One of the reasons I’ve been belaboring the limitations of top-down management is that I’ve found this is a subject on which some libertarians get confused. Because political debates often pit governments against businesses, there’s a tendency for those of us who are skeptical of government power to idealize the operation of businesses. (And conversely, of course, those who are skeptical of big business have a tendency to idealize government agencies.) Probably the most forthright and extreme example of this is free-market economist Ludwig von Mises. Here he is in Bureaucracy (page 26 of the Liberty Fund edition):
The entrepreneur is in a position to separate the calcuation of each part of his business in such a way that he can determine the role that it plays within his whole enterprise. For the public every firm or corporation is an undivided unity. But for the eye of its management it is composed of various sections, each of which is viewed as a separate entity and appreciated according to the share it contributes to the success of the whole enterprise. Within the system of business calculation each section represents an integral being, a hypothetical independent business as it were. It is assumed that this section “owns” a definite part of the whole capital employed in the enterprise, that it buys from other sections and sells to them, that is has its own expenses and its own revenues, that its dealings result either in a profit or a loss which is imputed to its own conduct of affairs as separate from the results achieved by the other sections. Thus, the general manager of the whole enterprise can assign to each section’s management a great deal of independence. There is no need for the general manager to bother about the minor details of each section’s management. The managers of the various sections can have a free hand in the administration of their sections’ “internal” affairs. The only directive that the general manager gives to the men whom he entrusts with the management of the various sections, departments, and branches is: make as much profit as possible. And an examination of the accounts shows him how successful or unsuccessful they were in executing the directive.
In a large-scale enterprise many sections produce only parts or half-finished products which are not directly sold but are used by other sections in manufacturing the final product. This fact does not alter the conditions described. The general manager compares the costs incurred by the production of such parts and half-finished products with the prices he would have to pay for them if he had to buy them from other plants. He is always confronted by the question: Does it pay to produce these things in our own workshops? Would it not be more satisfactory to buy them from other plants specializing in their production?
He goes on in this vein for several pages, fetishizing data in precisely the same way that Robert McNamara did. Mises seems to believe that senior management doesn’t have to know very much about what his subordinates are doing because the firm’s financial statements will tell him everything he needs to know. Yet a firm’s financial reports (like any leaky abstractions) can hide nasty surprises. In the case of Wall Street circa 2006, firms were taking reckless bets that offered short-term profits at the risk of long-term insolvency—examining the accounts wouldn’t have revealed this fact.
There are lots of other examples. A manager might be deferring critical maintenance or reducing research and development spending. He might be cutting corners in areas like customer service that will damage the long-term reputation of the firm. He might be hogging shared corporate resources to make his own division look good at the expense of other divisions. A competitor might be about to introduce a disruptive innovation into the market that will destroy the division’s profitability.
Financial results are also difficult to interpret because firms operate in a dynamic and unpredictable marketplace. Suppose a company’s widget division lost money last quarter. One plausible explanation is that the guy in charge of the widget division was incompetent and should be replaced. But there are other possibilities. Maybe the price of widgets collapsed, and the widget division would have lost even more money if not for the hard work of its management. Or maybe the division’s expenses are up because it’s spending money on developing an improved widget that will sell like hotcakes next quarter. There’s no way to distinguish among these cases (or many others) simply by examining the company’s books. You have to actually spend time understanding the business and its place in the larger marketplace.
This isn’t to deny that private firms (even large ones) in competitive markets will generally perform more efficiently than monopolistic government agencies. But it’s the “competitive market” part, not the “private firm” part, that’s doing most of the work here. Private firms are more efficient than public ones not because the private firms are inherently better-managed, but because private firms are more likely to face effective competition that keeps them disciplined, and because inefficient firms are allowed to fail.
This has important policy implications. Mises’s argument suggests that we should be indifferent between an industry dominated by a handful of large firms or one comprised of many smaller firms. After all, the large firms is effectively just a group of “hypothetical independent businesses” that happen to be grouped under the same corporate umbrella. But Mises is wrong about this. A single large firm is fundamentally different from a bunch of small firms. The group of small firms is likely to (collectively) be more efficient and more innovative than the single large firm would be. So policymakers should prefer policies that promote decentralization and competition, and they should be skeptical of policies that give large firms too much control over sectors of the economy.