Diane Coyle's Enlightened Economist Review of the Open Revolution

I’ve been dipping into a new … book by Rufus Pollock, The Open Revolution. I'm not great at reading whole books online, hence the dipping. What I've read, though, I really like. It states: …

Even if digital technology did create a world in which information could not be prevented from flowing freely, there would still be the question of who pays to create it in the first place. In the absence of a new financial structure such as proposed here, free information might well hinder innovation and creativity, by robbing innovators of their income, so impoverishing us all.

The book starts by underlining that any property rights – and all economists agree well-defined property rights are fundamental to economic growth – are defined by politics and law and social norms. The argument that the intellectual property rights that currently prevail are problematic for the digital economy has been made before, and is forcefully made again here. The book argues for ‘open information’, that “can be universally and freely used, built upon and shared.”

The book looks at patent and copyright protection, and draws the link to the accumulation of significant digital market (and political) power. It then jumps from the unquestioned powers such as Google and Facebook to a chapter on music streaming and Spotify – a marvellous service and one that can hardly be described as exercising undue power, especially when facing up against the recording industry titans. Pollock argues for replacing Spotify’s commercial model, which involves it in policing copyright protection, with a state-sponsored streaming model whereby governments collect a tax and allow citizens to access any streamed music:

“With no fee per track and no limitation on use, this all-you-can-eat buffet is a prototype for how one aspect of the Open world would operate. Money would of course have to be collected somehow to fund it, but instead of ten dollars a month to Spotify, this could be a special fee incorporated in your taxes or added to your internet or mobile bill. This money would then be distributed according to usage, through remuneration rights fees.”

Subscription models are also of course all-you-can-eat ones and so similarly fit well the low or zero marginal cost economics of online. The difference between Spotify and this government-sponsored version of music streaming is scale: the more users pay, the lower the average fee. This is the rationale for public service broadcasting, so in effect we already have this model in the UK. My household pays an absolutely bargainiferous £2.89 per week to have access to a vast array of content for which the BBC has dealt with all the IP negotiations. I’m a huge fan of this model. But I don’t see why poor old Spotify – another marvellous service – should be prevented from asking for £10 a month for its streaming services.

But do we really need government to put in place a levy, or can we leave private companies like Spotify to sort this out?

Spotify is quasi-Open, and consumers evidently like its flat-fee, unlimited-access model. Yet it has its drawbacks, such as a developing monopoly in commercial hands and the lack of universal access. Moreover, it will never be in Spotify's commercial interest to price at a level that gives access to everyone. It is therefore offering a more limited service to both artists and listeners than an Open system would.

Why will Spotify never provide access to everyone? Because users differ in their willingness to pay for the service. Some people would pay a lot for it, because they value it very highly – or simply have lots of money. Others are happy to pay a modest amount, and some can or will only pay a bit because music doesn't matter much to them or they don't have the cash. Spotify, however, cannot tell which is which and has to set a single price for everyone. Furthermore, a for-profit company like Spotify would never set the price that would maximize the number of users. Lowering the price would bring Spotify more customers, but probably not enough of them to compensate for the fall in income from each. So a private monopoly almost guarantees exclusion of a substantial number of potential users.[^8] The Open music model solves this problem by giving universal service and allowing a thousand services to flourish.

At root, platform owners have different interests to the users and artists: open platforms with competition may be great for users and artists, but a monopoly platform is more attractive to those investing in the platform. Furthermore, a platform like Spotify has strong incentives to use its power to shape the development of the ecosystem in ways that preserve and enhance its grip. In particular, it will want to restrict or kill off innovations or developments that threaten its monopoly – a monopoly which if unchecked will give it power not only over music-listening but over artists and record labels, and over technological innovation related to music access and discovery.

The Open model would not be a proprietary platform taking advantage of artists and users, but a neutral platform democratically overseen, mediating between users and suppliers, setting the rules, levying the charges and setting them to optimize the outcome for society as a whole, including users and creators.

There's no need for the government to set up its own Spotify. There is no need for the state to operate a streaming service or create apps for your phone or store your playlists. All it would do is establish a standardized, automatic, blanket-licensing regime, under which any firm could set up a service to provide music by offering the playlists and the apps. Unlike Spotify today, these providers would be mediating as technical distributors only, not as legal distributors. They would not negotiate licensing. In other words, in the Open music model what the state would provide is a universal legal protocol for the licensing of music. Not only would this provide universal access to music and a better deal for artists it would also enable innovation in the search for new technical ways to deliver music to users. This Open legal protocol would create a competitive market of music service providers in just the way that our Open internet protocols have created a competitive market of internet service providers (the companies we all rely on to connect us).

This is not a programme of nationalization. State-sponsored monopolies can be terrible, combining the disadvantages of a private monopoly with an added strata of bureaucracy. In this case, the state's role would be that of promoting competition. It was states which granted copyright monopolies in the first place; if they now introduce blanket compulsory licences instead, and in the process they will be diminishing those undesirable monopolies.

Anyone would be free to offer a new recording through any channel they wished and earn a share of the remuneration fund proportionate to the number of plays of the music. Reciprocally, the citizens would all be able to access music any time, anywhere from anyone. Finally, and perhaps most importantly, there would be no restrictions to prevent building upon the licensing platform: anyone could create a new business or a new kind of business related to music. Not only would this stimulate innovation in tech industries directly concerned with writing, making and distributing music, it would have an indirect impact on quite different industries, such as restaurants that play music or artificial intelligence startups that need music databases for their learning algorithms.

So I’m with the book on the over-reach of IP laws in the digital age. I’d strongly argue for intense regulatory scrutiny of the advertising funded model of Google and Facebook, which might force them to an alternative business model – a flat fee, and utility regulation for some of their services perhaps. Some jurisdictions have started inquiries into the online ads market, almost a duopoly. But I’m not persuaded public versus private ownership and control of the access services is the central issue. For sure, given that digital goods are essentially public goods and natural monopolies, there’s a free riding issue which argues for tax financing, but the first step must be to tackle the ridiculous IP laws.

There’s also a deeper question, I think, which is what one is paying for when it comes to fees to access digital content. The book argues: against “restrictions on which BBC programmes are freely accessible (having, of course, been paid for by the British public) and when, and where, to digital watermarks, paywalls around newspapers…”

Well, I happen to know something abuot restrictions on online acccess to BBC programmes, having chaired the original iPlayer approval process. One source of restriction is indeed a byzantine set of rights – different broadcast platforms in different countries may have specific rights to certain ‘windows’ or time periods. The licence fee payer in these cases has only paid for limited access rights, much as one might wish the limits to be less retrictive and the costs lower. Another source of restrictions was the desire on launching the BBC’s online services not to foreclose potential competing services, a worthy application of competition policy, and also not to cannibalise too quickly the BBC’s own linear services and commercial offer. These restrictions have been relaxed over time. The licence fee also buys not just the intangible asset of programmes – and I’m all for making the archive as freely available as possible – but also the continuing costs of investing in new content. The high initial fixed costs have to be covered. Similarly, I’m all in favour of newspaper paywalls – how else is new journalism going to get financed?