The relationship between market power and innovation

David Havyatt

Column: ACCC chief Rod Sims has kicked off a debate on merger law reform in Australia. How valid is the claim he mentioned by the International Monetary Fund that further increases in the market power of already-powerful firms could deter innovation?

As reported here, Mr Sims has used a speech to the Competition and Consumer Workshop of the Law Council of Australia to seek wide-ranging merger and acquisition law reform.

Unfortunately, like much discussion of competition policy, this is a proposal before an audience of legal practitioners and merger experts rather than informed economists.

Sims described the speech as kicking off two debates; the appropriateness of the Australian merger control regime and ‘whether we want an open, innovative and competitive economy.’

Australia’s competition watchdog Rod SIms. Image: Twitter

What market conditions best promote innovation is not as simple as market concentration. Fans of competition often cite Schumpeter’s term ‘creative destruction’ from Capitalism, Socialism and Democracy and his belief that technological innovation associated with business cycles was the disruptive force that sustained economic growth.

Yet Schumpeter was no fan of artificial devices to create competitive market structures. Instead, in the same book, he advanced the hypothesis that large firms with market power accelerate the rate of innovation.

More recently, Clayton Christiansen, in The Innovator’s Dilemma, has advanced the theory that disruptive innovation comes from new players who are focused on new markets, while established players refine their focus on existing markets. He therefore makes a distinction between disruptive innovation and sustaining innovation.

As Joshua Gans notes, disruptive innovation takes two forms. The first is demand-side innovation of the type Christiansen identified, but the second is architectural innovation identified by Rebecca Henderson.

(Gans expands on this in his book The Disruption Dilemma. Interestingly, Gans is the co-author of a paper that does the kind of post-merger analysis that we will see later that Rod Sims is proposing.)

Now let’s look at Rod Sims comments on the M&A regime. Sims rightly points out that parties are not required to seek clearance from the ACCC for a merger (or acquisition), and even if they do, the ACCC’s only power is clearance.

If the ACCC wants to prohibit the merger, it must go to the Federal Court to stop the merger. (More technically, if a merger that lessens competition occurs without ACCC clearance or court authorisation, the ACCC can bring an action against the firm.)

Whether a notification to the ACCC is required, the ACCC will never be the final arbiter prohibiting a specific merger. That is a judicial action that under our system of government will only be allowed to be made by a judicial authority (as determined in the Brandy case.)

Moving from who has to be convinced to assessing the consequence, Sims notes that proving anti-competitive consequences ex-ante is difficult. He advises that his team is now doing ex-post assessments of the outcomes of some of the merger cases they have lost. In taking this action, Sims draws comfort from the conclusion of a Treasury paper that finds increasing mark-ups across the Australian economy.

However, the increasing mark-up data alone is insufficient to demonstrate harm. If the consequence of increased industry concentration is cost reduction from economies of scale, whether consumers and society are better or worse off depends on the combined effects of the increased productive efficiency (lower costs) and the declining allocative efficiency (mark-ups).

It seems that the ACCC Chair is advocating for a position that merger disallowance should occur at a threshold lower than the merger is likely to reduce competition; however, what the lower standard might be is not made clear.

An option would be not to change the merger law but instead for the ACCC to attempt to use the divestiture provisions that already exist for mergers that result in a lessening of competition.

There are two particularly troubling elements in the reasoning applied. The first is Sims’ argument that courts places ‘significant weight on evidence from the companies’ executives about their plans for the future, despite the extent of self-interest involved’.

Sims further argues that this can be difficult since ‘parties are careful to ensure that company documents do not contain any reference to the forecast effect of the merger.’ It is hard to reconcile this claim with the obligations on the Board at least to maintain appropriate records and act in the best interests of the company.

One of the examples cited is the TPG/Vodafone merger that the ACCC opposed. On this website, I described the ACCC’s opposition to this merger as ‘competition wished into existence.’ That column noted that repeated attempts (by placing competition limits on spectrum auctions) had failed to see a viable fourth operator materialise.

