Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Journal of Economic Methodology 5:2, 263-297 1998
On the problematic link between fundamental
ethics and economic policy recommendations
Olof Johansson-Stenman
Abstract This paper provides a systematic survey of major simplifying
assumptions that economists make, and often have to make, in order to obtain a
useful theory for policy recommendations in practice. The aim is to consider the
whole chain of assumptions with an emphasis on such simplifications that
economists sometimes tend to ignore (at worst), or at best often tend not to take
very seriously. The paper concludes that the link from fundamental ethics to
economic policy recommendations is often very fragile, but that this is neither a
convincing argument for economists to ignore normative issues, nor a reason to
pretend that policy recommendations can be derived without value judgements. In
order to improve the link it is recommended that welfare economists sometimes
move beyond reductionist individualism and learn more from other social sciences
regarding preference formation, decision behaviour, and the creation of values,
institutions and social norms.
Keywords: ethics, moral philosophy, normative economics, welfare economics,
policy recommendations, simplifying assumptions
1 INTRODUCTION: POSITIVE VERSUS NORMATIVE
ECONOMICS
Policy recommendations deal with the normative issue of how one (e.g. the
government) should act, and not just with how people and other agents actually
do act, or why they act as they do. Thls difference is fundamental, although it
is sometimes argued, rightly, that there is no objective value-free scientific theory
either, since analysis has to be undertaken by a scientist who can never completely be independent of his or her experience, the cultural and moral climate in
which they live, and so on. As expressed by Myrdal(1969: 74):
no social science or particular branch of social research can ever be
'neutral' or simply 'factual', indeed not 'objective' in the traditional
meaning of these terms. Research is always and by logical necessity
based on moral and political valuations, and the researcher should be
obliged to account for them explicitly.
Furthermore, as noted by Solow (1994: 243): 'Where powerful interests are at
O Routledge 1998
ISSN 1350-178X
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
264 Articles
stake, some research will be consciously or unconsciously perverted, and the
critical mechanism will be diverted or dulled.' Still, there is a fundamental
difference between positive and normative analyses, since the expressed aim
of positive analysis is to be objective and value-neutral, even if this aim can
never be fulfilled. In the words of Friedman (1994: 181):
Positive economics is in principle independent of any particular ethical
position or normative judgments . . . In short, positive economics is, or
can be, an 'objective' science, in precisely the same sense as any of the
physical sciences. Of course, the fact that economics deals with the
interrelations of human beings, and that the investigator is himself part of
the subject matter being investigated in a more intimate sense than in the
physical sciences, raises special difficulties in achieving objectivity at the
same time that it provides the social sciences with a class of data not
available to the physical scientist. But neither the one nor the other is, in
my view, a fundamental distinction between the two groups of sciences.
(emphasis added)
For example, most people seem to think that economists (and other social
scientists) should try to evaluate a proposed tax-reform as objectively as
possible. What will the consequences be for the environment, employment,
equity, national income, terms of trade, etc.? We know that this ideal
objectivity is often far from fulfilled, and that it is never completely fulfilled,
but most people probably agree that it is a reasonable goal to strive for. Still, it
may sometimes be advisable for social scientists to follow Myrdal's advice
and explicitly account for their moral and political values, since economists'
values and 'ideologies' may strongly influence analysis which is normally
characterized as positive and descriptive.
On the contrary, normative analysis can by definition never be value-free.
The normative question: 'how should the government reform its tax system?'
can never be answered without a set of value judgements. Trying to answer
this question solely on the basis of descriptions of how the society works, and
why people and other economic agents act as they do, would be an attempt to
derive an 'ought' from an 'is'. As clearly expressed by Nath (1969: 2):
No prescriptions can ever be made without starting explicitly or
implicitly with some notion of what is desirable, good, or, simply, the
social objective. Again, no objective can be formulated, or questions of
desirability settled, without implicitly or explicitly assuming some value
judgments.
This is of course not to say that normative economics should be less objective
than positive economics in the sense of replicability. Given the ethical
premises, the analysis should ideally end up with the same conclusions
irrespective of the values held by the researcher, in exactly the same way as
these values should ideally not affect the outcome of positive analysis
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
265
(Bromley 1992). The choice of ethical premises has often been considered to
be the choice of policy makers, and hence beyond the scope of economists.
This view, although still common, is, however, rarely possible to maintain in
practice; see e.g. Hausman and McPherson (1993, 1996). For example, it is
not very common that policy makers express their inequality aversion in
quantitative terms. In any case, normative analysis which leads to policy
conclusions regarding what the government should or should not do involves
ethical values, irrespective of whether these values are held by the researcher,
the policy maker, or someone else.
The argument that one cannot derive an 'ought' from an 'is' was made
famous by David Hume in the eighteenth century, but it is still a controversial
statement. For example, Dasgupta (1993: 7) writes: 'The "descriptive" and
"evaluative" components of concepts like destitution cannot be separated.
They are entangled' (cf. Putnam 1993). Another even more problematic
example is the proposition 'it is a sin to kill'. This, in a sense factual
proposition, seems clearly to imply the normative proposition 'one should not
kill'. The difficulty clearly arises for semantic reasons since the word 'sin' is
defined with the use of values, e.g. as something we should not do. A similar
example which may be more important in economics, but is perhaps less
obvious, is the statement: 'the social costs of reducing greenhouse gas
emissions further exceed the benefits'. This may also seem to be a factual
proposition which may be either correct or incorrect. However, even
disregarding the tremendous uncertainties and economic interests involved in
this issue, it is still a value-laden proposition in the sense that the costs are
derived from an implicit or explicit social objective, which (by definition) is
ethically based. Thus, the separation between intrinsically value-laden and
purely empirical issues is not always straightforward; see also Blaug (1992:
chapter 5).
Another reason why the separation between factual and normative
propositions is often criticized is that many social scientists argue that
economics as an academic discipline will never be value-free, or a solely
positive science. But, as argued by Weston (1994), even if this certainly is true,
it is not a valid argument for not retaining the positivelnormative distinction in
economics, since there are no valid reasons to consider normative issues to be
less relevant for the academic discipline economics. Even though the
language is never completely value-neutral, and we have to be aware of the
fact that seemingly descriptive matters may sometimes be derived from value
judgements, it seems in many cases to be both possible and useful to
characterize an issue as either normative or positive, and either intrinsically
value-laden or descriptive, in the senses discussed above.
