[From Mike Acree (2003.08.26.1110 PDT)]
Rick Marken (2003.08.02.0850)--
Alice McElhone suggested that someone put together a conference
proceedings.
After less than a month, I'm getting around to sending a draft of my talk. I'm posting it to the Net, in case feedback on it may prompt me to withdraw it altogether.
OPEN- AND CLOSED-LOOP HUMAN TRANSACTIONS
Michael Acree
On a routine dental visit last May, my dentist asked me to schedule another appointment for a check-up and cleaning in 4 months. I declined because, as I reminded him, my insurance covered only two visits a year. There was a time when I thought of myself as conscientious in going once a year-and, in fact, there has never been any indication that more frequent visits were needed. Now, I go every 6 months because it doesn't cost me any more. But, as I left the office, I was reflecting on this distortion in my consumption. When I pay for dental services directly, I can be a "good consumer" in the sense of controlling value received for money spent. Third-party payments sever that loop, so that increased consumption has no effect on my checkbook. My perception stays the same regardless of my behavior.
It might look as though the concept of insurance were the problem. But carrying insurance on my house doesn't tempt me, to any appreciable degree, to indulge in more fires or floods. (The degree to which I may be more careless when I'm insured, as Dave Rothman pointed out during my talk, is referred to as the "moral hazard" of insurance.) The real issue, I think, is that we don't ordinarily insure against routine events, like having to reroof or repaint the house. Insurance companies would be willing to write such policies if there were a market, except that such policies would be very expensive and totally unnecessary. It makes much more sense just to do such things when we need them and can afford them. If such policies were nevertheless to become routine, we would all be painting and reroofing on a regular basis, and the increased demand would lead temporarily to higher prices, attracting more people to the field until prices were driven back down again. Th!
e net result would simply be a greater allocation of resources to these areas than before, with correspondingly less to be spent elsewhere-except that, being handled through insurance companies, which take their cut, the process would be less "efficient" in terms of giving consumers their money's worth.
It doesn't work quite this way in dentistry or medicine, since prices in those fields are capped. The increased demand without increased supply entails shortages, so we get medical treatment on the factory model-2 hours of waiting to see a doctor for 5 minutes. (After the CSG conference, the European edition of the New York Times carried an article (www.nytimes.com/2003/08/12/international/Europe/12WALE.html) about 600 people lined up to get one of 300 advertised appointments to see a National Health Service dentist in Wales. NHS dentists, according to the article, see 30 to 40 patients a day, compared with 12 in the U.S.)
It's interesting to consider how we should have gotten started on insurance against routine care in the first place-a case of a solution to one problem causing another. In World War II the government had introduced wage and price controls; and businesses, unable to compete for labor by offering higher wages, started adding benefits packages. After the war, people came to expect health benefits as part of a job offer; and 20 years later Congress decided to extend the idea as a national program. In some respects the U.S. has lagged behind other countries in implementation, where the effects are correspondingly more visible. Prime Minister Trudeau of Canada notoriously flew to the United States when he needed an operation, rather than wait 6 months for a Canadian hospital, and the hospitals of North Dakota are full of Canadian patients who have the money to go outside the system. There was a wonderful story this winter of a British doctor who sued a hospital for fo!
rcing him to perform an operation with a dessert spoon (BBC NEWS | UK | England | 'Spoon' surgeon loses case). He had to go out and spend $400 of his own money to buy the proper instrument. The government found against him, of course, contending that other doctors had operated with a spoon, and that he had moreover behaved unprofessionally in throwing the spoon on the floor in disgust.
These thoughts, as I say, led me to reflect on issues of control in various sorts of human transactions. Take first the relation of buyer and seller. Buyers are generally controlling for something like value for money, and sellers, reciprocally, money for value. (Whether the relation is ratio or difference doesn't really matter in the present context; so long as they are pulling in opposite directions, a compromise price will need to be negotiated.) We have to say "generally" because other considerations often intervene: We pay more for milk at the corner grocery because we don't want to drive to the supermarket for one item. Or, perhaps more appropriately, we could say that the convenience of the convenience store was part of the value we were controlling for, not just the price of the milk. In some markets a high price itself becomes a value: When I worked at Café Mozart, a customer once told me, I suspect correctly, that we could double our business if we !
doubled our prices. But these are exceptions.
