What Can Bosses Know?
Is it possible for bosses to really know about their own businesses and markets? How much foresight can they have? What tools do they have for acquiring such foresight? To what extent are bosses prone to the problems of bounded rationality, inherently limited knowledge or cognitive biases, as proposed by Herbert Simon, Friedrich Hayek and Kahneman and Tversky, among many others?
Journalists systematically avoid these questions. Every management failure is presented as an exception to a general pattern of competence. But is this really the case? Could it be that bosses have less control and foresight than they pretend, and that success often results not from management's deliberate interventions but simply from force of habit or the fact that success breeds success? Could it be that the power, importance and money we give chief executives is therefore unmerited?
I don't know the answers to these questions. But I do know that the dead trees rarely ask them. But then, you wouldn't expect journalists to challenge hierarchies, would you?
I believe the problems are inherent in hierarchy. As Samuel Edward Konkin III (SEK3) of the Movement of the Libertarian Left said somewhere (I can't find it--little help?) organizational inefficiency starts when you have one supervisor taking orders from another supervisor: that is, the point at which hierarchy replaces market contracting.
Under such circumstances, problems of moral hazard and imperfect knowledge, internal transaction costs, principal-agent issues, etc., become unavoidable. I've already written at length on the problems of incentives (or even tracking the effects of individual decisions in the first place) in large organizations, and the effects of hierarchy on information flow (see "On the Irrationality of Large Organizations" and "On the Superior Efficiency of Small-Scale Organization").
The central problem is that, since the costs of tracking the results of individual decisions becomes prohibitively expensive in a large organization, market incentives must be replaced by administrative ones. Milton Friedman pointed out long ago that people do a better job of spending money on themselves than on other people, and do better spending their own money than other people's money. That's the standard, and correct, libertarian argument for why government is so inefficient. It's spending other people's money on other people; and unlike a private firm not only can it not go out of business for inefficiency, it gets rewarded with more money. Well, the very same incentive problems apply to the decision-maker in a corporate hierarchy. He's a steward of other people's money, and the costs and benefits of any decision he makes can be determined only badly, if at all. Unlike a self-employed actor whose relations with others are mediated by the market, he is motivated by purely administrative incentives.
Internalization of costs and benefits of decisions is further hampered by the fact that administrative authority is separated from the actual performance of productive work. As a result, labor is likely to bear all the costs and inconveniences of increased efficiency, while owners and senior management recieve the rewards. A speedup or layoff, or combination of attrition and added workloads for the surviving employees, is likely to be followed by a big bonus to the CEO for "increasing productivity." The workers on the shop floor, on the other hand, are likely know the most about how to improve the work process. But they have no authority to restructure it on their own, and no reason to do so when somebody else will benefit at their expense.
And the flow of information within a hierarchy means that senior managent receives distorted or falsified information, and as a result the people doing the work receive irrational orders from above. As Kenneth Boulding put it, those at the tops of large hierarchies live in almost completely imaginary worlds. The only thing keeping the average large corporation going is that it coexists, in a cartelized industry, with a handful of other firms that share the same organizational culture and whose "competing" managements base decisions on the "industry trend." Market entry barriers enable a "Big Three" or "Big Five" to maintain a shared, pathological organizational culture without significant competition. Rather, the cartelized corporate firm and the large government agency become the hegemonic norm in society at large, so that even non-profits and mutuals adopt the standard form of management featherbedding, prestige salaries, mission statements, and a senior management dominated by resume carpetbaggers).
So what are the alternatives? There are two ways of internalizing the costs and benefits of every decision in an individual actor, and thus achieving rationality: 1) Replace administrative relationships with contractual ones within the organization, effectively dissolving the outer walls of the firm and replacing it with an internal market; 2) make authority within the organization flow from the bottom up.
The first alternative, the one that (as mentioned above) SEK3 found more congenial, also has a venerable history in the socialist strand of individualist anarchism.
Josiah Warren, in Practical Details in Equitable Commerce, proposed (in James Martin's words) "a system based on voluntary cooperation, but at no place rising above any individual within its structure..."