This paper will focus on the normative and intrinsically value-laden part of
economics and discuss the link from the most fundamental ethical premises
step by step all the way towards policy recommendations. However, since
many assumptions are needed on this way, there are of course many different
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
266 Articles
ways to end up with policy conclusions. This paper will focus on what may
be seen as the main track in normative economics, including
consequentialism, anthropocentrism, welfarism, utilitarianism and consumer
sovereignty, and not follow up all other possible tracks. For example, it will
not discuss which additional assumptions one has to make if one rejects
welfarism and revealed preference economics. For more general recent
surveys of ethics and economics, see Hausman and McPherson (1993,1996),
Hamlin (1996: introduction), Sen (1987) and Vickers (1997). For a thorough
collection of papers on this subject, see Hamlin (1996).
In Section 2, we will discuss the ethical foundation of normative economics
further, and discuss several common simplifying assumptions on the way
towards a more 'operational' theory, in the sense that it is useful for policy
purposes.' Section 3 links the theory of consumer behaviour to these ethical
foundations, and Section 4 briefly discusses the historical controversies on
how to deal with normative issues in welfare economics. Section 5 draws
conclusions.
2 THE ETHICAL FOUNDATION OF NORMATIVE
ECONOMICS
The betterness relation and consequentialism
Questions related to what should be decided or done are of a moral nature,
and the analysis must then be based on some ethical grounds. A fundamental
ethical criterion is the betterness relation (Broome 1991a) that an action
should be undertaken, and a decision taken, if they are better than every other
action or decision, in the sense that they are more 'good' than any alternative.
This is also often denoted a teleological or consequentialist theory as opposed
.~
this goodness-related ethical guideline
to a deontological t h e ~ r yAlthough
for actions is indeed very general, and might seem obvious, it is not
completely uncontroversial. For example, one may argue that some actions
are 'right' or 'wrong' independently of how 'good' they are, based on rightsbased ethics (e.g. Kantian duty-based moral). Dasgupta (1993: 27-8) remarks:
'The key to deontological reasoning lies in a recognition of the priority of the
right over the good. The hallmark of consequentialism is just the reverse: it
acknowledges the priority of the good over the right.' Economic theory
normally assumes consequentialism, i.e. that what makes an action good or
bad depends solely on how good or bad the consequences of the action are.
However, it should be noted that consequentialism as such does not rule out
that actions may have intrinsic values, since the action is a consequence in
itself of the action; see e.g. Broome (1991a) and Scheffler (1982).3 Still, in
economics one typically ignores intrinsic values in addition to
consequentialism. Although there are no a priori arguments why this would
follow from the betterness relation, many philosophers and economists seem
Ethics and economic policy recommendations 267
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
-
-
to agree (implicitly or explicitly) on the usefulness of this assumption.
A popular counter-argument, however, is based on whether it can be
considered good to lie, or not. A consequentialist ethical theory which ignores
intrinsic values implies that it would be justifiable to lie if the overall
consequences (except from the action itself) were better than if one had told
the truth. But with a partially deontological view there might be an intrinsic
value in telling the truth. Hence, even if the consequences of lying would be
slightly better than by telling the plain truth, the latter might be preferable in
such a view. This argument obviously has some appeal, and most people
probably do value truthfulness intrinsically, at least to some extent. However,
most of this appeal may be superficial, due to the fact that there exist long-run
dynamic effects from honesty and other social norms in a society. For
example, if people can trust each other, obviously considerable control effort
and costs are saved. Hence, there is no contradiction between consequentialism (in the stronger form which ignores intrinsic values) and
arguments in favour of telling the truth also when the direct foreseeable effects
are better than if one lies, due to possible long-run effects which may be
difficult to identify. Hence, what may seemingly be intrinsic values may
instead be valuations of indirect, dynamic and complicated consequences, see
e.g. Ng (1981a, 1983). Still, to rule out non-consequentialist ethical theories,
and that actions may have intrinsic values, are serious restrictions.
Anthropocentrism
A consequentialist ethical theory is not very 'operational' in itself, since we
have no guidelines or criteria yet on what makes an action better than other
actions. As we will see, there often seems to be a trade-off between the degree
of 'operationality' and the 'quality' of the goodness measurement obtained. In
order to make an ethical theory more operational or, loosely speaking, more
practical and useful, some further simplifying assumptions are needed. For the
theory to be able to say whether an action is better than another, we need to
know which consequences should matter, and what makes a consequence
good.
It is often (almost always) implicitly assumed in economics that it is solely
the consequences for human beings that matter intrinsically. The
consequences for animals and plants may also matter, but only indirectly
through valuation by humans. For example, in environmental economics,
'existence values' of animals and plants are generally assumed to reflect
individual human well-being measured as a willingness to pay for the
environmental good (Freeman 1993; Mitchell and Carson 1989), but it is very
rare that one tries to value nature intrinsically (e.g. based on some naturalist
ethics), i.e. irrespective of human welfare. Many moral philosophers, such as
Singer (1975, 1979, 1980), criticize anthropocentrism, in the sense that the
goodness measure of an action would solely reflect the consequences (in terms
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
268 Articles
of welfare) for human beings, and argue that animals' welfare should in
principle have the same weight as humans' elf are.^ Ng (1983) agrees in
principle with Singer and others, but also notes that the application of such a
theory 'may require man to sacrifice an enormous amount of his welfare'
(1983: 165). He continues: 'Most people (myself included) are not prepared to
sustain such sacrifices. What justification can we provide to reject mananimal parity? I can think of a pure and simple one - self-interest. To those
who object that this is hardly an acceptable justification, I have to agree'
(1983: 165). Furthermore, as shown by Spash and Hanley (1995), there is
evidence that many individuals (including so-called 'deep ecologists', cf.
O'Neill (1997)) think that nature should be valued intrinsically (at least to
some extent). But is it really possible to talk about values in any meaningful
way without considering human beings? Alternatively, should we value the
welfare of other species also after the extinction of the human race? Norton
(1987) proposes an intermediate way, entitled 'weak anthropocentrism',
between anthropocentrism and 'ecocentrism'. An object can then have a value
if and only if there is a valuer. But given that there is a valuer, the object may
have an intrinsic value; cf. Aldred (1994).
