Dominant companies may confer marginal benefits to consumers that are difficult, if not impossible, to estimate. However, antitrust authorities will do little wrong if they stick to the underlying principles of competition policy.
When Microsoft integrates new capabilities into its Windows software, it increases the functionality of its products. But it also makes it more difficult for competitive firms to offer innovative services. Do the advantages outweigh the detriments?
If General Electric owned Honeywell, would it use its position in the aircraft leasing market to increase its sales of aircraft components? And if it did, would its customers gain or lose?
Deciding these issues requires careful balance and fine judgement and anyone who is confident of the answers is a fool. Not even business titans like Jack Welch and Bill Gates know enough to see clearly how the markets they serve will evolve. And no economist has enough data to be certain of the cost benefit analysis. You can find economists, business people and lawyers arguing both sides of these cases, and different adjudicators come to different views. The Appeal Court’s view of Microsoft’s dominance differs from that of Judge Jackson – and the US Department of Justice and the European Commission came down on opposite sides in the GE-Honeywell merger.
How did we find ourselves in a position in which judges and public officials have to make decisions of this kind? Half a century ago, the administration of competition policy in the U S and Europe was often characterised by economic illiteracy and ignorance of basic principles of business organisation. It is less than 50 years since Britain’s Restrictive Practices Court held that a price-fixing agreement between makers of bolts and nuts was legal because it saved customers the trouble of shopping around.
There is a better understanding of how markets work in the application of competition policy today. And I need to declare a personal interest: I am paid to promote that understanding. But things have now gone too far. We are asking people to make judgements that no one is qualified to make. We cannot fine-tune the evolution of industrial structures to ensure that each market develops in the way most conducive to economic efficiency.
That does not mean that we should not have a competition policy. It does mean that we should go back to basics and remind ourselves what it is about.
The first anti-trust law, the Sherman Act, was passed in the U S in 1890. Its most important application came in 1911, in the break-up of Standard Oil and American Tobacco. That era was decisive in the development of modern capitalism.
The giant figures who dominated American business before the first world war – men such as John D. Rockefeller, Pierpont Morgan, and Jay Gould – genuinely believed that the concentration of major industries through a single, central organisation, the trust, was inevitable and desirable. Such an organisation would co-ordinate investment and guarantee supplies of key commodities, would impose rationalisation on inefficient and fragmented industries and would alone achieve the economies of scale that the huge US domestic market made possible. You hear similar arguments from many business people today.
The issue divided US politics and society. But the debate ended in a clear rejection of this model of economic development. The strongest motive was revulsion at the political and economic power that would be conferred on those who ran these businesses. They could not be expected to put the public interest before their business interests and it could not be assumed that the public interest and their business interests were the same.
And a century later, with so much experience of unsuccessful central planning behind us, we understand more fully why the decision to reject a concentrated structure of industry was right. Whenever a line of economic activity is controlled by a single entity, whether it is a grant firm, a government department or a trade association – we lose the vitality, rivalry and experimentation that are the essential features of a market economy.
The battle for pluralism in economic structures was keenly fought in the 20th century. Britain veered between the movement for “rationalisation” and support for competition. The militarily successful allies struck a decisive blow for pluralism when they broke up German cartels and Japanese zaibatsu. The motive was to make impossible the industrial complexes that had helped propel these countries into the second world war. The formation of the European Union represented a commitment to competitive markets, with neither laissez-faire nor state control as the guiding principle of the European economy.
That battle for pluralism has to be constantly refought. I am reasonably confident that a takeover of Marks and Spencer by Tesco or Sainsbury, or a merger of Barclays, BNP and Deutsche Bank, would be rejected by the competition authorities but I am equally confident that some aspiring Pierpont Morgan has analysed the cost savings and synergies and could garner political support for his plan.
Those who run competition policy must trust their instincts more and their models less. If the technical argument is close, it should always be resolved in terms of the underlying purpose of antitrust legislation. That purpose is to constrain the political and economic power of big companies – and to do so even if there might be marginal benefits in specific cases to consumers from allowing particular companies to extend their dominance. Favour competition over monopoly, entrants over incumbents and pluralism over concentration. If these are the principles, anti-trust authorities will rarely go badly wrong. And if these are to be the principles, there is not much doubt how the Microsoft or Honeywell cases should be decided.