Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
Civic Ethics as Solutions to Social Dilemmas
I.
Introduction: Rules and Civil Society
Many of the above philosophers argued in passing that a subset of
ethics and virtuous behavior helps to make civil society possible. Aristotle
suggests that reciprocity holds society together. Locke, Smith, and Mill
suggest that respect for one another’s life, liberty and property provide the
foundation for community life. To these men the practical importance of
internalized rules of conduct were obvious.
This was partly because all but Aristotle had read the work of another
philosopher, who is less known for his contributions to ethics and
economics than to political theory, Thomas Hobbes (1588-1679).
Whatsoever therefore is consequent to a time of War, where
every man is Enemy to every man; the same is consequent to
the time, wherein men live without other security, than
what their own strength, and their own invention shall
furnish them withal. In such condition, there is no place for
Industry; because the fruit thereof is uncertain; and
consequently no Culture of the Earth; no Navigation, nor
use of the commodities that may be imported by Sea; no
commodious Building; no Instruments of moving, and
removing such things as require much force; no
Knowledge of the face of the Earth; no account of Time;
no Arts; no Letters; no Society; and which is worst of all,
continual Fear, and danger of violent death; And the life of
man, solitary, poor, nasty, brutish, and short. [Hobbes,
Thomas (1651/2004-07-01). Leviathan. (pp. 70-71). Kindle
Edition.
The society described and its core dilemma are referred to as the
Hobbesian Jungle or Hobbesian Anarchy.1
Hobbes suggests that the natural order of mankind is disastrous. In the
absence of ethical or legal constraints, mankind would be reduced to bare
subsistence and life would be highly uncertain and short. Conflict over
scarce resources can easily escalate to the point that a civil society is
impossible. Resolving fundamental conflicts among human interests is a
prerequisite for civilization and markets.
Hobbes suggests that people would recognize the dilemma and agree to
create a strong central government--a Leviathan--via social contract to
escape from it. That government would enforce laws protecting life and
property, which would improve life for all, even if the government itself
could not be easily restrained. Hobbes was not the first to link prosperity
to law and order, but his statement of the problem and suggested solution
were clearly influential during and after the enlightenment.
The details of Hobbes’ social contract are not of interest for this book,
but his argument that human interest are not inherently harmonious is.
Subsequent enlightenment scholars did not all agreed with the Hobbes’
bleak assessment of the “natural state,” nor of the impossibility of
constraining government once created, but most regarded the problem
characterized by Hobbes to be serious and essential.2
Locke (1689), for example, also uses a natural state and social contract
to explain the emergence of legitimate government, but he paints a more
pleasant picture of the natural state. Nonetheless, as in Hobbes, agreements
to delegate law creating and enforcing powers to a single organization
account for the legitimate authority of government. Good government and
good law from these contractarian perspectives are those which all persons
The Leviathan was written in the relative security of Paris during the English civil war, a war that may have inspired this idea.
Contemporary archeologists have argued that the Hobbesian jungle was a reasonably correct depiction of society from the dawn of agriculture through
the iron age. Although the warfare was not literally man against every man, but tribe against every tribe, life was poor, nasty, and short. See, for example,
Keeley (1996). More recent evidence from the world’s remaining stone age tribes is more or less consistent with this characterization, although not all
anthropologists agree about the importance of warfare and warrior virtues in such societies.
1
2
Page 1
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
would agree to because they advance the shared interests of essentially
everyone living in the community of interest.
Table 6.1: The Hobbesian Dilemma
Locke (1689) suggests that civic ethics may play a role in solving the
Hobbesian dilemma , as noted in chapter 3. We’ll explore that possibility in
this chapter. This chapter demonstrates that internalized ethics and external
law enforcement are substitutes for one another. Elementary game theory is
employed in the analysis, because it provides sharper illustrations of both
the social dilemmas faced and their solutions than prose alone is able to.
Thomas
Although the ideas of ethics, conflict and social contracts were all
developed before game theory emerged in the mid-twentieth century, the
essential logic of many social dilemmas and their solutions can be most
clearly illustrated with elementary game theory. Two archetypal social
dilemmas are focused on in this chapter: the prisoner’s dilemma game and
the coordination game. A third choice setting is added to the list in the next
chapter, namely assurance games.4
A. The War of Every Man
The natural place to begin our analysis of the role of ethics in civil
society is with the Hobbesian dilemma. The essential features of this choice
setting can be characterized with the following 2x2 game matrix.
(J , T)
(J , T)
Don’t
(10,10)
(0, 14)
Attack
/Steal
(14,0)
(2,2)
John
The analysis shows that internalized codes of conduct can reduce a wide
variety of problems associated with life in communities. Indeed, Spencer
argued that the evolution of such norms could in principle solve all the
problems associated with life in a community, without the need for
government via social contact or otherwise.3
II. The Hobbesian Dilemma
Don’t
Attack
/Steal
Two persons are assumed to have equal liberties and be unconstrained
by internalized ethics or external legal sanctions. Each is assumed to
initially control their own labor and a stock of some useful item such as
food, water, or firewood. Labor (time) can be used to harvest more of that
good a natural resource freely available in the area, or it can be used to
attack the other and attempt to take their stocks of the item. Warfare,
however, destroys part of the stock being contested.
In this context, steal, attack, and defend are all equivalent strategies.
The alternative (don’t) is to employ one’s time and the resource harvested
for peaceful purposes. For example, both persons may initially control 8
units of the good in question and be able to harvest 2 more during the
period of interest. Alternatively, they may attempt to steal the horde of the
other, the process of which spoils or consumes between 2 and 6 units of
the good struggled over, depending on whether the theft is resisted or not.5
The emergence of civil ethics does not necessarily reduce warfare among communities, but does reduce it within communities. Montesquieu (1748)
argues that international law may serve a similar purpose among communities.
4
Elementary game theory was not available to philosophers writing before W.W.II.
5
Only the rank order of the payoffs matter for the illustration, however, not the details of the choice setting. Archeologists speculate that it was the
emergence of settled communities with significant stocks that set off the war of every tribe against every other.