If we still adopt an anthropocentric perspective (in the stronger sense) a
natural question follows: which human beings? The ethically most-justifiable
answer would perhaps be: all individuals, born and unborn. But, obviously, the
task of valuing the consequences for people living 10,000 years from now
would most often be rather problematic. Therefore, one often limits analysis to
the currently living population. The consequences for future generations may
be reflected to some extent, but only through the valuation of the current
generation (as for animals and plants). For comparison of some actions, such
as the choice between different unemployment insurance systems, this is not
very controversial. But the issue becomes more complicated when dealing
with actions which may cause long-term environmental effects, such as
emissions of greenhouse gases; see e.g. Spash (1993). In such cases it
becomes obvious that one has to undertake some lund of explicit intertemporal
analysis, and include also the welfare of future generations. However, it is still
common to count the welfare of future generations less than the welfare of the
current generation, primarily through the choice of discount rate. This choice
appears to be a typical example where (some) economists try to be valueneutral by simply choosing the actual observable interest rate (or some
average of existing interest rates which are considered relevant) and applying
this in a cost-benefit analysis; see e.g. Nordhaus (1991, 1994). But this is of
course an illusion since the discount rate is a calculation instrument which
should be derived from the social objective. Consequently, the common
practice of discounting utilities (and not just consumption) through the socalled 'pure rate of time preference', without explicitly stating that the welfare
of future generations should count less than that of the current generation, has
been criticized for a long time by many authors including Ramsey (1928) and
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
269
Harrod (1948); see Broome (1992) and Azar (1995) for recent treatments of
this issue in relation to the greenhouse effect. Hence, one cannot observe from
actual behaviour which interest rate one should apply in cost-benefit analysis
in general, and in particular not in cases which have consequences for
hundreds or even thousands of years. There is also an ethical problem
associated with the simple fact that it is this generation which decides how the
welfare of different future generations should count; weighing these nonuniformly may be seen as a form of 'deep' anthropocentrism. In a similar way,
but in another dimension, one often ignores (or puts less weight on)
consequences for people in other countries, or in other areas.
Furthermore, some decisions may not only affect a large number of
individuals, but they may also affect the number of people (living in different
time periods or in different areas). Blackorby and Donaldson (1984, 1991),
Broome (1996a, b), Hammond (1988) and Ng (1983, 1986) provide some
possible criteria for evaluating such population changes, but due to the
obvious ethical difficulties and complexity of this issue, it is neglected in most
applied analysis.
Welfarism
A further restriction is that the overall goodness of different actions is often
assumed to be reflected by an individualistic social welfare measure, which
we call a social welfare function (SWF). In western societies of today such an
assumption may seem natural and straightforward, but in a large part of the
world this is not at all uncontroversial. As recently expressed by Chow (1997:
3 24) :
The society is more than a collection of individuals. Hence the welfare of
the society is more than the sum of the welfares of its individual
members. People in many developing countries are striving for
nationalism and may consider the common good and national unity more
important than individual rights.
Furthermore, as discussed by Thurow (1996), individualism of the western
type is a very recent phenomenon in human history.
An SWF which depends solely on individual utilities is what Sen (1979)
denotes a welfaristic social welfare function. Utility is here taken to be a
measure in broad terms of individual subjective welfare or well-being; see e.g.
Haslett (1990). The social welfare W can then, in a society with n individuals,
be written:
where ui is the utility of individual i, and where social welfare W is normally
assumed to be an increasing function in all its arguments u. Hence, at this stage
we do not rule out that an individual k's utility depends on individual j's
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
270 Articles
consumption. Sen has frequently argued for a broader 'information base'
for the social objective or SWF, and his focus on capabilities and
functionings (Sen 1985, 1992, 1993) can be seen as such attempts. Ng
(1981a) provides a powerful defence of welfarism and argues that most (if
not all) alternative arguments (such as a fundamental human right of not
being tortured) in the SWF are unnecessary since these arguments are
basically motivated by welfarist reasons. He asks rhetorically: 'Why do
we hear about "human rights", "animal rights", etc., but no "stone rights"?
An obvious answer is that stones do not feel pleasure and pain' (Ng 1981a:
530). Sen (1981) in a reply agrees that welfarist arguments may be
important for the creation of human rights and other social rules, but he
still argues that these welfarist arguments are far from sufficient: 'Cases of
exploitation, sexual discrimination, racial asymmetry, etc., are standardly
criticised on grounds that do not depend exclusively (or perhaps at all) on
utility considerations. Traditional welfare economics is not robust enough to
accommodate this type of values' (Sen 1981: 532). However, given a
sufficiently broad definition of utility, it appears possible to capture some, if
not all, of these features in a welfaristic framework, cf. Hammond (1991). One
could simply construct a function which depends on happiness as well as
various other elements which one would think contribute intrinsically (and not
only instrumentally) to a person's well-being, and then call this measure
'utility'. Broome (1991b, c) argues against such broad interpretation of
utility to mean something rather close to good: 'it is perfectly redundant.
We already have an excellent word with the meaning of good:
"Good"'(Broome 1991b: 11). Instead, Broome argues that the notion of
utility should be reserved for a representation of preferences. However, Sen
(1991) in a comment argues against such an abandoning since the term
'utility' is used, and has for a long time been used, with many different
meanings and, even if it would be better if this were not so, the important
point is to be as clear as possible about which meaning one is currently
attaching to the term. In particular, it is important not to use the term with
different meanings simultaneously.
Since it is common practice to focus on utilities as the only arguments in the
SWF, it appears reasonable to include as much as possible which corresponds
to a person's well-being in the utility function for the social welfare function to
be a measure which corresponds as closely as possible to an ideal goodness
measure. Still, as frequently pointed out by Sen, there may be other elements
of ethical relevance which are not covered by the concept of well-being (as it is
However, since it is difficult enough to consider
most often inter~reted).~
individual subjective well-being, one typically ignores other possible
elements of what constitutes a person's interest.
If we still accept an individualistic welfaristic SWF, we can deduce the
Paretian concept that an increase in the utility of one individual, without
decreasing the utility of any other individual, unambiguously increases social
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations 27 1
welfare. If a welfaristic SWF is our measure of goodness, the action should
correspondingly be undertaken. Note that the same cannot yet be said about an
increase in consumption for one individual, even if utility is increasing in the
individuals' own consumption, since utility may decrease as a function of
others' consumption.