Judging value for money usually requires a context, a range of potentially substitutable goods or services and prices. Again there are exceptions: If I want a monopoly good, like a new copy of Behavior: The Control of Perception, I may have to pay what is asked or do without. But even there substitutes are usually available, like buying a used copy or making an illegal photocopy.
Prices convey some information to us about costs, even if the connection is not tight. A large profit margin is just what many people make it their business to spot, so they can move in, get a portion of it, and reduce the margin for all. When these processes are operating, some degree of control is possible for buyers seeking the most for their money. Similar considerations apply to sellers. Sales, or changes in sales, across various prices give them information about what they need to do to improve their return, and hence to control their own perceived profits within the constraints of consumer demand.
We can contrast this arrangement with an open-loop transaction. An open loop I'm thinking of as a control loop which has been cut, in some way, so that it no longer functions as a control loop: The perception stays the same regardless of the behavior, and action becomes irrelevant. Whether a loop is open or closed has implications not only for control but for knowledge. Powers has made clear how knowledge depends on control. Fundamentally, that we reach limits in our efforts to control is our principal evidence for an external world. If we know a particular loop is being held open, we at least know that our perceptions are uninformative. Otherwise, we may be led to a distorted view. A barometer reading of 28.7 tells us something about the weather on the way-unless someone has set the barometer to read 28.7, in which case it says nothing at all.
If prices are held fixed, like the barometer reading-if a third party says, "If you sell at any price higher (or lower) than $6, I'll kill you"-then the prices contain no information about costs or demand. Consumers can still choose to buy or not, at the stipulated price; and sellers can determine whether they can afford to stay in business or not; but neither has the information to judge how well she or he is doing, nor to exercise control beyond the dichotomous decision of buying at the fixed price or not, or going out of business or not. Resources can be depleted, or inventories can pile up, when the negotiation connection between buyer and seller, which conveys information both ways, has been severed. Ultimately, it is difficult to keep an economy going without the information exchange mechanism of negotiated prices, as the productivity of command economies attests.
Interestingly, when government intervenes in the market, it most often does in the direction of keeping prices higher than they would be on an unregulated market. This was the case with airline fares before deregulation, and is still the case with commodities like sugar, dairy products, and labor. In these cases, production exceeds demand at the elevated price. With dairy products, the government buys the surplus, and maintains a stockpile-currently over 1.28 billion pounds of dry milk alone, over $1 billion worth (www.twincities.com/mld/pioneerpress/news/local/6557034.htm), not counting storage costs. With labor, the surplus shows up as elevated unemployment rates. If a price is held artificially low, of course, the result is underproduction relative to demand. There is a chronic extreme shortage of human organs since the price is legally set at zero. The lack of control by both consumers and producers is clear enough.
But there's a familiar saying that, when government opens a loop, it closes another one somewhere else, and we can see that process in action in the examples above. The bureaucrats and politicians who make the regulations are controlling primarily for keeping their jobs and budgets-for reappointment or reelection. In contrast to the situation with business, keeping their jobs doesn't mean pleasing consumers, but their funding sources-campaign contributions or Congress. (This is the perspective of public choice theory, for which the seminal text is Buchanan and Tullock, 1962.) Since the institution of ("temporary") price supports in the New Deal, the dairy lobby, for example, has been one of the largest and most powerful. Like most government programs, dairy price supports are a case of concentrated benefits and diffuse costs. The extra amount each of us pays is hardly worth complaining to our Congressperson about. The fact that dairy products are about twice w!
hat they would cost on a free market can make a difference to a family with several adolescents, but few of them probably even know about the legal floor on prices. If they were better informed, there is still little they could do about it; they are unlikely to be able to mount a larger campaign than the dairy industry, to whom the issue makes an enormous difference. And similarly for all sorts of commodities and services, from sugar to psychotherapy, for which prices are held artificially high by law.
Consider now a nonconsensual action: I've stolen your car or sold you tainted meat. In the Anglo-American tradition of 1000 years ago, it was understood that you would now have a claim against me. If we didn't resolve it between us, we would be brought together by an intermediary mutually agreed upon, a court or an arbitration agency more generally, which would determine the restitution I owed you (or your survivors, in the case of murder). Restitution would generally be designed to cover the direct loss incurred, plus some compensation for the inconvenience. If it were set too low, thefts would increase; if it were set too high, people would leave their cars with the keys in them in the hope that they would be stolen.
This kind of set-up is at least one way of closing the loop for the victim, allowing that person, and by implication the rest of us, some degree of control of security in our person and property, as well as of the perception of justice, of wrongs righted. The loop is not closed for the perpetrator, since the system does not generally allow control of the perception of getting something for nothing.
