Free, marginal cost pricing and policy

Chris Anderson managed to get an article, and then a book of the article (a pet peeve of mine, but we’ll move on) out of the idea that ‘free’ is a big deal. Better than a low low price, free avoids ‘mental transactions costs’ and is all round a Big New Thing. One thing that I don’t think he mentioned is that the ‘free’ of the internet revives to practical relevance an old insight from textbook economics which was theoretically clear but always of less use than some economists appreciated because of the practical difficulties of implementing it. (Sound familiar?)

There’s always been a ‘textbook’ case for governments to help subsidise from general revenue their own or even other service providers’ fixed costs to make marginal cost pricing (as supported by Economics 101) financially viable. But the incentive perversities and institutional complexities of doing so were usually formidable. If you issue subsidies for fixed costs, how will you get those fixed costs met at lowest cost? How will you prevent various tricks to maximise fixed costs to further lower marginal costs? How do you even define fixed costs in contradistinction to marginal costs – there can be difficulties making the distinction in practice.

Web 2.0 platforms will often provide a much cleaner opportunity to trial these ideas. Take, for example, the Australian firm CultureAmp, which successfully sells browser based employee engagement software to large and small firms whilst also offering a cut down free version. Government could negotiate with CultureAmp and its competitors to fund one of them to supply a full suite of web-based services within a certain jurisdiction for free. (It might make more sense to do this for businesses within a certain size also as larger businesses may well require customisation).

One could imagine this driving small- and medium-sized business productivity more effectively than any number of programs putatively directed to such ends. One could extend this to more fundamental software services like accounting software. (In fact I’m surprised banks aren’t doing this now, but then the easy life isn’t the best way to drive innovation). In any event, in the digital age, the benefits are not simply greater access to the service and any productivity growth this may produce. Additional public goods would be generated. The new platform would generate a standard that would enable comparisons between firms and the generation of a rich data set that could help unlock our understanding of what practices and what government programs promoted business productivity and wider benefits. The existence of the standard would then create incentives for the best firms to open up their metrics.

Last Friday at a workshop of the Harper review of competition policy, I realised the argument extends further than this and that. One of the workshop participants argued that today’s new disruptive innovators are tomorrow’s monopolists and that, while one wanted to ensure there was adequate incentive for the initial disruption to occur, there was a case at some stage for some action against the monopolisation of the market. At this one of the participants argued that the original action against the great trusts under the Sherman Act was mostly not particularly effective in lowering costs. Even with barriers to entry in steel and oil, a monopoly will often not be able to raise prices much before there is entry. So given the economies of scale to which the monopoly leads, it ends up being quite efficient. And busting up businesses is quite disruptive and expensive.

But here it can be pointed out – it was pointed out – that in the digital world often the monopoly problem is not baked into the economics of production where there are low barriers to entry, but rather into network externalities. And there, it is often possible to devise what look to me to be surgically precise interventions to interdict the monopoly aspects of the business leaving businesses free to compete on the merits. For instance if it comes in the form of a standard that the monopolist controls one can require others to be given free access to the standard.

Thus in this column nine years ago I argue:

On the logic of the National Competition Policy we should impose an access regime on Microsoft. As with utilities, it should only apply to the natural monopoly aspects of the business. With programs like Word, that’s mainly the ‘standards’ in which files are written like .DOC in Word and .XLS in Excel. . . .

I’m amazed that no country has imposed an ‘access regime’ on Microsoft requiring it to open its file standards.

The other thing one would do is to open the ‘look and feel’ of the software. Thus a software manufacturer could build and distribute software that enabled anyone who had learned how to operate a dominant product to operate it in a similar way to the dominant product.

I’m not sure if how much if any of this logic can be applied elsewhere. I’m not sure how to impose an access regime on Google. For Facebook and Twitter (and Google) one might impose access regimes on various aspects of their operations, and in all cases I think there’s a good case for requiring firms to provide to users full, machine readable records of all data they have on them. The UK has gone some way to legislating such a right. This doesn’t deal with all their monopoly power in the way that I think the access regime I’ve proposed for Microsoft Office does for it, but it does improve competition, efficiency and data liquidity substantially.

One could also extend this to a ‘realtime’ right so that social networks could be forced to degrees of openness of operation so that I could access any ‘declared’ social media via a third party app that would generate the functionality required for me to network with my ‘friends’ via that app with the social network as the ‘back end’. This configuration existed in many early messaging apps which were accessible via open clients and no doubt remains today in some services.

This entry was posted in Economics and public policy. Bookmark the permalink.
Notify of

Inline Feedbacks
View all comments