The Bergson-Samuelson social welfare function
The SWF is often assumed to fulfil the more restrictive properties of a
Bergson-Samuelson social welfare function (B-S SWF).This means that
social welfare is a function of individual utilities, which in turn depend
(positively) solely on their own consumption of different goods.6In addition,
the utilities can depend on many other variables which, however, are assumed
to be independent of consumption. The SWF can then be written:
where xi is i's consumption vector of different goods. This may perhaps look
like a rather general and uncontroversial way of writing what the government
should try to maximize. However, the weak separability property of (2) is
crucial both for positive and for normative analysis. The separability implies
that individual i's utility from consuming different goods is independent of
individual j's consumption. However, we can still not say that the individual
choice between different goods would not be affected, because we have said
nothing yet about the connection between individual utility and choice of
consumption. The operationality of the theory increases drastically with this
assumption. For example, we are now able to propose that Pareto
improvements in terms of consumption should be undertaken, provided that
income is increasing in consumption. Thus, if the consumption of a specific
good increases for one individual and is unchanged for all other individuals,
social welfare increases. Unfortunately,the degree of correspondence with the
original goodness measure has correspondingly decreased. Indeed, this may
be one of the weakest links on our way from the original betterness relation to
actual policy conclusions.
Externalities, such as air pollution and noise, constitute obvious examples
But even if we disregard conventional
where the theory is too re~trictive.~
externalities, the theory has been restricted to a large extent. For example, it
precludes the influence of relative consumption on individual well-being. In a
widely quoted paper, Easterlin (1974) surveyed several studies which tried to
measure how happy people were in different countries, in different time
periods, and in different circumstance^.^ The result for a specific country in a
single year supports the view of conventional economic welfare theory: richer
people tend, on average, to be happier than poorer people. However, if one
compares the result in one country for different years, people seem, on
average, to be about equally happy over time, even though income has
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
272 Articles
increased dramatically during the same period. The same holds when
comparing the degree of happiness in different countries, i.e. people seem
to be about equally happy in different countries, although the average
income in those countries may differ dramatically. Abramovitz (1979) has
called this result the 'Easterlin paradox'. The most obvious explanation is
that people's well-being, or happiness, depends to a large extent on
their relative income, i.e. their own income compared to others' income,
as proposed by Duesenberry (1949). This is discussed extensively by
Abramovitz (1979), Frank (1985a, b), Layard (1980) and Scitovsky
(1992), among other^.^ In a recent paper Easterlin (1995: 35) concludes:
'Will raising the incomes of all increase the happiness of all? The answer
to this question can now be given with somewhat greater assurance than
twenty years ago . . . It is "no".'
A weak point in these empirical happiness studies is the link from stated
happiness in various kinds of surveys and actual happiness. As observed by
Brekke (1997) and Osmani (1993), it may be that people answer these types of
questions in relation to what they consider to be an average norm of happiness,
and that this norm may depend on income too. If so, happiness may depend
primarily on the absolute level of income even though this is not reflected by
the answers in surveys.
Still, psychologists have often been aware of these problems and tried to
compare with alternative measures such as asking friends of the persons how
happy they seem to be. Furthermore, as Frank (1985a) has observed, the
hypothesis that relative income matters appears to be more compatible with
evolutionary biology than the more conventional view that utility depends
solely on the absolute level of income. This is because in an evolutionary
perspective, the probability (and extent) of reproduction has historically
generally been positively related to the control of resources. For this reason, as
evolutionary biologists such as Alexander (1987) explain, our utility is often
positively related to income or consumption in a given context. Alternatively
speaking, there is evolutionary value in controlling more resources than do
others in the neighbourhood. According to Frank (1985a) it seems most
reasonable to assume that individual subjective well-being depends positively
both on absolute income (particularly at low income levels) and relative
income. See also the recent policy forum in the Economic Journal on
economics and happiness (Dixon 1997; Frank 1997; Ng 1997; Oswald 1997).
A related issue, introduced by Hirsch (1977) and discussed by Frank
(1985a, b), is the existence of so-called 'positional goods', in which the
relative component is more important than for other (non-positional) goods.
For example, the utility derived from a new car or jewellery may depend
strongly on others' consumption of similar goods, whereas the utility derived
from consumption of leisure may depend less on the amount of others'
consumption of leisure. This may have consequences, e.g. in the derivation of
optimal taxes as well as for the appropriate provision of public goods. For
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
273
example, Ng (1987) showed that pure public (and hence non-positional)
goods should in general be over-provided relative to the basic Samuelson
(1954) rule (XMRS = MRT) in the presence of such relative income effects.
Furthermore, a B-S SWF implicitly neglects all kinds of altruism,
benevolence or malevolence. Boulding (1969) remarks on this assumption:
The plain fact is that our lives are dominated by precisely this
interdependence of utility functions which the Paretian optimum denies.
Selfishness, or indifference to the welfare of others, is a knife edge
between benevolence on the one side and malevolence on the other. It is
something that is very rare. [. . .] We either rejoice when they rejoice, or
we rejoice when they mourn.
(1969: 6)
However, in the last decades there has been an increased interest in economics
on interdependent utility functions, and in particular on economics and
altr~ism.'~
Utilitarianism
Since the functional form of the SWF is not yet defined, we can still make
no general judgement of what to do in situations where utility increases for
some individuals but decreases for others. This is obviously a very serious
limitation, since almost all public decisions have such consequences. One
way to deal with this problem is to assume that social welfare can be
represented by a weighted sum of individual cardinal utilities, what is
sometimes called 'generalized utilitarianism', which is of course a further
restriction compared to the more general B-S SWF." This procedure is
proposed by Diamond (1967) among others, and the SWF can then be
written:
n
W
=Ia i u i ( x 3
(3)
i= 1
where the a ' s are welfare weights which can reflect the government's
distributional concern in terms of utilities. For example, social welfare
may be higher for an equal distribution of utilities compared to the same
amount of total utility more unequally distributed. In addition,
distributional concerns can also be reflected by the concavity of the utility
functions themselves, i.e. through the fact that a unit rise in income may
increase utility (or well-being) more for a poor individual than for a rich.
Rawls's (1973) theory of justice is often considered to be incompatible
with utilitarianism (even by himself), although it, interpreted in a very
narrow sense, is sometimes seen as a special case of (3), where the weights
are zero for all but the most unfortunate member of the society; see e.g.