3
Page 2
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
If the person attacked is not fully engaged in attack or defense, a less
effective defense is undertaken to repel the attack, which imposes a cost on
the attacker, but not sufficient to nullify the attack. Poorly defended
surprise attacks thus pay for the victor (14>10). On the other hand, if both
parties are fully engaged in military efforts their resources are largely
consumed or spoiled by those efforts, with the result that civil society is
poor or impossible for each, although still better than being the victim of a
surprise attack (2>0).
Given these payoffs, consider the thought process that each of these
individuals may go through. If “I” know that the other(s) will not attack,
then I can also not attack in which case I get a payoff of 10 or I could
attack in which case I get a payoff of 14. On the other hand, if the other(s)
attack, then by not attacking I get a payoff of 0, whereas if I attack (or
strenuously defend) I get a payoff of 2. I am better off in each case by
attacking (16>10 and 2>0). So, I should attack.
Since both persons go through the same reasoning, the result is that
everyone attacks and the result is the bottom right-hand cell that
characterizes the Hobbesian dilemma. This cell is a Nash equilibrium,
because no person can improve their result by acting alone--that is by
adopting another strategy. It is, of course, the relative magnitudes of the
payoffs rather than their absolute magnitudes that imply the Hobbesian
equilibrium. The numbers associated with the various possible outcomes
are simply illustrative.
Although this game is highly simplified, it should be clear that the basic
logic of choice setting extends to ones in which thousands of people are
involved and in which a variety of strategies between the all out attack-steal
strategy and the “mind my own business” strategy exist.6
The Hobbesian outcome is stable, although it is not the most attractive
one for either person. Given the other’s strategy, neither can do better by
changing their choice, although both people could be made better off if
they could agree not to attack one another (10>2).
It is this potential advantage that Hobbes suggests can be achieved
through an agreement to establish a law-enforcing regime. Without such an
enforcer, the neither attack (10,10) cell is not an equilibrium, because both
John and Thomas have an incentive to renege on their agreement and
pillage the other (16>10).7 Escape from the Hobbesian jungle is physically
possible, but how to do so is not obvious or trivial.
Hobbes’ solution was the creation of a strong central state that would
punish persons for violating the law. There are gains to trade that can be
realized by all participants by establishing an organization to protect the
original distribution of resources by punishing all persons that choose the
attack/steal strategy.
B. Civil Ethics and the Hobbesian Dilemma
An alternative solution is the emergence of internalized normative
systems. A wide variety of ethical systems can produce rules of conduct
that disapprove of the attack strategy. Such systems can also solve the
Hobbesian dilemma if sufficient numbers of persons in a community have
internalized the implied codes of conduct. Such persons feel better off
(virtuous) when they follow the rule of conduct and worse off (guilty)
whenever they violate the rule.
Consider for example a community of devout utilitarians. The payoffs
can be thought of as physical resources, money, or as happiness--that is to
say as utilities. If one interprets the payoffs as utility levels for the two
players, the numbers can be added together to create an exact measure of
For more detailed analyzes of anarchistic societies, see Tullock (xxxx) or Skaperdas (xxxx). A repeated version of the above game can be represented in
extended form or in normal form. The matrix above can be used to shed light on repeated games as well. In such cases, the payoffs represent the expected
present discounted values of the net benefits associated with the pure strategies of attack and don’t attack. Such games often have sub-game perfect Nash
equilibria of the sort depicted.
7
That this property is not immediately obvious to readers who are not familiar with game theory is evidence that the readers have internalized a “keep
your promises” ethic. Such norms allows agreements to move from (2,2) to (10,10) to be relied upon, for reasons indicated by table 6.2.
6
Page 3
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
aggregate utility. The highest aggregate utility possible in the game is that
associated with the (10,10) cell. This outcome yields a social utility of 20,
which is greater than the aggregate utilities associated with the other
outcomes (16 or 4). Utilitarians would thus recommend that steps be taken
to move to the (10,10) cell and escape the Hobbesian jungle.
If all the players were pure utilitarians, they would act as if they received
the total payoffs in the cells rather than their personal payoff and would
always choose the non-aggressive strategy (20>16 and 16>4). A society of
utilitarians would thus not confront the Hobbesian dilemma. For
utilitarians the magnitude of the payoffs are critical to their analysis, rather
than merely illustrative, but in this case the relative magnitudes are
sufficient for these conclusions.
Such a representation of the subjective effect of violating a norm or
rule implies that internalized codes of conduct are not totally binding, as
they would be if the losses from violating the norms were infinite. Instead,
the strength of the norms affect decisions at the margin.
The effect of internalized norms may be incorporated on the
Hobbesian dilemma as illustrated in table 6.2. In this particular case, the
dilemma is solved whenever the “ethical cost” associated with violating the
norm (G) is greater than “4” for each person. In that case, each person
chooses the “don’t” strategy and the result is the upper lefthand cell.
Table 6.2: Solving the Hobbesian Dilemma
with Ethics
Several other ethical systems can also potentially solve the problem,
without requiring such detailed information. Examples include
nonaggression ethics (defend, but never attack), “respect other person’s life
and property” ethics, or a “promise keeping” ethic sufficient to assure that
agreements are followed out.
The effects of such internalized rules of conduct can be modeled in
several ways. They may be represented (i) as taking some strategies off the
schedule of life’s possibilities, (ii) as reductions in the perceived payoffs
associated with the “attack” choices, (iii) as increases in the perceived
payoffs of the “don’t choices, (iv) combinations of the two, and so forth.
The second characterization is adopted below. This allows one to analyze
how the strength of internalized norms affects behavior.
Suppose that the internalization of an internalized rule of conduct
means that one feels somewhat “guilty” when one violates the rule. Guilt,
however, is not necessarily of the all-consuming variety. Finite feelings of
guilt reduce the subjective payoff to the forbidden or wicked strategy by
amount G. Such reduction in the subjective reward (utility) of a particular
action might be associated with a variety of ethical systems: with the
reduction in praise from an impartial spectator, the failure to behave
virtuously, or increased fear of being condemned to an unpleasant afterlife.