Although the merits of such a restitution-based system of justice would seem obvious, it wasn't long before kings realized they were missing a huge opportunity for expanding their wealth and power. If actions like murder were defined as crimes against the state instead of against persons, then kings could collect the fines and administer punishments. The system remains closed-loop for society in general in the sense that actions against others still entail costs. But it is open-loop for the victim, who gets nothing, and may even be forced, through taxation, to pay for the perpetrator's incarceration. Nothing the victim can do will make any difference; the circuit is broken. This system has worked so well, for the state, that it has been used throughout the West for the last 800 years.
Government, as Bill Powers has argued, is based on punishment, or the threat thereof. He has also analyzed the limitations of punishment as a basis for social organization, in terms of the resistance it invariably arouses. I have argued in addition that punishment, as a means of controlling the perception of justice, creates a new relationship between the perpetrator and the enforcing agent, while leaving the victim's error uncorrected. There are also many cases nowadays where the act being punished has itself no victim-the War on Drugs is an obvious example-and the action by government simply creates an uncompensated victim.
In general, I think it can be established that all transactions involving government entail opening loops. That's actually no surprise, when you think about it; the whole point of government is to keep people from controlling what they want to control. I hasten to add that this analysis is purely neutral and descriptive; each of us will have different notions of which loops we want open and which closed.
It is characteristic of government operations in the first place to be a legal monopoly-police, fire, postal service. The open-loop character of these enterprises is apparent. If I'm dissatisfied with my appliance repair service-if they don't respond promptly to calls, for example-I take my business elsewhere and correct the error. If I'm dissatisfied with police services, I'm stuck. Dissatisfaction is not unknown. Indeed, anybody who has anything substantial to protect-banks, department store, even supermarkets-has to hire private security because police protection is inadequate-and they have to pay twice since they have to pay for police whether it does them any good or not. I can make a written complaint; but, because of sovereign immunity, I can't even file a lawsuit unless the government agrees to it. And the police are aware that keeping their jobs is not dependent on satisfying customers. Indeed, the U.S. Supreme Court has ruled (starting with South v.!
Maryland, 1856) that the police have no obligation to protect us. The police control for satisfying their funders-City Hall-but the loop back to consumers is tenuous and indirect.
I can illustrate the process in more detail in an area where I've worked: research. The Center for AIDS Prevention Studies at the University of California, San Francisco, was created in 1986, and has a current annual budget of around $25,000,000. As far as I can tell, from 7 years of working there, nothing has been accomplished. HIV rates increased dramatically in San Francisco, our own backyard, in the late 1990s. In some ways that's not surprising at all. Nobody knows how to change the kinds of behaviors that are relevant, like condom use and needle sharing, especially on a mass basis-and they're not likely to find out, given the theories of behavior and of psychological research that they're using. Whatever consumers there might be for AIDS prevention programs, they're not getting their (tax) money's worth. We might expect private funders to require some results for their investment, but public funding could well continue indefinitely. The constituency CA!
PS is serving-whose satisfaction they are controlling for-is their funders: Congress and agencies like NIH. What Congress and the granting agencies need to be able to say to their constituencies is: "Look, we're giving lots of money to the very best people and institutions we can find. After that, it's out of our hands." And indeed they do a creditable job of selecting good people to support, by conservative, orthodox criteria; and those people are working earnestly. But when they are not succeeding, which is all the time, they simply need to remind the funders that the problem is large and difficult, and they simply need more money and time. Indeed, they need to be careful not to solve the problem outright, or their jobs would be cut. It seemed to me that there was a distinct air of panic in the room on the day in 1996 when protease inhibitors were announced. In fact, the requirements of continued funding lead to describing every situation as a crisis. I was assi!
gned for awhile to an expensive HIV prevention project in Peru, where
you hav
en't heard much about an impending epidemic. The HIV prevalence was estimated at 0.25%; a recently published survey of 360 women in family planning clinics in Lima found exactly 0 cases. It's hard to power a study to show a reduction in those rates, so the target of the intervention had to be extended to all STDs.