Atkinson and Stiglitz (1980).12
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
274 Articles
Harsanyi (1975), Ng (1981b) and more recently Kaplow (1995), have
criticized this kind of weighted utilitarianism for being inconsistent with the
Pareto principle in the presence of uncertainty (as opposed to unweighted
utilitarianism). Their main argument, which follows Harsanyi (1955), can
easily be illustrated as follows. Consider a society consisting of two
individuals, Allan and Boris, who are assumed to maximize their own
expected utilities. The government can choose between two feasible
alternatives, I and 11. In state I Allan receives 10 utility units and Boris 5 utility
units with probability 0.5, and vice versa, i.e. Allan receives 5 utility units and
Boris 10 utility units with probability 0.5. In state I1 Allan and Boris receive 7
utility units each. Should the government choose I or II? According to the
Pareto principle the government should choose I because both prefer I over 11,
since both Allan's and Boris's expected utilities are larger in I compared to 11.
Still, if the SWF is according to (3), it would for sufficiently large inequality
aversion choose 11, and hence violate the Pareto principle.I3 Sen (1976, 1977a,
1986) has frequently objected that Harsanyi's theorems are not really about
utilitarianism, since the utilities are not defined independently of the
individual choice; see also the rejoinders by Harsanyi (1975, 1977), and
Weymark (1991) who largely seems to support the view of Sen.I4
The classical unweighted utilitarian SWF in the Benthamian tradition,
supported by Harsanyi and others, is given by:
=zu i
n
W
(xi)
(4)
i= 1
Both of these SWFs are frequently used in economic analysis. A common
critique of utilitarianism deals with the fact that some individuals may have a
much higher marginal utility of consumption than other individuals with the
same income (Friedman 1947; Nozick 1974).15Consequently, some people's
well-being may increase more from an additional unit of income than may
others'. Consider two individuals with the same income. A1 is a modest stayat-home kind of guy who likes to go fishing in his local pond more than
anything else, and whose utility does not vary much with his consumption. His
utility would not increase much with extra income. Bob is a party-guy who
loves spending money and hates not to. His utility would increase rapidly with
an increase in income. Everything else kept constant, would it then be a good
thing to redistribute income from A1 to Bob? Most people would probably not
think so, even though it would increase our measure of social welfare. Then
we clearly have a problem, since there seems to be a contradiction between
people's intuitive perception of what society should do, on the one hand, and
our definition of social welfare on the other.I6
However, in practice this seems not to be a major problem in applied
economics, because these kinds of individual differences are generally not
easily measurable at the necessary disaggregated level at which the economic
analysis is undertaken; cf. Lerner (1944) and Sen (1973: 81-5). Hence, in
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
275
applied economic studies one typically either assumes an identical utility
function for all individuals, or imposes some specific restrictions such as
equal marginal utility of consumption for equal income. Consequently,
differences in marginal utility of income for people with identical incomes do
not seem to play any role for actual policy recommendations. This illustrates
the fact that, sometimes, the quality of the goodness measure may actually
increase with additional restrictions."
3 CONSUMER BEHAVIOUR
Consumer sovereignty and revealed preferences
So far, nothing has been said about the relation between individuals' wellbeing and how they act. In economics, the dominating doctrine is simply that
individuals do what is best for themselves. Such an assumption simplifies the
analysis tremendously. We would then simply know that, for a given amount
of resources, individuals would maximize their own well-being. Alternatively
speaking, individuals' preferences are assumed to be revealed by their actions,
following Samuelson (1938). With this assumption, together with equation
(2), we are able to say that the individual's consumption choice is independent
of other individuals' consumption (provided that prices will not be affected by
their consumption). This assumption is crucial in most normative economic
analysis, such as optimal-taxation literature. For example, it implies that an
individual is always better off (or equally well off) from paying a certain
amount of money in a lump-sum manner compared to income or consumption
taxes.
Although this may often be a very well-motivated simplification, it is of
course not in any absolute sense correct. As expressed by Sen (1987: 11): 'The
coolly rational types may fill our textbooks, but the world is richer.' There are
several reasons for this, including obvious cognitive limitations; see Conlisk
(1996) for a recent survey on bounded rationality. Scitovsky (1974) provides
several examples based on 'money illusion', 'effort illusion' and 'time
illusion' in a paper entitled 'Are men rational or economists wrong?'.
Furthermore, psychologists have presented much evidence that expected
utility theory is often unsatisfying as a descriptive model under risk; see e.g.
Tversky and Kahneman (1974) and Kahneman and Tversky (1979), who show
that people instead often tend to apply some simplistic heuristic decision
criterion. It is also well documented that individuals often tend to overestimate
small probabilities and underestimate large probabilities (Thaler 1992;
Viscusi 1992).
We have already noted that individuals may be altruistic and derive
utility or well-being from others' utility or consumption. In addition, as
pointed out by Sen (1977b), individuals may take others' well-being into
account through some kind of commitment or social norms without
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
276 Articles
deriving any utility (or well-being) from this choice. In economics, this type
of assumption, where utility maximization cannot fully describe individual
behaviour, is still quite rare.I8 Sen (1977b) denoted people who are always
selfishly rational as 'rational fools,' since such behaviour will often imply a
sub-optimal outcome for all involved. As argued by Thaler (1992: 20) after
reviewing much experimental evidence that people often choose the
cooperative alternative, even in situations where economic theory would
predict them not to: 'Perhaps we need to give more attention to "sensible
cooperators"'.
Exogenous tastes
Implicit in (2) we have that the utility functions per se are exogenously given
and will not change with regard to consumption of various goods, i.e.
preferences or tastes are assumed to be fixed. This common assumption has
been questioned by Galbraith (1991), among others. For example, he argues
(1991: 129) that the 'central function [of modem advertising and salesmanship] is to create desires - to bring into being wants that previously did not
exist'. In a formalization of similar ideas, Dixit and Norman (1978) argue that
the amount of advertisement seems to be socially too large under a quite
general set of assumptions.
In their famous paper 'De gustibus non est disputandum', Stigler and
Becker (1977) argue, on the contrary, that it is not necessary, or useful, to
explain the obvious influence of advertisement on behaviour as change in
tastes.l9 Instead, they argue that the changed consumer behaviour can be
explained by the changed information (true or false) that the consumers
receive, and by changes in the 'technology' through a special kind of
household production function in order to produce 'commodities' such as
prestige." However, as explained by Pollak (1978), these arguments may be
useful in the explanation of consumer behaviour, but the analysis needed is
much more complicated in normative analysis, e.g. when we are dealing with
the welfare effects of advertisement. The reason is of course that if utility is
seen as a measure of what the government should maximize, it is important
how utility is defined. Consequently, if the relevant measure for what the
government should maximize is changing, the utility measure also needs to
change. If, on the other hand, utility is not primarily seen as a measure for what
the government should maximize, then we need to define a new independent
concept for this purpose (e.g. individual welfare or well-being), since the
maximization of an SWF that solely depends on utilities would not follow
from any (reasonable) ethical principles.