Thomas
Don’t
Attack
/Steal
(J , T)
(J , T)
Don’t
(10,10)
(0, 14-G)
Attack
/Steal
(14-G,0)
(2-G, 2-G)
John
Note that the guilt associated with “bad” behavior serves the same role
as penalties imposed by a strong central government. In either case, a code
of conduct changes private incentives and provides greater security for life
and property.
Of course, not all ethical systems have this property. Internalized
rewards for military bravery or stealth, for example, might perpetuate the
Hobbesian Dilemma, rather than curtail it. Such norms provide additional
reward for strategies that are already problematic. The payoffs of the
aggressive choices, 2+V and 14+V, will be clearly regarded to be better
than the payoffs associated with the passive choices of 0 and 10.
Page 4
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
With respect to norms that tend to perpetuate the Hobbesian dilemma,
it is clear that such societies, proud of their military virtues though they may
be, would live closer to the margin of survival than more prosperous
communities whose norms reduce counterproductive conflict rather than
encourage it, as might be said of the ancient Greek cities of Sparta and
Athens.
That relatively few people in most communities routinely engage in
criminal actions suggests that the combination of external and internal
supports for appropriate conduct are sufficient to encourage most people
to behave ethically. The strength of internalized norms and expected
formal penalties (G+P) are sufficient to deter theft, murder, by most
people and so allows a community to escape from the Hobbesian dilemma.
Only a subset of possible norms and virtues actually solve the
Hobbesian dilemma. Moreover, the greater the temptation, the greater the
feeling of guilt or virtue required to over come it.
It bears noting that internalized ethical systems have advantages over
external law enforcement. An internalized normative system is less likely to
be avoided, can be more fine-grained than a uniform legal system can be,
and does not require a costly organization to detect and punish
transgressions. Given an internalized ethical system, guilt is imposed by the
perpetrator him or her self, rather than by outside observers. Although one
may occasionally “fool oneself,” this is less likely than avoiding detection
by others. Thus, an internalized normative system is more likely to bind
than an externally enforced system is.
C. Survival in the Long Run and the Evolution of Ethical Theories
Most of the philosophers that we surveyed in Part I believed that ethics
and formal law often support the same conduct. Stealing may be both
“wrong” and punished by the legal system. Just conduct may consist largely
of following one’s community’s formal laws.
That civil ethics and criminal law often discourage the same behavior is
itself of interest and may be explained in several ways.
Theists would employ religious texts to argue both in favor of particular
codes of conduct and formal laws consistent with them. Both morally
depraved and criminal persons have violated God’s commandments,
whether written in scripture or based on natural law. Virtue advocates
would argue that laws can help encourage virtue by penalizing vice. Those
employing the Kantian imperative favor both behavior and formal
legislation that is compatible with universal law. Utilitarians would argue
that criminality occurs when persons undermine the aggregate happiness of
their communities.
An evolutionist like Spencer might argue that broad agreement about
the actions classified as “criminal” among ethical theories reflects
survivorship pressures for communities. Communities that failed to punish
or discourage such actions are less likely to be self sustaining. Many civil
and criminal laws are in this sense “natural,” although not necessarily of
divine origin.
Ethics, however, may not be sufficient to solve all the critical social
dilemmas faced by a community, and so communities may require formal
legal systems and penalties as well. As Thomas Paine once wrote:
Here then is the origin and rise of government; namely , a
mode rendered necessary by the inability of moral virtue
to govern the world; here too is the design and end of
government, viz. freedom and security. [Paine, Thomas
(1776/2014-07-31). Common Sense (Kindle Location 97-99).]
III. Beyond Law and Order: The Provision of Public Services
After a degree of law and order have been established, a civil society
may begin to emerge, as the war of everyman becomes less all consuming.
Law and order may be enhanced by encouraging civil ethics and through
better law and law enforcement. Other useful services, both relatively
important ones and minor ones may also be initially under-provided from
the perspective of most members of a community.
Page 5
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
Table 6.3 illustrates the nature of problems associated with producing
services that advance shared interests within a community, but which
cannot be easily “priced” and sold to individual consumers.
As in the Hobbesian dilemma, the essential free-riding problem
emerges from the relative size of the payoffs rather than their absolute
values. Each person would better off if all contributed to the provision of
the public service (3>0), but narrow self interest leads all to free ride.9
Suppose that there is a service that can be provided at a cost of 10 that
provides benefits of 8 to each person in the community of interest. If only
one person contributes to the public good, that person pays the full cost of
the good. If more than one person contributes to providing that good or
service, then the cost is shared one pro rata bases. If we normalize the
payoffs associated with the no public service state to zero (the presumed
status quo ante), the net benefits of voluntarily contributing to the
provision of the community service for a two person community can be
represented as in table 6.3.
It bears noting that free rider problems may be existential or trivial
depending upon the service at issue. For public services such as national
defense or potable water, the free rider outcome could be as threatening to
society as the original Hobbesian dilemma. In other cases, the result may
simply be a society in which life is a bit more difficult or less satisfying for
members of the community; which may, for example, have fewer than the
ideal number and variety of flowers planted along it’s sidewalks or in its
window baskets.
The incentives imply that free riding is the dominant strategy. If Alfred
believes that Paul will contribute, then he is better off free riding (8>3). If
Alfred believes that Paul will free ride, Alfred is still better off free riding
(0>-2). In either case, free riding is the best choice. The same logic applies
to Paul’s analysis and the result is the (0,0) outcome of mutual free riding.8
In either case, both Alfred and Paul would favor low cost steps to
move from the (0,0) cell to the (3,3) cell where the service is provided and
the cost shared among all members of the community. The same
government formed to solve the Hobbesian problem may be called on to
solve this problem by punishing free riders.
Table 6.3: The Public Goods Dilemma, Free Riding
Paul
Contribute
Free
Ride
(A , P)
(A , P)
Contribute
(3, 3)
(-2, 8)
Free Ride
(8,-2)
(0, 0)
Alfred
Of course, the problem may not be as severe as the game matrix
indicates because internalized norms have already emerged that reduce
public goods problems. Free riding may evoke strong feelings of guilt.