The Osher Center for Integrative Medicine at UCSF, where I work now, is another interesting example. It was started 4 years ago with seed money from a private philanthropist, Bernard Osher, who also gave $10 million to Harvard for a similar center. There is a lot of important work potentially to be done by such centers-but it won't be done, at least by them. There are hundreds of herbs and supplements, and alternative cancer treatments, which have been the subject of only small-scale research, often with mixed results; and they would be the obvious subject for more definitive studies at the Osher Center. But, as a matter of explicit policy, these are not a priority for the funding agencies, including the National Center for Complementary and Alternative Medicine. Money is going instead to psychological research-on yoga, meditation, and the doctor-patient relationship: to complementary, rather than alternative, medicine, in a word. I don't assume that these are!
the personal priorities of the bureaucrats in charge. I do believe they're controlling for keeping their positions and their budgets, which means pleasing Congress. It is well known that the pharmaceutical industry is one of the largest and most effective lobbyists, and that they have been working hard to destroy the supplement industry. Research which showed any of these substances to be effective would not be welcomed by the drug companies. Government-Congress and research funding agencies-thus establish their own loop, controlling for re-election and retention of positions and budgets.
In situations like these, when we see chronically ineffective, wasteful programs, or where we see actually destructive programs, it is useful to look for which variables are actually being controlled, and which presumed loops are actually open, with errors held artificially low. A good example of a policy is that is commonly deplored as destructive is the clear-cutting of forests. The result is not only unsightly, but, far more importantly, the practice destroys the land through greatly accelerated erosion; the eroded soil also frequently clogs streams, destroying the fish population. How could anybody be so stupid-especially the lumber companies who depend on future harvests from that land? The answer is that they're not destroying their own property. Clear-cutting occurs on land owned by the government, which grants short-term lumbering rights to companies. The errors from clear-cutting don't loop back to those performing the action; "public" ownership of the land di!
ffuses the effects much more broadly.
So why would government implement such a policy? Employees of the Forest Service are not controlling primarily for preservation of forests or land, but for keeping their jobs and budgets, which ultimately means pleasing Congress. Members of Congress aren't controlling for taking care of forests, either, but for retaining their own jobs. That means answering to their lobbyists (they insist otherwise, but a recent study by the Associated Press (www.foxnews.com/story/0,2933,92473,00.html <http://www.foxnews.com/story/0,2933,92473,00.html>\) found Congressional votes highly predictable from campaign contributions)-among whom the lumber companies are prominent.
A similar analysis applies to Western rangeland, much of which has been destroyed by overgrazing. The overgrazing occurs primarily on land administered by the Bureau of Land Management, which grants short-term grazing rights to ranchers. Environmentally conscious ranchers know that if they hold back the number of animals grazing on their tenure, the next leaseholder will not, and their sacrifice will gain nothing. The chronic crisis of land destruction means job security for the BLM; there are always more problems to be addressed, requiring more money.
In situations like these, there is a straightforward solution, namely the elimination of "public property." Converting government ownership to private ownership eliminates the tragedy of the commons. In other situations, like overfishing of the oceans, private ownership has yet to be defined.
In some cases where we see errors held artificially low, that is the express purpose of government policy. The various devices of income redistribution are a good example. The effects of socialism are particularly clear in the circumscribed environments of the first English colonies in the New World. Sponsors of new settlements commonly prohibited private property because they feared they wouldn't get their money back. The first Jamestown settlement starved to death, and the Massachusetts Bay Colony nearly did, until Governor Bradford set aside the colonists' contract and permitted private ownership (Bethell, 1998). There was unlimited land to be cleared and planted, and animals to be killed, and fish to be caught. But since the fruits of any such labor had to be shared with everyone else, the difference in how much any worker got to eat was negligible. Perceptions stayed essentially the same, in other words, regardless of behavior. Socialism cuts the loop. !
The effects are more gradual and less obvious when an existing system adopts redistributionist policies. Decades ago, before the introduction of such policies, Sweden had the highest per capita income in the world. Now the system survives partly through its loopholes: Wealthy Swedes who can afford to travel abroad buy as much as possible overseas, whereas those too poor to travel are stuck with a large value-added tax.
Clearly, a PCT analysis doesn't specify which values are to be controlled. Survival itself can even be a secondary consideration to the desire to hold certain other kinds of errors low. But PCT does offer a useful tool for the analysis of social systems, in directing our attention to what perceptions are actually being controlled in a given arrangement, where loops may be held open and errors held artificially low.
References
Bethell, T. (1998). The noblest triumph: Property and prosperity through the ages. New York: St. Martin's.
Buchanan, J. M., & Tullock, G. (1962). The calculus of consent. Ann Arbor: University of Michigan Press.