Another economist who has frequently questioned the assumption of
exogenously given tastes is Kenneth Boulding. Several years before the
Becker and Stigler article he wrote (Boulding 1969: 1-2): 'The most absurd of
all pieces of ancient wisdom is surely the Latin tag de gustibus non
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
277
disputandum. In fact, we spend most of our lives disputing about tastes.' He
then goes on exploring this:
One of the most peculiar illusions of economists is a doctrine that might
be called the Immaculate Conception of the Indifference Curve, that is,
that tastes are simply given, and that we cannot inquire into the process
by which they are formed. This doctrine is literally 'for the birds', whose
tastes are largely created for them by their genetic structures, and can
therefore be treated as a constant in the dynamics of bird societies. In
human society, however, the genetic component of tastes is very small
indeed.
(1969: 2)
Akerlof and Dickens (1982) and Elster (1982) provide other examples of
preference formation, as a result of cognitive dissonance and similar
arguments, which may have important policy-relevant welfare effects; see
also Veblen (1.899) for an early and powerful discussion.
Maybe a change in attitude towards the assumption of exogenous tastes is
under way. Recently, in his AEA presidential address, Victor Fuchs (1996: 16)
stated: 'I believe there is an analogy between the economics of values and the
economics of technology. Over the past several decades some economists
have begun to treat technology as endogenous. Now, a similar effort must be
undertaken for values.' Interestingly, one of the founders of the endogenous
growth theory, Paul Romer, discusses in a new book (Romer 1999) the
importance of changing preferences, both for explaining human behaviour
and for policy recommendation^.^'
We have seen that the social norms may be important for individual
behaviour and choice. But these norms, of course, are not constant over time.
The sociologist Coleman's (1990) most fundamental critique of neoclassical
economics seems to be the neglect of the formation of social norms and the
corresponding build-up of a 'social capital'.22In frequently discussed work,
Putnam et al. (1993) and Helliwell and Putnam (1995) emphasize the
importance of social capital to explain a large part of the big difference in
economic development between southern and northern Italy.
There seems to have been an increased interest in the importance of social
norms in economics for the last fifteen years or so. Elster (1989a, b), Frank
(1988) and Sugden (1986, 1989) provide thorough explanations of why and
how social norms occur. Lindbeck (1995, 1997) discusses the role of social
norms as incentives in the labour market, Ostrom (1990) gives several
empirical examples of successful local norms to overcome 'prisoner's
dilemma' and 'tragedy of the common' situations in relation to natural
resources management in a third world context, and Akerlof (1997) and Cole
et al. (1998) discuss social norms and the existence of distinct social classes.
Kreps (1997: 363) concludes a recent paper on the interaction between norms
and economic incentives: 'The results are likely to be messy. They will
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
278 Articles
involve activities unfamiliar to economics (e.g. theories of how preferences
are formed and reformed). But messy or not, they are important and must be
pursued.'
Well-behaved preferences
In order to end up with an operational ethical theory we must typically impose
some structure on individual preferences, such as reflexivity, transitivity and
local non-satiation; see, for example, Varian (1984). However, although
routinely made, these assumptions have in many cases been demonstrated not
to hold. For example, Thaler (1980) presented evidence supporting the
existence of an endowment effect, i.e. that people often demand much more in
compensation to give up a good, than they would be willing to pay to get it, and
that this seems to be the case also for small amounts of money where a
marginal analysis could be argued to be a reasonable approximation.
Kahneman and Tversky (1984) and Samuelson and Zeckhauser (1988)
discuss the related concepts of loss-aversion and status quo bias, respectively.
Slovic and Lichtenstein (1983) and Tversky et al. (1990) demonstrate and
discuss preference reversals, that people in a choice between goods A and B
prefer A, but that they still would be willing to pay more for good B than for
good A.
Chaos and evolution
The contributions in evolutionary economics and chaos theory often focus on
the dramatic outcome that may result from small changes in a crucial variable,
due to strong non-linearities and unstable equilibria; see Boulding (1991) for
an introduction to evolutionary economics and Kemp (1997) for a nontechnical introduction to chaos and non-linear dynamic systems. But in a
complex dynamic system such phenomena are often extremely difficult to
predict before they occur. Consequently, economists tend to assume they will
not occur, at least not due to a discussed policy measure. Again, stress of
ecological systems due to human activities constitutes a perhaps obvious
example when such assumptions may be doubtful. But, as history shows,
human systems, cultures and political systems may also be very unstable, and
relatively small changes in the environment may cause dramatic effects, such
as a switch to a development path towards democracy instead of dictatorship
(or, of course, the other way around). In a recent and thoughtful book Thurow
(1996) discusses the rapidly increasing inequalities in the capitalist world, and
in particular in the US, during the last two decades: 'How far can inequality
widen and real wages fall before something snaps in a democracy? No one
knows, since it has never before happened. The experiment has never been
tried' (1996: 261). Cf. Ayres (1998) on the same issue.
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
279
4 THE 'OLD WELFARE ECONOMICS', THE 'NEW
WELFARE ECONOMICS' AND THE 'NEW, NEW WELFARE
ECONOMICS'
Basically all decisions taken by a government at a national or local level will
affect a large number of individuals, and it is almost always possible to find
both people who gain and people who lose from political decisions. To deal
with such issues, one has to trade what is good for some against what is bad for
others. This was the natural conclusion drawn by classical economists and
philosophers in the utilitarian tradition, such as Mill, Bentham and
Edgeworth.
The revolution of the new welfare economics
However, in the twentieth century, this view has been considered
'unscientific' and impossible to test and become gradually less common.
Marshall, for example, who was first a convinced utilitarian became more
influenced by these 'new' ideas. In particular, ethics and moral values were
considered to be unscientific, and should therefore be removed from
economics. Robbins (1932: 148) argued that 'it does not seem logically
possible to associate the two studies [economics and ethics] in any form but
mere juxtaposition'. As noted by Sen (1987: 2), 'he was taking a position that
was quite unfashionable then, though extremely fashionable now'.
