4 June 2018

Is capitalism in need of a radical transformation?


Building a new railway in Britain is a herculean task. When Crossrail eventually starts operating at the end of this year, it will have been 75 years since the first proposals were made, and it will have cost the taxpayer tens of billions of pounds. If Crossrail 2 ever happens, it will have been at least 50 years in the making, and will likely be a similar burden on public finances.

It wasn’t always like this. Along with most of the country’s railway lines, London’s commuter rail network was built privately, paid for in large part by the uplift in land values that railway access creates. Buy a field in Surrey for pennies, connect it to Waterloo by train, and suddenly you have prime real estate to develop. The system worked – and in some countries, still does.

One of the things that stops this today is hold-ups. Landowners along the route can refuse to sell their property to a railway company except for an enormous amount, capturing a large amount of the value of the line for themselves and possibly making the line uneconomical to build at all.

Hold-ups are a problem in areas like radio spectrum and patents as well. Because radio spectrum is most efficiently supplied in wide blocks of continuous spectrum, licensees who hold narrow, less useful pieces of spectrum can end up demanding more for the licenses than they would be worth on their own from an operator who needs that piece to create a continuous block. Television broadcasters in the United States have obstructed the creation of mobile data spectrum blocks in this way. Patent markets have similar problems.

For Glen Weyl and Eric Posner, the authors of Radical Markets: Uprooting Capitalism and Democracy for a Just Society, these are not isolated examples but emblems of the whole rotten system.

The hold-up problem is, really, a monopoly problem, where control over an asset allows the owner to extract more from would-be innovators and investors than it would be worth in normal times. In their words: property is monopoly.

Weyl and Posner suggest a system of taxation where, in principle, everything is available to buy, all the time. A homeowner or patent holder, for example, must declare the price they would sell their house or patent for – and if someone comes along and offers them that, they will have to sell. Assets are taxed annually on that valuation – they suggest a rate of 7 per cent – with the proceeds distributed evenly among a country’s citizens via a cash “dividend”.

The effects would be dramatic. Asset prices would fall by between a third and two-thirds of their current level, since the taxes that will be owed in the future would be priced in today. The large 7 per cent tax proposed by Weyl and Posner would raise roughly 20 per cent of national income – enough to eliminate other taxes on wealth and capital, and in the US pay for a $5,300 per person “dividend” as well. Where hold-ups are a barrier to innovation and investment, the problems would largely be solved. Property ownership would become an inherently temporary phenomenon. And capitalism would be transformed.

Well, maybe. The elegance and ambition of their scheme is admirable. For things that do face significant hold-ups, this could be a solution. But for the economy as a whole, their proposal seems like an answer in search of a problem.

Hold-ups may be an issue in markets where goods are effectively unique and there’s not much we can do to invent more. We may not be able to create more radio spectrum, say, or more land on Oxford Street.

But most income-generating assets aren’t like that. If you don’t want to sell me your shares in Amazon for the market price, no problem – someone else will. That’s how almost all other markets work, because supply usually isn’t as constrained as the market for radio spectrum or land. I don’t have to worry about Tesco holding out on selling me a sandwich to extract monopoly rents from me. I’ll just go to Pret.

And a dynamic view of the economy, where new ideas can create new things, changes the calculus altogether. In Weyl and Posner’s static model of the economy, monopolies are irredeemable. But when we add innovation into the model, the prospect of monopoly rents creates an incentive to try to become a monopolist yourself by inventing something new.

That there was only one patent for Viagra, giving Pfizer a monopoly, is the whole point – the rents Pfizer could raise by overcharging was what created the incentive to research Viagra (which was discovered in the process of researching new heart disease medicines) in the first place. The state grants ownership over some ideas because monopoly can incentivise innovation, as well as stifling it.

And land’s heterogeneity is less important than it may seem – the more potential routes there are for a new railway line or road, the weaker potential hold-ups are. Tokyo’s railway system is largely private and funded the same way the old English lines were funded, with minimal use of compulsory purchase. The major difference there is not hold-ups, but Japan’s far more liberal planning laws.

In cities like London and San Francisco that are experiencing housing shortages, planning laws are a much more important barrier to almost any kind of development than individual property owners refusing to sell. The fact that people cannot make more efficient use of their land by building more densely (or by building at all) is the issue at hand in most of the developed world’s property markets. Where hold-ups do exist, compulsory purchase orders and eminent domain are not obviously inferior to Weyl and Posner’s plans.

It feels boring to raise mundane economic objections to such a radical proposal, but Weyl and Posner are clear that this is no thought experiment. So the usual arguments against taxing capital apply to this proposal too. Taxing an asset reduces the incentives to invest in that asset. A 7 per cent tax on all assets would capture two-thirds of all returns on investment, wiping out an awful lot of productive investment and, hence, economic growth.

