pass information across them) just as you couldn’t move between differently gauged tracks. And now of course we see the same worry amplified in early debates about the metaverse — a spatial version of the internet — which likewise is playing out as a war for territory fought between walled-off private kingdoms.
Wrapping up
I hope you’ll empathise if I say I’ve bitten off more than I can swallow in the remainder of this post. We’ll have to come back another day.
For now, though, one thing feels clear: the policy answer to these questions will be fiendishly complicated, just as it was last time around.
When we eventually worked out how to govern physical infrastructure, we did it through a sprawling and intricate machinery, including: new institutions (from the Board of Works to standard-setting agencies); legal devices to secure competitive prices even in natural monopolies (franchise agreements, auctions, subcontracting); funding mechanisms to enable adequate investment (levies, taxes, subsidised loans); and a whole host of attendant norms, cultures, and practises (VfM, procurement, civil engineering).
For digital infrastructure, we’ll need a settlement that’s just as expansive and probably even more sophisticated, and guessing the details of that settlement now would be a mug’s game, quite aside from exploding my word count.
I guess, though, if I was pushed to tee up a few headlines — areas where our current policy settlement feels self-evidently under-developed — I’d point to the need for:
- A new legal regime to reassert permissive standards for the internet, for example by opening up the ecosystems of big tech, mandating interoperability, and forcing the biggest platforms to open up their data for researchers and entrepreneurs to use.
- A suite of new institutions to build, fund, and mandate the use of shared digital platforms in the public sector, including, for example, a central body to promote/mandate the use of shared platforms and data registers across government; an institution to collaboratively develop and mandate shared standards across the charity sector; and a body to drive the development of a shared digital infrastructure across local government.
- New organising mechanisms to enable the collaborative development of software/platforms, to underpin markets like the ones I described above. This will need to run alongside new ways to reward people for contributing to the development of these shared platforms, such as through new approaches to IP.
- Cultural norms, so that we develop digital infrastructure with the right mindsets and skillsets. This feels analogous to the role played by civil engineering, and the mindset/values that were applied to build and design physical infrastructure in the Victorian age.
Some of these suggestions might seem a little sweeping or radical, and in general they’ll probably feel like an agenda of government overreach. My guess though, for what it’s worth, is that they’ll look laughably timid when we look back from 2050, by which time I suspect we’ll have made far more radical changes than these.
The good news, however, as ever in digital reform, is that a whole community of dedicated people have been working on all these areas for years, laying the groundwork for bolder reforms or even setting up what are in effect nascent or shadow versions of the new institutions we need. If one thing’s missing it’s political leadership with the same boldness, energy, and depth of understanding of the internet as the practitioners.
Let’s wrap things up there. As I said at the top of this post, sketching a future state is hard, and I guess at worst I’ve proven that point! In future posts, I’ll look at other planks of a new institutional settlement, from the question of how we give people security in a digital labour market to economic regulation in digital markets.
If you’d like to read along, you can follow me on Medium for free here or, if you’d like to support the project (and get a free book), you can join the £3 a month club over on Substack.
If you’d like to read the big (and I promise optimistic) story that sits behind all of this, you can find it in my book, End State: 9 ways society is broken — and how we can fix it.