Society must be so converted as to preserve the SOVEREIGNTY OF EVERY INDIVIDUAL inviolate. That it must avoid all combinations and connections of persons and interests, and all other arrangements which will not leave every individual at all times at liberty to dispose of his or her person, and time, and property in any manner in which his or her feelings or judgment may dictate. WITHOUT INVOLVING THE PERSONS OR INTERESTS OF OTHERS.As he developed the idea further in Equitable Commerce,
If governments originate in combined interests, and if government and liberty cannot exist together, then the solution of our problem demands that there be NO COMBINED INTERESTS TO MANAGE. All interests must be individualized--all responsibilities must be individual, before men can enjoy complete liberty or security, and before society can be completely harmonious....
When one's person, his labor, his responsibilities, the soil he rests on, his food, his property, and all his interests are so disconnected, disunited from others, that he can control or dispose of them at all times, according to his own views and feelings, without controling or disturbing others; and when his premises are sacred to himself, and his person is not approached, nor his time and attention taken up, against his inclination, then the individual may be said to be practically SOVEREIGN OF HIMSELF....
He proposed a system for labor accounting within the enterprise, based on the severability of all individual interests and the coordination of all internal transactions by contract or exchange. This was, in part, supplemental to a point made by Thomas Hodgskin in Labour Defended Against the Claims of Capital. Hodgskin raised the question of how the labor-product could be apportioned among laborers engaged in work which was, by its nature, collective. The answer lay in the ability of each to withdraw his labor from the collective enterprise if his pay did not compensate the perceived toil and trouble of his labor. The lesson for worker self-management is that wages will be apportioned among workers within the enterprise through the "higgling and bargaining of the market." Warren supplemented this with the caveat that the problem should be avoided as much as possible by de-aggregating work units as much as feasible.
Of course, Warren's model applies mainly to an artisan economy in which the largest work units are small shops. It's of limited relevance for capital-intensive forms of production with larger work units, in which some steps in the production process by their nature involve collective work. At some points, in such a process, contractual relations must be supplemented by worker self-management as alternatives to traditional managerial hierarchies. But, as I will discuss in greater detail below, there's no reason that contracting and self-management can't coexist within a single production process.
It's something Jane Jacobs touched on in The Economy of Cities. Contrary to what technocratic liberals like Schumpeter and Galbraith had to say, the large corporation isn't much good at innovation. It's possible for an organization to do essentially the same functional job within a larger production process, either as a division within a vertically integrated corporation or as an independent firm contracting to supply a higher stage of production. For example, 3M might just as easily been an abrasive sand division with a larger metal castings corporation, rather than an independent firm doing the same thing. But it would have been a lot less likely to attempt sandpaper production as a sideline, or to use its failed experiments with sandpaper backing as the basis for a new venture in adhesive tape. Such side ventures would have been shelved as irrelevant to the larger organization's goals. For purposes of innovation, the ideal economy is one of small, specialized firms related contractually in a free market.
The second alternative, workers' control of production, internalizes the costs and benefits of decisions in the same people by making decision-makers directly responsible to those doing the work. It has a long track record of success stories, with worker self-management resulting in steep increases in productivity and morale, and decreases in absenteeism and scrap rates.
Unfortunately, the increased productivity is outweighed by the imperatives of social control within the workplace. For one thing, a self-managed work unit has a lot more bargaining strength: it's considerably harder to replace your workforce with unskilled labor in the event of a walkout when production is planned by blue collar workers on the shop floor. This was one of the central attractions of automated control systems, like numeric control technology in the machine tool industry. It shifted control of production from master machinists on the shop floor to white-collar engineers in the managerial hierarchy. Numeric control technology was developed largely by Army Air Corps/USAF money in the 1940s, by the way, and first introduced in the aircraft industry and its suppliers. For more on this, read Forces of Production, by David Noble.
