69. Selfishness, Altruism and Rationality, Part 2

Note: This post is the sixty-ninth in a series about government and commercial ethics. Click here for the full listing of the series. The first post in the series has more detail on the book 'Systems of Survival' by Jane Jacobs which inspired this series.

This week's topic is a continuation of last week's post on the book, Selfishness, Altruism and Rationality, by Howard Margolis

In the last post, we talked about how Margolis explained the failings of the traditional rational choice theory to explain human behaviour in situations where self-interest conflicted with group interest. This is a pretty common observation, generally taken for granted outside of economic circles, but Margolis goes a step further and proposes an alternative model of human motivation.

Margolis calls his model the Fair Share model and it is based on the notion that people feel a desire to 'contribute their fair share' to the public welfare. He describes the underlying motivation of people in this model as follows:

"The larger the share of my resources that I have spent unselfishly, the more weight I give to my selfish interests in allocating marginal resources. On the other hand, the larger the benefit I can confer on the group compared with the benefit from spending marginal resources on myself, the more I will tend to act unselfishly."

Margolis imagines that a person (who he calls 'Smith') contains two separate components, 'G-Smith' who values (Smith's perception of) the general welfare, and 'S-Smith' who values only Smith's personal welfare. Smith stays in equilibrium by adjusting the level of his spending on the public interest so that the marginal value of more public spending by Smith (to G-smith) equals the marginal value of more selfish spending (to S-Smith).

Margolis argues that, from an evolutionary point of view, it would be easier for this sort of limited, 'fair share' altruism to be maintained over time, because it would be less vulnerable to being exploited by selfish people than an unlimited altruism that didn't keep track of how much a person had already sacrificed their personal interests for the public good.

Margolis further notes that,

"The notion that human beings might have the kind of dual preference structure posited by this study is very old, going back at least to Plato's distinction between man as a private individual and man as citizen."

In chapter 6, Margolis goes into more detail on how his model differs from the classical rational choice model, and helpfully unpacks some of the assumptions which are embedded within the rational choice model (but often go unstated or unnoticed):

1. Smith can be treated as narrowly self-interested
2. Smith's utility function is a goods function (i.e. he only cares about what goods people possess, not how they got them or what role he played in determining the allocation)
3. Smith chooses in conformity with the principle of consumer sovereignty (i.e. Smith thinks what's best for society is that everybody get what they wants, as opposed to Smith having a vision of what's best for society which might conflict with what other people want).
4. Smith has only one utility function (as opposed to having one for his own interests and one for the social interest).

As Margolis explains, economists, if confronted with these assumptions might deny that they are a necessary part of the model, but after their denial, they will then go right back to building models and making predictions that only make sense if those assumptions are there.

He also explains how, in the marketplace, where the public interest and private interest are in alignment (subject to all the caveats he have discussed in this series), the difference between the predictions of his 'fair share' model and a traditional rational choice model that posits self-interested behaviour by all participants is not that big. It is primarily in political situations where the differences will be clearer, because here the contrast between private and social objectives is sharper.

If you've had the same struggles as I have over the years trying to pin down how economists come to the (often wrong-headed) conclusions they do, this chapter is a must read. Margolis is that rare bird who knows enough economics to be able to explain things clearly using the language of economics but has still retained enough common sense to be interested in models of people as they actually are as opposed to making unrealistic assumptions so as to have a model that is easier to work with mathematically.

---
Later on, Margolis talks about how his fair share model explains why people might act differently in different circumstances, pursuing their own interest in one and the public interest in another,

"If I am a producer facing reasonably competitive markets (even the experience of Ford in trying to promote safety features in automobiles in the mid 1950-s is instructive here), then I will scarcely be in a position to do anything very different than produce what the market seems to want. Even if my choices affect only me and my customers, I will not have any customers to benefit unless I offer them things they want at a price they are willing to pay. If there are external effects (environmental side-effects), the dilemma is even worse.

However, if I am in a senior position in my government, my decisions on public matters often affect society in a large way. This being so, there will be no necessary inconsistency between my behaving as a narrow profit maximizer (to a good approximation) as a private businessman; as a rather casual decision maker, as a voter, and as a very serious decision maker, working very hard and feeling great personal responsibility for the social effects of my decisions as a high public official.

...I wish to say enough here to indicate why [James] Buchanan's and [Gord] Tullock's 'paradox of bifurcated man' seems, from the [Fair Share model] view, to reflect a mistaken assumption that an internally consistent model could not account for a disposition for the same individual to behave in a very public spirited way in some circumstances and as a profit-maximizing economic man in other contexts."

Now this just seems like common sense to me, but then consider the surprise in the reactions of experimenters when they found, exactly as Margolis and his model would have predicted, that when they ran the exact same Prisoner's Dilemma experiment on the same people and only changed the name (in one case 'Wall Street Game' in the other 'Community Game') that people behaved very differently. As the abstract states,

"The results of these studies showed that the relevant labeling manipulations exerted far greater impact on the players’ choice to cooperate versus defect—both in the first round and overall—than anticipated by the individuals who had predicted their behavior." (emphasis added).

My one disagreement with Margolis in the passage above is that after stressing the role played by competition in preventing public interested behaviour in the marketplace, he then fails to note how the lack of competition in the public sector is an essential component of allowing the pursuit of the public interest there (although to be fair, it is somewhat implied in the text).

Finally, Margolis offers some interesting speculation on how caste might be partially explained by the fair share model. In the early stages of society, people who are more disposed towards public action would be more willing to undertake key tasks such as organizing irrigation schemes or a defensive army. In successful societies, these actions lead to large gains for the whole society, and those who were among the early organizers of the action would claim some of that gain for themselves, leading to greater wealth and influence. But under the fair share model, the more wealth you have, the more resources you will donate to the public interest. Richer people have more resources to donate in the first place, plus they donate a higher proportion of their resources, so there is a positive feedback whereby people with more power put more effort into the public realm which gets them more power in return and so on.