Contributing to the civil services of the community may evoke a strong
feeling of virtue. Either effect can be sufficient to solve free rider
problems.
The matrix above can be modified to illustrate the effects of
internalized codes of conduct and norms. These can be represented either
as an internal loss associated with violating one’s duty or as an internal
benefit from feelings of virtue or approbation associated with dutifully
“doing the right thing.”
Note that if this service could be priced, which is normally easier if exclusion is possible, a private organization or club might be formed to produce the
service. It would recover the cost of the service provided by charging for it (or them), in which case no taxes (coerced payments) are necessary.
9
As in the Hobbesian game, this game can be represented in an infinitely repeated form. In such cases, the relative payoffs should be considered present
discounted values of the pure strategy outcomes, and the equilibrium depicted as a sub-game perfect Nash equilibrium.
8
Page 6
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
The moral solution may also require resources, often at the family level,
but these tend to be routinely provided in societies with a viable civil order
and so as a “fixed cost” may be neglected for the purposes of this
illustration.
Table 6.4: Solving the Public Goods Dilemma through
Ethics or Social Norms for Virtue
Paul
Contribute
Free
Ride
(A , P)
(A , P)
Contribute
(3+V, 3+V)
(-2+V, 8)
Free Ride
(8,-2+V)
(0, 0)
The illustration also shows that the extent to which norms can be relied
upon to solve public goods problems varies with the cost of the service
problem of interest. If the strength of civic duty is modest, as with V=2,
public services costing up to 10 can be overcome via civic ethics, but not
ones costing more than 10.
Alfred
Table 6.4 illustrates how virtuous conduct may solve the problem
characterized in table 6.3. In the case illustrated, the benefits from the
public service are nearly sufficient to justify private provision, so the “virtue
supplement” does not have to be very great to solve the problem; V>2 is
sufficient to assure that the public service is provided, although not to
assure that the costs are shared, which requires V>5.
Contributions for especially critical public services such as national
defense may also be supported by both theories of virtue and pragmatic
interests in survival. Such public virtues may be reinforced by public esteem
and approbation, as with military parades, honors, and privileges. Such
norms may be supported by public policy, instead of financing the desired
service directly through coercive means, although other free rider problems
would also have to be overcome to do so.
From a utilitarian perspective, the civil ethics solution is arguably
superior to the tax solution, because the tax solution consumers resources.
It requires financing a tax collection regime and a subsidy regime for the
public service. Coercion itself is a cost of the tax system. All these costs
may must be charged against the net benefits of service provision.10
10
11
Sufficiently intense norms can induce Minutemen to grab their guns
and rush to the defense of their community at a moment’s notice, to grab a
bucket and put out a fire down the street, or to induce contemporary folks
to sort through and separate their trash into designated recycle bins.
However, as ethics diminish in force or as the cost of the public service of
interest increases, coercive means are more likely to be at least part of the
solution to community free rider problems.
The above analysis suggests that one would not, for example, expect
each member of the community to purchase and drive their own tanks or
missile launchers to the national borders to provide national defense or
their own fire trucks to a downtown department store in the central city to
help put out a three alarm fire. For such capital intensive services, reliance
on ethics and volunteers is likely to be replaced or supplemented with
tax-financed professional armies and fire departments.11
The logic of the above games can be easily extended to communities
with larger numbers in which the internalization of codes of conduct varies
among residents. For example, in the above case, as the number of persons
in the community increase, the net benefits of sharing costs tends to
increase, because the costs can be spread among more members of the
community. If the choice setting is otherwise similar, the payoffs of the
upper lefthand cell become (8-10/N), but the remainder of the payoffs and
See Martinez-Vazquez and Winer (2014) for a recent analysis of the welfare effects of coercion within utilitarian and contractarian frameworks.
It bears noting that in the past these services were often provided by volunteer organizations such as the “minute men” and volunteer fire departments.
Page 7
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
losses are unaffected. A uniformly distributed virtue payoff of V>2 is still
sufficient to solve the free-rider problem.
In larger communities the uniform ethics assumption is less likely to
hold, although this is not necessarily a problem for the community. It is, for
example, sufficient for at least one person to have a virtue payoff greater
than 2 in the public goods illustration for the public good or service to be
provided out of sense of duty or for other ethical reasons.
As the cost of the service increases however, it may not be feasible for a
single person to provide the service, so more persons are required
contribute. The illustration suggests that the shift from single person
provision to shared provision will require stronger norms. However, large
communities would tend to exhibit a wider distribution of values for “S”
and “V,” and so more free rider problems would tend to be solved by
private contributions and public service clubs of various kinds.
With respect to non-utilitarian ethics, it is interesting to note that
persons and groups that contribute far more than the typical person are
often regarded as heroes or public benefactors. And it is partly from the
praise associated with such remarkable conduct that ideas about virtue
originate according to Aristotle and Smith.
financing of community services. The higher the benefits of virtue or
stronger the sense of civic duty, the lower other subsidies or (expected)
formal penalties need to be.
Insofar as the cost of tax systems tend to rise with the revenues raised,
including increases in DWL, a community with stronger civic virtues tends
to have a less burdensome government than one without ethical support,
other things being equal.12
IV. On the Value of Conventions: Coordination Games
Beyond peace, essential public services and public amenities, there are a
variety of other problems that can be overcome to make life in a
community easier, more pleasant, and productive. Among the simplest of
these are more or less arbitrary conventions that solve coordination
problems. Examples include, language, measures and weights, calendars,
and many simple customs of day-to-day life such as greetings.
Coordination games have the property that all participants are better
off when everyone chooses the same strategy, but no particular choice is
better than another. In such cases, what Kant referred to as universal law is
not unique. This makes coordination games substantially different from the
prisoners dilemma games analyzed above. Only one of the strategies of a
PD-game necessarily yields better results when everyone adopts it. Another
important difference is that the desirable outcomes of coordination games
are stable once they emerge. Solutions to coordination games are self
sustaining once they emerge, in contrast to those of PD games.