The second is Sims’ argument that overseas jurisdictions have an easier path to stopping mergers because courts look at the structural conditions for competition and start from the assumption that increased concentration will result in a greater likelihood of enhancing market power.

Yet the market Sims is most concerned about is ‘digital platforms’ where the merger activity is global, and these supposedly more favourable US rules apply.

Is there something fundamentally different about business in the 21st century that makes industry concentration not only likely but also desirable? The short answer is yes; it is the feature captured in digital platforms described just before the start of the century by Shapiro and Varian in Information Rules as ‘demand-side economies of scale’ (and is also referred to as two-sided markets).

Demand-side economies of scale arise when more people buying a product (say a VHS rather than Beta VCR) results in a growth of services that benefit the product owners (more films on tape in that format).

Economically it has the same consequence as the network effect with the telephone; the product is of more value to an individual consumer the more other consumers use it, but the cause is what happens in the related market.

I observed that the nature of industry is changing in a presentation to the ACCC Regulatory Conference way back in 2002. I noted ‘the nature of many of the industries of the information age represent a greater potential for economies of scale, be they demand or supply-side economies, and the greater potential therefore for near-natural monopolies to emerge.’

The ACCC really can’t wish competition into existence. The nature of 21st-century industry is that it will be concentrated and that markets will very rapidly tip into a monopoly.

A corollary of this observation is that regulators may need better powers to develop access regimes so that other technology providers can innovate over the top of the existing dominant firms. The proposal to restrict the monopoly the two app stores exercise over app payments is an example of this kind of access.

Another example would be the equivalent of any-to-any connectivity to mandate ‘interconnection’ of messaging platforms (For example a Twitter DM (direct message) addressed to a LinkedIn address).

As we have seen above, sustaining innovations are more likely to be made by large incumbent firms. When one firm thinks it can gain a greater share of the industry mark-up by enhancing its products or reducing its costs, the incentive to innovate is strong.

Disruptive innovations are more likely to come from small firms and new entrants of various kinds. These innovators face one of three potential futures in a world of scale economies on both the demand and supply side; dominance, cash out or get out.

The first case is that their innovation is sufficiently novel that they can develop into a dominant firm in their sector. Despite the dominance of some platforms, new platform sectors continue to emerge, such as AfterPay and Canva.

The second is that they are acquired by one of the large platforms, and their offering is integrated with the larger platform.

Sticking with Australian developments, Where 2 Technologies and the tiling technology developed by the Rasmussen brothers is core to Google Maps.

But the Google Maps product required other acquisitions, most notably geospatial data visualisation technology from Keyhole (from which came the Keyhole Markup Language or KML).

The third is that they never achieve scale and either drift with a small market share, substantially reducing industry concentration, or collapse. One of my favourite merger cases was the ACCC action not to oppose the acquisition of Australis Media by Foxtel.

The ACCC thought a sustainable competitive local call market could arise. As the calling competitor was Optus Vision, the ACCC commenced proceedings in the Federal Court to stop Foxtel’s proposal to acquire Australis because of its impact on OptusVision and hence local calls.

The response was worked out at a meeting at Malcolm Turnbull’s home that (from memory) included at least Turnbull, Lachlan Murdoch, Austar chair Don Hagans, and XYZ chief executive Patrick Delany together with a smattering of lawyers and your humble correspondent.

The conclusion was that it was better to let Australis fail and the various assets to be bought piecewise rather than defending or modifying the merger proposal. Less than a month after the ACCC’s action, Australis announced it was insolvent and would be in liquidation within weeks.

If over-zealous merger and acquisition laws close down or restrict the second pathway for innovators, the incentive to innovation is dramatically reduced. Thus, the thesis that less concentrated markets and more restrictive merger and acquisition laws would promote a more innovative economy is simply false.

David Havyatt is a former adviser to Federal ministers and a longtime observer of Australian innovation policy.

Do you know more? Contact James Riley via Email.

Leave a Comment

Your email address will not be published.

Related stories