Do not believe those who tell you that concentration or globalisation is inevitable. Think of your hairdresser.
Wherever you go, the oil market, the banking sector, and the car industry are dominated by a small number of firms. But wherever you go estate agency, or agriculture, and plumbing are not. In most countries, retail banking is organised into chains which operate on a national basis (except in the United States, where legislation was introduced to stop it). This is also true, but less true, of food retailing. And much less true if you are selling clothes.
This is evidence that industrial structure is the product of economic forces which are general in their geographic application yet are specific in their application to particular sectors. It is not generally a matter of historical accident or inevitable trends that are uniform in their incidence across the economy. Yet only recently have there been systematic attempts to identify these forces or to detail general explanations.
The size of the market is one important factor. Many services have to be delivered locally. So the effective market is really much smaller, and so more concentrated, than crude measures of shares of total output would suggest. No company has a very large share of retail newspaper sales, but that is not the point. Most customers have an effective choice of more detergents than of newsagents, even though there are very few detergent manufacturers and very many news stands.
Thirty car manufacturers, or ten brands of oil, or six hairdressers, provide all the choice that anyone would need. But it is cheap to ship cars and oil, but not haircuts, around the world. So thirty car manufacturers can provide not only all the choice that anyone could need, but all the choice that everyone could need. This is not true of hairdressers, and so there need to be many more hairdressers than automobile manufacturers.
So deciding whether an industry structure is concentrated or not depends on how you define the market. Perrier have a large share of the bottled water market, but a small share if you include tap water. And is the world pharmaceutical industry highly concentrated or not? On one measure no: even the largest drug companies, like Merck and Glaxo Wellcome, have less than 10% of total sales. But if you look at most particular therapeutic categories you will find them dominated by two or three products. Since the efficiency of any market structure is determined by a trade off between low costs and breadth of choice, the industry is probably concentrated enough. These are the issues with which anti-trust agencies wrestle all the time. And the same issues confront the business analyst who has to decide whether an industry is particularly fragmented or particularly concentrated.
But it is not just the balance between scale economies and market size that matters. Banking is dominated by large firms. But I could set up a bank in my garage with only a computer, a phone and a licence from the Financial Services Authority. And with no need of an expensive lunch room, well remunerated executives, or a branch network, its not obvious that my costs would be higher than Citicorp’s. The reason not many self-employed people go into banking is that credibility matters a lot in a bank, and credibility tends to go with size.
This is one example of a market which has a winner-takes-all property. You look for the most secure bank, the best (rather than the second best) drug for an illness, the credit card that is most widely accepted, and the operating system for your video recorder or your personal computer that has the largest range of software associated with it.
These winner-take-all markets tend to become concentrated. Microsoft dominates the personal computer world, Visa the credit card business, VHS (though not JVC, which devised the standard) rules the video recorder industry. That is why therapeutic categories (but not the pharmaceutical industry as a whole) show high concentration.
And there are other winner-takes-all, or winner-takes-most, structures. In some industries the largest research and development budget scoops the pool (perhaps less common than you might think, but an important factor in IBM’s former dominance). In other industries the largest advertising spend scoops the pool. That is why most fast moving consumer goods markets are relatively concentrated.
The strength of competition is also relevant. More intense competition tends to mean fewer firms. Weak competition protects weak companies; deregulation or decartellisation removes that protection. American airline deregulation illustrated that well. The immediate outcome was a wave of entry but in the end there were fewer firms than before. This will happen for European countries too: these trends are already evident. The financial services industry is evolving in a similar way.
And yet there are many industries to which none of these factors apply. Our picture of the industrial economy is often distorted by the nature of the quoted sector. If you put your faith in the FTSE 100 index, your view of the economy is skewed towards banks, pharmaceutical firms and oil companies. But if you want to invest in plumbing or farming, you cannot do it through the stock market, and your choice of estate agents is limited to a handful of small firms.
There are generalisations to be made about the principles of how the structure of industry evolves. But there are no general conclusions applicable to all industries. And do not believe those who tell you that concentration or globalisation is inevitable. Think of your hairdresser.