At the core of this debate between the old welfare economics and the new
welfare economics lies the concept of utility. As we have seen, the term
'utility' is used with many different meanings. There are at least two
fundamentally different meanings and uses of the concept utility frequently
used in economics; they are sometimes called the experience model and the
preference model, see e.g. Kahneman and Varey (1991). The different
meanings have often caused, and still cause, great confusion. Briefly, the
experience model derives utility only through individual experience, i.e.
through mental states or states of consciousness. Originally, in the
Benthamian tradition, these mental states were taken to be pleasure or
happiness. Later on, this has been modified to a much broader concept; see
e.g. Hausman and McPherson (1994). Even though it is of course difficult to
measure (or estimate) such a vague concept as utility per se, there do exist
various psychological methods of measuring individual well-being; see e.g.
Tversky and Griffin (1991) and Kahneman and Varey (1991).23
The preference model instead assumes that what increases an individual's
utility is simply what the individual prefers, which is revealed by their actions.
In its purest form, no assumption is made of whether the individual will be
happier or better off (defined independently of the actions). This model is of
course in general much easier to work with empirically when it comes to
purely explanatory models, if we accept revealed preferences, since we need
then only to observe market transaction^.^^ According to the new welfare
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
280 Articles
economics, it is beyond the domain of economists to ask why an individual
prefers a certain good or service to a certain extent.
The statement 'Individuals will always maximize their own utility' is then
either wrong or a tautology. It is wrong if we define utility as a measure of
well-being, as in the experience model, or as some other goodness-related
measure defined independently of the action.25As expressed by Hausman and
McPherson (1994: 263): 'What people prefer may not be good for them,
because people make mistakes, and because they may prefer to sacrifice their
own well-being in pursuit of some other end.' On the other hand, it is a
tautology if the maximization of utility is defined solely by the fulfilment of
preferences, and the preferences are defined solely by the actual actions
~ n d e r t a k e nThe
. ~ ~new welfare economics, which is sometimes considered to
be a basic pillar of neoclassical theory, therefore avoids interpersonal utility
comparisons, which are frequently said to be unnecessary, unscientific,
useless, etc. Since interpersonal comparisons of utility were ruled out, the
Pareto criterion became the only guidance for action. But since almost all
decisions imply that some will gain and some will lose, the Pareto criterion
provides of course little help in practice.
Some extreme Paretians go even further and argue that one (the
government) should not do anything unless a Pareto improvement is achieved.
This does not necessarily follow from the Pareto criterion itself,27but if it is
interpreted in this way it will obviously bias the decisions in a conservative
direction. For example, Robinson (1962) pointed out that with this view, the
only justifiable economic policy would be to do nothing. Furthermore, it may
even be that any economic policy (including doing nothing) would be
incompatible with the Pareto criterion, interpreted in this (rather extreme)
way. Consider the following tax-collecting paradox which arises from the fact
that we act in a historical context. The question 'Should we collect any income
taxes next year?' is then indeed difficult to answer. We can with reasonable
certainty assume that most people would prefer not to pay any income taxes
themselves (even though most people would prefer others to pay taxes). For
this reason, we should not decide to collect any income taxes for the next year
(unless people would like to make any voluntary contributions). However, we
do collect such taxes now, and many people, notably the poor and ill, would
obviously be much worse off if we did not collect them. Hence, we could then
not decide not to collect any income taxes for the next year either. Thus, we
can neither collect taxes nor not collect them! Harrod (1938) pointed out early
the difficulties for the new welfare economics to say anything about policy.
If the incomparability of utility to different individuals is strictly pressed,
not only are the prescriptions of the welfare school ruled out, but all
prescriptions whatever. The economist as an adviser is completely
stultified, and, unless his speculations be regarded as of paramount
aesthetic value, he had better be suppressed completely.
(1938: 397)
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
28 1
Consequently, the huge amount of social choice literature, which, following
Arrow (1951), most often rules out interpersonal comparisons of utilities,
seems to be largely focusing on negative results and has very little to say
which has direct policy relevance; see e.g. Sen (1986) for a survey and Elster
and Hylland (1986) for an extensive treatment of social choice literature. As
noted by many, this has to do with the fact that the 'information base' on which
the decision should be taken is so drastically reduced without interpersonal
comparisons of utilities. According to Hammond (1991: 206): 'Indeed, after
Arrow's original contribution, not much of this literature seems all that useful
in retrospect.'
Potential compensation criteria
Since a situation where economists could not give any policy
recommendations whatsoever was not considered very appealing, at least by
the economists themselves, there was a large demand for some other criterion
which would be more useful in practice and which was still 'scientific' and
'objective'. To some, the well-known potential compensation criteria
proposed by Hicks (1939) and Kaldor (1939) were the answer to that
demand, and they have frequently been associated with the new welfare
economics.28It is obvious that these criteria proved extremely useful, and
they are generally considered to constitute the welfare-theoretic
foundation of cost-benefit analysis.29For example, in a recent textbook on
environmental economics, Freeman (1993: 87) writes about the HicksKaldor criterion:
This is the efficiency criterion of the new welfare economics. According
to the efficiency criterion, the objective of social policy is to maximize
the aggregate value of all the goods and services people receive,
including environmental and resource services.
Whether this is more 'scientific' and 'objective' than the old welfare
economics is not at all clear, however. Samuelson (1950) showed that no
policy recommendations could be derived on the basis of the Hicks-Kaldor
criteria, or from the extension proposed by Scitovsky (1941).30According to
Boulding (1969: 8): 'Any decision involving other people obviously involves
these interpersonal comparisons.' This is clear since if the losers are not
compensated, and we still argue that a decision should be taken, what we are
doing is exactly an interpersonal comparison of utilities: we are saying that,
from a social perspective, the utility increase from those who gain is worth
more than the utility loss from those who lose. Of course, this is by no means
more objective, or less controversial, than classical utilitarianism. On the
other hand, if we do compensate the losers 'then the overall outcome - after
compensation - is a Pareto improvement, and then there is no need for a
compensation test as a supplement to the Pareto principle. So the
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
282 Articles
compensation criteria are either unconvincing or redundant' (Sen 1987: 33,
footnote 4, italics in original).