The opposite kinds of persons, who might be referred to as unethical
free riders or malefactors, fail to contribute at all. Persons who require the
most stringent penalties to encourage payment of their cost shares or to
discourage violations of the law are often regarded as immoral, criminal, or
depraved. Their actions are normally spurned rather than praised by fellow
members of the community.
In cases in which ethics are not strong enough the solve free rider
problems, civil ethics nonetheless tend to reduce the overall cost of
governmental solutions. It is the sum of the expected formal legal penalties,
losses from disapprobation, and subjective losses associated with failures to
perform one’s civic duties that induce persons to provide or contribute to
Table 6.5 illustrates the payoff structure of a coordination game for the
case of passing people on the left or right on a road or sidewalk. Note that
in this case, a pattern of community behavior is likely to emerge that is
stable and requires neither ethical nor legal support. Self-interest is
sufficient to sustain it. However, there are two possible equilibria, so it will
not be clear to community members which will emerge.
It bears noting, for example, that relatively low expected penalties are sufficient to induce very high levels of tax compliance in the US and most
Western Europe. Although the penalties are not trivial, the probability of being punished is very low. That trust plays a role in taxpayments has been
demonstrated by Feld (xxxx).
12
Page 8
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
As conventions emerge, the unfavorable off diagonal results may occur
with some frequency, a frequency that tends to increase with the size of the
community.
Table 6.5: Coordination Games: Walking on
Community Paths and Sidewalks
Harold
Pass on Left
Pass on Right
(D , H)
(D , H)
Pass on Left
(1, 1)
(-1, -1)
Pass on Right
(-1,-1)
(1, 1)
A convention following speaker may be rewarded with applause and an
unconventional speaker booed as incoherent.13
As conventions become internalized norms, a driver that places his or
her car on the “wrong” side of the road or a speaker that makes a poor
word choice or confuses tenses, would be embarrassed by his or her
“error” (failure to follow the convention). As norms and habits of thought
and behavior emerge, a subset of the possible equilibria in new
circumstances becomes more likely than others to emerge as conventions.
Table 6.6: Internalization of a Convention for Walking
on Community Paths and Sidewalks
Duncan
Harold
The role of ethics and other types of norms in coordination games is to
shorten the period of disequilibria. They also do so by encouraging
generalizations of norms that resolved other coordination problems.
General convention simplify the community’s life, by increasing the
likelihood that a benevolent equilibria occurs or emerges rapidly. A
normative principle that extends easily to new circumstances--escalators
and slideways--has value. Although, one could imagine different rules for
sidewalks, roads, stairs, and hallways, a uniform one is easier to remember
and easier to generalize to new situations.
When a convention becomes internalized, one of the strategies
increases in value relative to the other. It becomes the “right thing to do”
whereas the other becomes the “wrong thing to do.” In a society with a
convention to drive on the right, driving on the left is punished. Similarly,
but less formally, conventions for language and grammar are taught at
home and correct grammar rewarded with praise, high grades, and esteem.
“Incorrect” grammar is punished with criticism, low grades and disesteem.
13
Pass on Left
Pass on Right
(D , H)
(D , H)
Pass on Left
(1+V, 1+V)
(-1+V, -1-G)
Pass on Right
(-1-G,-1+V)
(1-G, 1-G)
Duncan
Such norms are often taught to children and foreigners. The simplest
conventions may be taught in a moment, as children may be taught, for
example, to always pass on the right of someone approaching them from
the front (and on the left of any person they approach from the rear).
Others are so complex that it takes a lifetime (or more) of practice to
master, as arguably true of English grammar
Even a small nudge provided by an internalized normative principle is
often sufficient to induce the “right” strategy choice in unfamiliar
circumstances within a given community. The combination of “pull” and
“push” from the community’s conventions and direct losses from
deviation simply have to be sufficient to dominate the losses associated
See Brennan and Pettit (xxxx) for an analysis of markets for esteem.
Page 9
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
with deviating from other possible equilibrium (or rebellion), as with
(V+G)>2.
Although conventions are often supported by strong norms,
conventions are not always interpreted as matters of ethics. This is partly
because many conventions are entirely arbitrary. For example, utilitarian
theory provides no guidance about which of the above two conventions
should be followed, although it supports the existence of such conventions.
Many conventions also tend to be beyond the domain of ethics from
the perspective of Aristotle or Smith insofar as particular norms do not
advance virtue or happiness, or necessarily attract approbation from an
impartial spectator. Conventions often qualify as universal laws under
Kant’s approach, but many universal laws serve may equally well.
community norms may be supported by the virtue of prudence and a sense
of duty. Violations of internalized norms are routinely discouraged through
disapprobation.
It also bears noting that there are generally more social conventions
and uniformity norms than coordination problems that need to be
addressed. Examples from ancient and more recent history include
sumptuary laws, dress codes, dietary restrictions, and religious beliefs. In all
these cases, there turned out to be benefits from eliminating uniformity
norms that appear to have been larger than their long run costs.14
V. Partial Takings and the Internalization of Externalities
Morally neutral conventions include methods for writing of dates and
fractions, the voltage of electrical outlets, the width of railroad tracks, the
units and nature of money, the sizes of doors, male and female names for
children, the order of family and personal names, and many others.
Nonetheless, violating internalized norm-supported conventions, as with a
boy named Sue, tend to be associated with guilty reactions and
disapprobation from fellow community members, in much the same
manner as for actions regarded to be unethical.
Some conventions are also supported by legal sanctions, although this is
not required, given the logic of coordination games. Such laws can be
supported on pragmatic or utilitarian logic, if they help reduce the period of
disequilibria, or discourage irrational deviation (rebellion) from a
convention once an equilibria emerges. Thus, once established, following
The idea that law and order is the foundation of a civil society is not
entirely unambiguous. From the perspective of Hobbes, rights, allowed
spheres of personal conduct--are determined by the state. From the
perspective of Locke and Grotius, rights have divine origin and are
intuitively obvious. Spencer’s support for equal liberty, suggests that legal
rights should be similar for all persons, although that is not necessarily
implied by Hobbes or Locke.15 What rights are to be protected is not
entirely obvious, although a strong case can be made to protect individuals
form death and harm at the hands of their fellow citizens. A similar,
although more utilitarian case can be made for supporting rights of
ownership. Without such rights, incentives to work and save would be
absent, and less food, clothing, and shelter would be produced.