And once workers get a taste of deciding the how, there's always a danger they might get the idea that they can decide the what and why just as well. If they can do a better job managing production without the foreman, they might start wondering whether they can do the jobs of the CEO, the vice presidents, and the Board of Directors better, as well. For that matter, they might decide they can do a better job than the wankers at Gosplan, in the Ministry of Non-Ferrous Metallurgy, and the branch ministry for copper. Experiments in worker self-management met the same reaction in the corporate West, as in the Leninist East (Lenin quoted Taylor to justify his suppression of the factory committees), for largely the same reasons.
Of course, there's no reason the two approaches can't be combined in a decentralized, bottom-up economy, with self-managed work units functioning both as firms in a market economy (producer co-ops), and as contractors carrying out each separate stage of production in what used to be vertically integrated corporations. In both approaches, the idea is to decentralize decision-making to the smallest feasible units; to replace hierarchical relations between such units with market relations; and thus to unite decision-making authority, responsibility for results, and direct experience of the consequences in the same persons.
presto wrote: Excellent post, Kevin. I spent nearly two decades working in the corporate world in quality control analysis, where my job was (supposedly) to gather data and analyze consequences of management decisions regarding the production process, and have found that in addition to the difficulties in gathering good information regarding management decisions, they tended to shoot the messenger (me) regarding bad news than using the information to make better decisions.
I later realized that there is a significant difference in a corporation between what your job description says and what your job actually is. My job was to tell middle management what they wanted to hear, and middle management's job was to do the same to senior management. Senior management's job is to do whatever it takes to have stock prices go up every quarter, including manipulation of data to make things look better than they are. Outright falsification is not that common, but picking the right measures to guarantee a rosy picture is. Accurate information was not wanted, especially if it interfered with the yarn that senior management was trying to spin to stockholders and major customers. I couldn't do that and look myself in the mirror, and so have quit working in a field that I was very good at. Gave up a lot of money, too.
Enron wasn't an unnatural oddity, it was the natural outgrowth of the corporate structure. CEO's are not rewarded for making good business decisions for the long-term health of the company if they hurt the stock short term, because the majority of large stock-traded corporations are owned primarily by institutional investors, who are only interested in short term stock price.
For one example, look at mergers. Mergers rarely benefit the companies involved in the long run, so why do they happen? Because the news of the merger boosts the short-term stock price of the acquired company and bonuses are paid to the senior management of both companies. Long term effects on the companies health are not an issue.
You are right, Kevin, that managers are not able to make informed decisions regarding the workings of the business, but that is not their real job in a large corporation. Their real job is pumping the short-term stock price. Stock price manipulation has been perfected into an art form, and the CEO's who are best at it are rewarded handsomely.
Sorry about being so long winded, but this goes to the heart of why I started the journey from Republican to Mutualist. Seemingly irrrational decisions are actually quite rational (but often immoral)if seen in their proper context.
thebhc wrote: "Such side ventures would have been shelved as irrelevant to the larger organization's goals."
I am immediately reminded of the history of the nascent company, Apple. Jobs, et al., had discovered that the Xerox PARC had developed a GUI for their operating system, which also made use of "mouse" that had been developed at Standford Research Center. Much to the chagrin of Xerox engineers (the low level workers), Xerox granted Apple access to this technology in exchange for some pre-IPO Apple stock. Xerox managers had little understanding or appreciation for what was being developed on their own research floor and obviously had no idea what it might mean in the long term. To them, a quick million in Apple stock for what was no doubt considered the worthless product of indulged geeks was a "no-brainer."
thebhc wrote: I quite agree with presto that, if Enron is anything, it is an excellent example of the inherent problems with the modern market. And the modern market is now considerably more volatile in the short term and it certainly no coincidence that the raft of enormous corporate accounting scandals that have taken place lately have occurred within such an environment. Pressures on stock gains now exist at the minute, or less, level. No one is thinking about six days down the road let alone six months. And years? Ha. This creates a very irrational climate with little or no thought in the long run. I always find it amusing that retirement investors will tell everyone to keep calm and take the long view when Wall Street itself can't do that.
[Note--This originally appeared in September of 2005 in Uncapitalist Journal. Since that blog is unfortunately offline, I decided to reproduce it here.]