Taxing the rich more than the poor is a very good thing. But we should tax them when they consume their wealth, not while it’s being invested. Eventually all invested wealth will be consumed – you invest so you have more to consume later.

It’s consumption where the resources are effectively destroyed, and no longer of any value to anyone apart from the person consuming them. In the investment period, those resources are being used productively by others to make more wealth.

Since we have to tax something, taxing consumption, especially the consumption of the rich, should be the goal. We have tools to do that already, like the value-added tax, and others that would tax the rich once they consume (destroy) their wealth, not while they’re investing it.

In Weyl and Posner’s world, people would spend more of their money on things that cannot be taxed or valued – expensive holidays, fine dining and other expensive “experiences” – as well as assets that are valuable to one person but unlikely to be valued by others. Expect crystal statues of beloved family members to become the next big thing in Weyl-Posner-land.

The book’s other proposals are a bit of a grab-bag of ideas.

Quadratic voting, a system that allows people to express how strongly they feel about a particular outcome at the cost of losing say in other votes, is worth exploring. By allowing people to trade votes between different issues, quadratic voting might allow people who feel very strongly about something to win elections even when they are in a minority, if the majority is relatively indifferent. That could build protection for minorities into the electoral system and make election outcomes more reflective of how people really feel about the options available to them. On the other hand, giving more say to political obsessives may have its downsides.

The idea of allowing individuals to auction off work visas to foreigners, in order to return more of the gains from immigration to a receiving country’s citizens, is elegant too. But it could come dangerously close to indentured servitude.

Whether that would make the programme bad overall is much less clear. Weyl and Posner are refreshingly clear-thinking about the misery of developing world poverty and the improvement that even life in very brutal migrant programmes like the Gulf States’ can bring to someone from a very poor background. If their proposal did make the public more accepting of more migration from poor countries, and we felt we could trust our government to protect migrants from exploitation, it may be worth the risks.

Most topically, the “Data As Labour” chapter argues that today’s internet users – who “pay” for services like Google Search and Facebook by providing data – are being badly exploited.

As data-hungry machine learning algorithms grow in importance, these users will end up being like medieval serfs or Victorian workers before the rise of labour unions, giving up their data “labour” for pennies – individually weak, but collectively strong.

This is unconvincing. How do we know that people are being underpaid? Software like Google Maps and Search, Instagram or Android, which effectively are traded for users’ data and attention to advertising, are very good compared to paid-for alternatives, and do not look like bad value, for most users.

Would you prefer money instead of those services? The most wildly generous estimate of how much money there is to pay users still amounts to very little: all of Google and Facebook’s global revenues – that is, ignoring the money they spend on staff, R&D, marketing and sales – combined and divided only among their American users would still only add up to a few hundred dollars a year.

Where is the money for “data labourers” supposed to be coming from? The argument is pure conjecture – that in the future, user data will become much more valuable, while the services they receive in return will not. Maybe, or maybe not.

The idea of a “data labourer” is misconceived as well. Most of the data I produce, like my search engine queries or photo uploads, is a by-product of other activity, and unlike work does not come at the expense of leisure. Most of the data that tech giants use is valuable precisely because of this. If people were paid to search the web more than the cost of the time it took to do so, search engines would be flooded with junk data. Goodhart’s Law would apply: when a measure becomes a target, the measure will no longer be useful.

Weyl and Posner’s case here is rife with non sequiturs. The fact that software firms spend much less on wages than traditional businesses should make us wonder about whether infinitely scalable “capital” will change the position of workers in the economy.

That is the sort of question that Stian Westlake and Jonathan Haskel consider in Capitalism Without Capital and that David Autor et al’s work on the labour share of national income (cited elsewhere in the book but ignored here) has explored. But Weyl and Posner ignore the possibility that software may be fundamentally different to physical machinery, and assume that Facebook’s small wage bill must mean that somewhere, somehow, it is exploiting someone.

Throughout the book, examples of narrow problems (such as hold-ups in spectrum or land markets) are taken to show a general market failure in need of revolutionary solutions. Weak evidence and speculation is used to justify colossal claims. It’s superficially impressive, but ultimately shallow.

Weyl and Posner see themselves in the mould of Henry George, an economist who railed against land ownership and whose Progress and Poverty was profoundly influential in its day, as well as genuinely popular – the only real economics bestseller ever written.

Radical Markets is filled with ideas that, modestly applied, might be useful. Where it fails is in trying to present them as revolutionary alternatives to neoliberal capitalism. It is an interesting and original book – but a frustrating one too.

Sam Bowman is an Associate at Fingleton Associates.