Ownership of self and physical property, does not necessarily imply
freedom from all sources of discomfort. One’s neighbors may create
annoying noises or smells on their properties that drift over to yours. In
The rise of tolerance can be regarded as a reduction in the number and extent to which unconventional behaviors that are formally or informally
discouraged by fellow residents of a community. Nonetheless, some conventions are productive for reasons developed in this subsection and for those
tolerance would be harmful to a community in the long run. Coordination problems exist. The principle that some risk of physical harm is required for
disapprobation, rather than psychological discomfort, implies that differences in religion, politics, etiquette, attire, and speech should be ignored, rather than
discouraged.
15
See for example Buchanan (xxxx) for an analysis of the emergence of civil law (property rights) from Hobbesian anarchy. He argues that esssentially
anything could in principle emerge, given assymetries in the initial powers or positions of individuals.
14
Page 10
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
economic terms, neighbors may impose externalities (spillover costs or
benefits) on others in the community.
That some actions impose costs or benefits on others is an important
source of civil ethics as is evident from the philosophical writing of Locke,
Smith, and Mill. Heroes receive approbation because they are public
benefactors, which is to say because their activities produce benefits for
most members of the community. Villains receive disapprobation because
they are public malefactors, which is to say their activities impose losses on
fellow members of the community. And, of course, it is not only these
extremes that exist, but a broad range of costs and benefits that one’s
liberty and property may impose on others in a community.
noise, smells, and pests. Many other nuisance producing activities would
serve as well as with, trash storage and disposal, activities that produce loud
noises, and ones that create risks for neighbors as with bonfires and
shooting practice. Urban farming, as it turns out, is an issue being debated
in many town councils in the US.
Three intensities are used below to illustrate cases in which the choice
at hand is not an all or nothing one.
Table 6.7: The Externality Problem
James
Ethical norms and rules of etiquette, as well as more fine-grained
descriptions of ownership rights, often emerge to discourage behaviors that
harm others and encourage behavior that confers benefits to others. Such
rules evidently attempt to “intenalize” what economists refer to as
externalities. Some of these rules explicitly encourage individuals to take
account of the spillover costs of their actions: ”do onto others as you
would have others do onto you.”
Others forbid some kinds of actions that clearly impose losses on
others. For example, tricking another into an action, as with fraudulent
claims about the properties of a good or services, is often regarded to be
unethical and punished by laws against fraud. Neighbors that share access
to a stream or lake may differ in their planned use of the water. One may
want to use it to carry away or dilute waste products, the other as a source
of drinking water. Similar, a neighbor’s smoky fire may be unpleasant or
unthealthy to his neighbors.
Craig
1 Chicken
10 Chickens
100 Chickens
1 Chicken
(C , J)
(4, 4)
(C, J)
(3,6)
(C , J)
(1, 8)
10 Chickens
(6,3)
(5,5)
(2, 6)
100 Chickens
(8,1)
(6,2)
(3, 3)
In this case, the Nash equilibrium involves sizable poultry farms in the
village. These provide benefits to their owners, but impose significant costs
on neighbors. From a Utilitarian perspective, the result is the worst one
listed in the table. Contractarians might also note that both James and
Craig would be better off if they agreed to reduce their flocks to size 10.
An externality occurs when one person’s behavior imposes costs or
benefits on others. In the former case, the problem may be said to be the
over provided of some activity by members of the community, rather than
under provision, as with free riding.
The example below considers problems associated with raising chickens
within a village. The assumption is that chicken farming is profitable for an
individual family farm, but imposes costs on their neighbors in the form of
Community norms might emerge to support this equilibrium, as a
norm that suggests that chicken farms as opposed to a few chickens should
be out of town, where neighbors are more dispursed and so less affected
by each other’s behavior than in the village. Alternatively, the villagers may
hold a town meeting to consider formal rules on chicken holdings in the
village. Self-interest alone, as per the contractarian logic, would imply that
the community would agree to impose such rules upon themselves.
Other normative perspectives tend to be mute on the issue of chicken
farming itself, insofar as chicken farming has little to do with the
Page 11
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
production of virtue or long term happiness, nor is it obvious that a unique
rule for chicken farms could serve as a universal law and so create a duty
consistent with Kant’s categorical imperative. The numbers are consistent
with the extremes being vices as in Aristotelian theory, although it is
doubtful that owning ten chickens would be regarded as a virtue.
ago, creating uniform ordinances will rarely produce perfect solutions,
whereas in internalized norms may be able to more appropriately take
account of relevant variations in circumstances. In this respect,
internalized norms may be regarded to be superior to the use of formal
legal machinery to address many externality problems.
A general internalized norm that made one feel guilty or suffer a virtue
loss whenever one’s actions imposes a “significant” losses on fellow
members of the community would improve life in this community if it is
appropriately strong. As demonstrated in table 6.8, a guilt level for large
flocks greater than 2 solves the problem, as would community fines of an
equal amount.
Where there is a continuum of possible choices and the best one is
(arguably) among the middle ones, one cannot simply punish the activity
per se with approbation, fines, or taxes. Such cases also illustrate the
challenge of Aristotelian, Smithian, and Kantian ethics. When multiple
intermediate cases exist and the best of them is not obvious, their ethical
theories provide little guidance.