As Margolis says,

"As generations pass, the resulting division between those who manage and defend the state (often enough at real personal cost and risk) and those who labor comes to seem to accord with the natural order. What gives that presumption special potency is that there is some substance - something more than a self-serving myth - in the presumption that the noble and commoner are motivated in different ways. In terms of [the Fair Share model], that presumption is false at its root [because all people have an interest in their own welfare and the public welfare] but nevertheless consistent with observed behaviour. Our modern colloquial usage of words like 'noble' and 'peasant' is an anachronism but not necessarily a libel."

If you like economics but generally find economists irritating, and you are interested in the public welfare and the interaction between the two topics, this is a great book.

68. Selfishness, Altruism and Rationality, Part 1

Note: This post is the sixty-eighth in a series about government and commercial ethics. Click here for the full listing of the series. The first post in the series has more detail on the book 'Systems of Survival' by Jane Jacobs which inspired this series.

This week's topic is the book, Selfishness, Altruism and Rationality, by Howard Margolis

Margolis' goal in this book is to extend the Economic Theory of Rational Choice so that it covers political situations as well as Economic ones.

He opens the book with a quote from James Coleman which eloquently outlines the problem, while also covering our now familiar choice between two versions of self-interest,

"Classical economic theory always assumes that the individual will 'act in his interest'; but it never examined carefully the entity to which 'his' refers. Often, as when households are taken as the unit for income and consumption, it is implicitly assumed that 'the family' or 'the household' is the entity whose interest is being maximized. Yet this is without theoretical foundation, merely a convenient but slipshod device. In this case, as in many others (e.g. when a man is willing to contribute much, even his life, to national defense, rather than use a strategy that will push the cost onto others), men act as if the 'his' referred to some entity larger than themselves. That is, they appear to act in terms, not of their own interest, but of the interest of a collectivity or even of another person. Indeed, if they did not do so, the basis for society could hardly exist.

Yet how can this be reconciled with the narrow premise of individual interest ... we could simply solve the problem by fiat, letting 'his' refer to whatever entity the individual appeared to act in the interest of. This would obviously make the theory trivially true, and never disconfirmable. A more adequate solution is one which states the conditions under which the entity in whose interests he acts will be something other than himself."

We saw in the last post that James Buchanan was willing to settle for a theory that based human motivation solely on the desire for material gain, arguing that the desire for material gain is always present to some degree in people.

But Margolis isn't willing to settle so easily,

"A satisfactory theory of social choice requires a model of individual choice that is consistent with the way human beings are observed to behave. Yet, even after a generation of work on the problem of applying the economic 'rational choice' perspective to social choice, often leading to striking results, this fundamental problem remains unresolved. We still lack a model that accommodates (without fudging) such obvious observations as that citizens bother to vote and do not always cheat when no one is looking. A resolution of this difficulty can be expected to require some departure from conventional assumptions."

Margolis goes on to indicate that, in his opinion, the main difference between situations which can modelled fruitfully using the traditional model and situations requiring a new model is that situations where the old model works are economic in nature whereas situations where a new model is required are political in nature (echoes of Mancur Olson specifically indicating that this theories on collective action only applied to economic groups, not groups formed for no-economic reasons.

Says Margolis,

"This classical model is profoundly shaped by its root concern with the problems of the marketplace. But in politics we are dealing with goods allocated largely through some coercive process, not through voluntary market transactions; and political 'goods' (such as justice) are often inherently unmarketable. Nonmarket effects (externalities) which are aberrations - market failures, which one seeks to correct - for most economists are the central feature of political life for political scientists.

We can expect that Samuelson's notion of public goods (which can best be understood as a generalization of the notion of externalities) would play a central role in any viable formal theory of politics, and indeed that is the case. It is not too strong a statement to say that societies, and hence politics exist because public goods exist."

Margolis spends a chapter illustrating his argument that the classical rational choice models fails to handle political situations via a series of 3 examples:

* Voting
* Repeated Prisoner's Dilemmas
* Public Goods

In the case of voting, the rational choice model fails to explain why people might go the trouble of voting even when they know their vote won't affect the outcome.

In the case of the repeated Prisoner's Dilemma, the model fails to explain why people will generally cooperate even though on any given iteration they could gain by defecting against the other player in the dilemma.

In the case of public goods, the model fails to explain why people will make contributions to things that are publicly available to everyone. Margolis asks us to imagine a hypothetical man named Smith who is planning a $10 donation to his favourite charity. The classical economic model says that Smith would do this because he wants the charity to have $10 more available to it than it does currently.

But now imagine Smith finds out that someone else has just donated $10 to the charity. Under the classical model, Smith, realizing that his favourite charity is now $10 richer just as he wanted it to be, no longer feels a need to make a donation.

Of course in reality there may be some relationship between how much money a charity has raised and how much people contribute, but it is nowhere near this strong a relationship. Clearly there must be something more to Smith's motivation than simply wanting the charity to be $10 richer, but the classical model has no answer to what that might be.

Margolis argues that there are two altruistic motivations that need to be taken into consideration. We have an altruistic motivation based on wanting other people to have more, and an altruistic motivation based on wanting to contribute our fair share (what Margolis calls 'participation').

Margolis also notes that all of his examples are prisoner's dilemma type situations, which is not surprising since the Prisoner's Dilemma is the formalization of situations where what is in the self-interest of participants is opposed to the group interest.

In the next post we will look at the solution that Margolis proposes in order to create a model of rational choice that can model human behaviour accurately in the case of prisoner's dilemma / public goods type situations.