“Perfect” amelioration from a utilitarian perspective will in many cases
require perfect internalization of these spillover benefits or costs. In such
cases, the moral code and its associated feelings of virtue and guilt serve as
a Pigovian tax and subsidy schedule for the activities in question. Such
internalization is more natural for utilitarians than for most other ethical
systems, but similar results would follow from the golden rule, the pursuit
of approbation and for a subset of problems, for dutifully following
appropriate Kantian universal law.16
Table 6.8: Solving the Externality Problem
James
Craig
1 Chicken
10 Chickens
100 Chickens
1 Chicken
(C , J)
(4, 4)
(C, J)
(3,6-G10)
(C , J)
(1, 8)
10 Chickens
(6-G10,3)
(5-G10, 5-G10) (2-G100, 6-G10)
100 Chickens
(8-G100,1)
(6-G100, 2--G10) (3-G100, 3-G100)
Nonetheless, any systematic norms or ethical principles that require
individuals to take account of the external costs and benefits of their
actions is likely to make life in a community more pleasant and sustainable
than it would be without them. The results need not be perfect to be
improvements over the natural state.
In a world of utilitarians, the obvious result would be the middle one,
although no general principle for the ethics of chicken farming would be
associated with it. Whether utilitarian or not, a local norm that spillover
costs be taken account of when making personal choices would help to
ameliorate a broad range of externality problems.
VI. Conclusions: Civic Norms Promote Life in Communities
Note that addressing the problem with a town ordinances and formal
penalties, in effect, redefines property rights. The choice of chicken
holdings beyond 10 is no longer part of the package of rights associated
with ownership of a particular parcel of land, nor part of equal liberty. The
same could be said of regulations that ban loud noises after certain hours or
insist that lawns be kept mowed and neat as a pin. As Aristotle argued long
16
The analysis of this chapter suggests that life in a community tends to
become more attractive as civic ethics become better adapted to the
problems associated with it. In the limit, as Spencer suggests, ethics could
in principle “do it all,” and eliminate all the conflicts and coordination
problems among persons living in the community. This is not to say that
the “ethical solution” is unique, but it is to say that a fine-grained
Solutions to some externality problems may require different persons to engage in different, rather than uniform, actions.
Page 12
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
community ethics could in principle solve the problems of living together in
a village, town, or country. In such cases, as Spencer suggested as a young
man, there would be no need for the centralized authority of a state.17
That states are almost universally thought to be necessary is evidence
that such norms have not yet emerged or been sufficiently internalized.
Changes in codes of conduct may emerge from progress in ethical
theories and from changes in the problems confronted by individuals and
communities. Civic ethics may be said to improve when they make a society
more attractive. They may do so by better resolving a given set of
problems, or by addressing previously neglected problems. Changes in
circumstances may also cause new rules of behavior to emerge and be
adopted by community members. The political, economic, or natural
climate may change the coordination and PD-like problems confronted,
and these may call forth new norms of behavior or new applications of
existing normative theories.
Perfection, however, is not necessary for communities to survive and
flourish. It is sufficient that life in communities be regarded as superior to
life outside of them by a subset of the people living at a particular time.
A. Civic Virtue and Urbanization
With respect to the main aim of this book, ethical solutions to the
problems of life in community are important because towns and cities are
where so much of economic life takes place. Life in cities requires
commerce insofar as less food is produced within cities boundaries than
required to sustain their populations. For this reason, cities have always
been centers of commerce and nearly always traded urban services for food
and raw materials from rural areas.
The extent of urbanization reflects the strength and sophistication of
civic ethics and associated civil law insofar as these solve problems of life
within a community. However, the extent of urbanization also reflects the
effectiveness and sophistication of local markets for food and raw
materials. When markets function poorly or transportation costs are large,
urban centers will tend to be small. In the next chapter, the effects of ethics
on the size and effectiveness of markets is explored. Given the results of
this chapter, we might anticipate that as ethical supports for markets
increase in strength and sophistication, we would expect markets to expand
allowing larger cities to be viable, for a given level of civil ethics and civil
law.
B. Changes in Norms and Urbanization in the Nineteenth Century
The nineteenth century was a century in which considerable
urbanization took place in the West. Not only did old cities expand, but
new ones emerged.
Some of this may be regarded as technological in nature, as innovations
in metallurgy and steam-based technologies reduced transportation costs,
increased the power of pumps for water and sewage, and better central
heating. These changes made life in cities more comfortable and imports
and exports to and from cities less expensive to undertake. The same
technological advances also created new economies of scale in production
and specialization that attracted large number of persons to communities
surrounding new manufacturing complexes.
The extent to which changes in norms also played a role in the
urbanization of the nineteenth century is evident in the policy debates that
took place within cities and national parliaments during that time.
Utilitarians, for example, lobbied for a broad cross section of policies that
tended to make life in cities more attractive and more feasible.
Policy debates in the nineteenth century took place with respect to
public education, sanitation, and economic regulation. As a rule of thumb,
right of center liberals believed that relatively few of the secondary and
tertiary problems should be addressed by urban or national governments
(as with Spencer) and left of center liberals thought many of them should
be (as with Mill).
Remarks in Spencer’s autobiography suggest that at some point after 1851, he changed his mind about the feasibility of a evolutionary cultural
equilibria, partly because he came to believe that society and human nature change more slowly than new problems emerge.
17
Page 13
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
Urban water and sewer systems were expanded and the use of trash
cans introduced. This was partly in response to lobbying by sanitation
lobbies who insisted that public health would be improved through
improved water supply, trash disposal, and sewage systems.18 A variety of
steps were taken to improve sanitation in the second half of the nineteenth
century, from improved trash removal, cess pools, and water supplies. A
variety of steps were also taken to increase the scope of markets as well, as
with improved highways, rail networks, and canal systems and reduced
tariffs.
That the rise of utilitarianism in the nineteenth century played a role in
all of these public policy debates is a subject that we’ll return to in part III
of the book.
The remainder of part one shift the focus from effects of ethics on the
civil society to that of ethics on the extent of the commercial society. It
demonstrate that the extent of commerce also tends to vary with the extent,
strength, and domain of ethics within and among communities.
VII. Appendix: Some Notes on Non-Cooperative Games
Theory
The use of game theory in economics began in the nineteenth century,
but accelerated after World War II. The Cournot duopoly model was
worked out in 1838, and provides an example of a non-cooperative game
with a Nash equilibrium. Analysis of other forms of duopoly as with
Stackelberg’s model and models of monopolistic competition developed in
the 1930s were also based on models and intuitions that would later be
referred to as game theory. However, it was not until shortly after W.W.II.
that game theory emerged as a separate field. The classic book that brought
the field to the attention of persons outside the small group of applied
mathematicians initially working on it is the Theory of Games and Economic
Behavior by von Neumann and Morgenstern (1944). A second more
accessible classic work is Games and Decisions by Luce and Raiffa (1957).
The application of game theory to economic problems continues to be
the most active area of theory in contemporary economics and philosophy.
A quick look at any economics journal published and many philosophy
journals in the past three decades will reveal a large number of articles that
rely upon elementary game theory to analyze economic behavior of
theoretical and policy interest.
Modern work on: the self-enforcing properties of contracts, credible
commitments, the private production of public goods, externalities, time
inconsistency problems, models of negotiation, and models of political and
social activity have used game theoretic models as their engines of analysis.
Game theory can be used to model a wide variety of human behavior
in small number and large number economic, political, and social settings.
The choice settings in which economists most frequently apply game
See Melosi (2005) for a history of the sanitation movement in England and the United States in the late nineteenth century, and of changes in urban
sanitation policies. Of course, sanitation was not invented in the nineteenth century. The old testament of the Bible includes several passages with
recommendations for cleanliness. Aristotle’s Politics also includes recommendations for water supplies and house placement. Sophisticated sewage systems
can be found in ancient Greek and Roman cities. However, concern for public sanitation in the West became relatively more important in the new urban
centers of the West in the nineteenth century, because they were expanding rapidly.
18
Page 14
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
{ (3) player A may choose S1 and player B may choose S2,
theory, however, are small number settings in which outcomes are jointly
determined by the decisions of independent decision makers.
{ or (4) player A may choose S2 and player B may choose S1.
In "non-cooperative game theory" individuals are normally assumed to
maximize their own utility without caring about the effects of their choices
on other persons in the game.
The particular combination of strategies is the result of the
independent decisions of the two players, A and B (Al and Bob).
{ The outcomes of the game, however, are usually jointly determined by the strategies
chosen by all players in the game.
{ Consequently, each person's welfare depends, in part, on the decisions of other
individuals "in the game."
For example, in Cournot duopoly, each firm's profits depend upon its
own output decision and that of the other firm in the market. In a setting
where pure public goods are consumed, one's own consumption of the
public good depends in part on one's own production level of the good,
and, in part, on that of all others. After a snow fall, the amount of snow on
neighborhood sidewalks depends partly on your own efforts at shoveling
and partly that of all others in the neighborhood. In an election, each
candidate's vote maximizing policy position depends in part on the
positions of the other candidate(s).
Game theory models are less interesting in cases where there are no
interdependencies among game players. For example, a social setting in
which there is little or no interdependence is that of a producer and
consumer in perfectly competitive market. A consumer (or firm) is able to
buy (or sell) as much as they wish without affecting market prices. Game
theory can still be used in such cases, but with little if any advantage over
conventional tools.
The simplest game that allows one to model social interdependence is a
two person game each of whom can independently choose between two
strategies, S1 and S2. There are four possible outcomes to the game:
{ (1) both players may choose S1,
{ (2) both may choose S2
A game is be said to have a Nash Equilibrium when a strategy
combination is "stable" in the sense that no player can change his strategy
and increase his or her own payoff by doing so. There may be more than
one Nash equilibrium. For example, the coordination game has two
equilibria. Neither person can make themselves better off by changing
their strategy (alone) given that of the other player(s) in the game.
A state of the world or game outcome is said to be Pareto Optimal or
Pareto Efficient, if it is impossible to reach another state where at least one
person is better off and no one is worse off. Note that the (Trade, Trade),
equilibrium is Pareto optimal, but not of the other outcomes are.
The Prisoners' Dilemma game is probably the most widely used
game in social science. What characterizes a PD game is that the
"cooperate, cooperate" solution is preferred by each player to the "defect,
defect" equilibrium. And also that the value generated by defecting is a bit
higher than the cooperative solution regardless of whether the other player
cooperates or not. Often the payoffs are represented ordinally with (3, 3)
for the mutual cooperative solution and (2, 2) for the mutual defection
result. The other payoffs are then (1,4) and (4,1) with the defector receiving
4 and the cooperator 1.
The PD payoffs can be represented algebraically with (abstract)
payoffs. (C, C) and (D, D) are the payoffs of the mutual cooperation and
mutual defection outcomes. And (S, T) and (T, S) for the "temptation"
and "sucker's" payoffs when one person defects and the other is "played
for a sucker. In a PD game, T>C>D>S.
The PD game's principal limitations as a model of social dilemmas are
its assumptions about the number of players (2), the number of strategies
(2), the period of play (1 round). However, these assumptions can be
Page 15
Moral Foundations of Capitalism: Chapter 6
Ethics and the Civil Society, An Elementary Game Theory Approach
dropped without changing the basic conclusion of the analysis. Essentially
the same conclusions follow for N-person games in which the players have
an infinite number of strategies (along a continuum) and play for any finite
number of rounds, as we will see later in the course.
Note that the mathematical requirements for completely
specifying a game are met in the Prisoner's Dilemma game. The possible
strategies are completely enumerated. The payoffs for each player are
completely described for all possible combinations of strategies. The
information set is (implicitly) characterized. (A player is said to have
perfect information if he knows all details of the game. A perfectly
informed player knows the payoffs for each party, the range of strategies
possible, and whether the other players are fully informed or not.)
In economics two person models of exchange, the Edgeworth Box, is
often used to illustrate the principle of exchange between two persons,
although we know that exchange in the real world is much more complex.
Game matrices can also be employed, as in the next chapter. 2 person - 2
strategy are often used in a similar way because such games can capture the
essential features of many choice settings of interest to social scientists.
However, as the above game theoretic representation of the "problem
of exchange" demonstrates, the usual economic representation of exchange
misses some details that may be important. On the other hand, the
Edgeworth box very nicely illustrates why the trade, trade equilibrium tends
to be Pareto optimal, which we used to determine the relative sizes of the
game's payoffs.
Page 16