However, although there is nothing ethically appealing about the
compensation criteria per se, they may still be useful in normative economics
as approximations to a utilitarian view (which, in turn, is an approximation of
our original goodness measure). Under some conditions, such as relatively
small distributional consequences, such an approximation may be j~stified.~'
The silent counter-revolution
The new, new welfare economists, sometimes called the 'new cardinalists',
have returned to the view that interpersonal comparisons of utility, however
difficult these may be, are necessary for any policy recommendation regarding
more than one individual. The classic paper on optimal income taxation by
Mirrlees (1971) is generally considered to be the beginning of the new, new
welfare economics; see e.g. Stiglitz (1985) and Arnott (1994). This 'counterrevolution' has typically been made in a rather 'quiet' and discrete manner,
without much rhetoric, at least compared to the 'revolution' of the new welfare
economics, even though there exist notable exceptions such as Blackorby
(1990) and Ng (1997):
if ever we are to have something consistent and correct to say about
policy in the real world, we shall have to learn how to make and make
use of, interpersonal comparisons of well-being. This is an activity that is
often dismissed as simply non-scientific or worse, impossible. The word
'scientific' is often invoked as an almost magical incantation. . . . The
general view behind this position is that it is more scientific to use
efficiency values rather than distributional values. To describe the choice
of the former as scientific renders the word practically meaningless. In
any case, my claim is that our choice is between making interpersonal
comparisons of utility or in having little or nothing to say.
(Blackorby 1990: 749)
Furthermore, as argued by Davidson (1986), we often make interpersonal
comparisons of well-being without finding this very difficult:
Some cases are hard, of course; the same can be said concerning some
decisions which affect one person only. But on the whole we do not
experience the problem of comparing the interests of different people as
being harder in kind or degree than comparing the interests of our own.
Naturally we are more apt to be ignorant of the interests of others than of
our own; but this is a variable we have no trouble in accounting for.
(1986: 195)
A possible difference between the old welfare economics and the new, new
welfare economics may be that the latter typically interprets the utility concept
Ethics and economic policy recommendations
283
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
--
-
in a broader way, and that it also often simultaneously deals with incentive
effects and distributional issues in a second-best general equilibrium
framework. The objective is then to maximize a (welfaristic) SWF given the
fact that individuals (and other economic agents) will maximize their
objective functions, and given that the government does not have all relevant
information, or access to all possible policy instruments. For example, in the
classic Mirrlees tax problem, the government could not observe individual
abilities to work or the amount of leisure consumed.
However, it should be made clear that although interpersonal comparisons
of utility are in general necessary for deriving policy conclusions, they are not
suflcient, since we still need some social choice rule for how we should act
given this utility i n f o r m a t i ~ n . ~ ~
Consumer surplus and 'exact' welfare measures
In order to measure welfare changes in practice we must typically add a few
standard simplifications including perfectly competitive markets, zero
transaction and search costs, that all other prices (except the ones we focus on)
are non-distorted, and that firms can be characterized as profit-maximizers.
Although these assumptions are never strictly fulfilled, they may (or may not)
serve as useful approximations; cf. Blackorby (1990) and Rappaport (1996).
With discrete non-continuous changes, as most changes obviously are, the
most commonly used welfare measure is probably still the change in
Marshallian consumer surplus; i.e. the area between the demand curve and the
price line in a price-quantity diagram. Besides ignoring all distributional
effects, this measure has some other theoretical drawbacks arising from the
fact that individual marginal utility of income is not constant during this
discrete change (of prices or quantities). To deal with these problems, one can
instead use the well-known monetary welfare measures introduced by Hicks
(1943), i.e. the compensating variation (CV) and the equivalent variation (EV)
corresponding to the maximum willingness to pay for an improvement and the
minimum willingness to accept for not having an improvement, respectively;
see e.g. Varian (1984) or Freeman (1993) for definition^.^^ Willig (1976)
derived bounds for the differences between the Hicksian and the Marshallian
welfare measures for price changes, and concluded that the differences are
most often negligible (if the income effect is not very large). Randall and Stoll
(1980) attempted a similar exercise for quantity changes (of a public good),
and Hanemann (1991) clarified their result and showed that the differences
will be larger if private consumption is far from being a perfect substitute for
the public good; see also Becht (1995) for a survey of related literature. The
current practice is to call these Hicksian measures exact welfare measures (see
e.g. Hausman 1981, or Becht 1995) in order to separate them from the
Marshallian consumer surplus, which seems, in the light of the simplifications
made above, somewhat m i ~ l e a d i n g . ~ ~
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
284 Articles
In modern applied welfare economics this problem, arising from
differences in marginal utility of income within a single person (due to
changes in prices or provision of public goods), is generally emphasized much
more than the distributional effects arising from differences in marginal utility
of income between different persons. This is so even though the variation of
the marginal utility of income, for most reasonable SWFs, will be much larger
between the individuals (at least for relatively small changes in prices or
provision of public goods). Presumably, this is a result of the still-lasting
strong influence of the new welfare economics; perhaps it also reflects a timelag before an influential breakthrough of the new, new welfare economics in
applied welfare economics.
Furthermore, so far we have discussed the problem of welfare change given
perfect information. But, of course, in reality the uncertainties involved are
often large. Although economists have begun to study imperfect information
extensively during recent decades, it is still often neglected in applied studies
in order to keep the complexity at a manageable
Sometimes this may
be a reasonable simplification; sometimes it is not. An example where this
simplification seems highly inappropriate is cost-benefit analysis related to
global warming and the greenhouse effect. Focusing solely on what is thought
to be the most likely outcome may seriously underestimate expected costs,
since, for example, the expected cost of an unlikely catastrophic event may be
much larger than the cost of the most likely outcome.36 Although most
economists seem to agree on this point (at least, few argue oppositely), most
work in this area still ignores this problem.
5 CONCLUSIONS
We have seen that the link from fundamental ethics towards policy
recommendations is often very weak and that it depends on several very strong
assumptions. This is so even though the economics literature is becoming
fairly rich in questioning these restrictive assumptions separately. However,
in applied work it is unfortunately much more difficult to avoid these strong
assumptions. As expressed by Goodwin (1991: 286):
Hence we have a relatively unscientific applied welfare economics, and a
scientific theoretical welfare economics, with no better bridges between
the two today than was to be found at the beginning of the century in
Marshall's system of inconsistent, or ambiguous, inclusiveness.
What is left of our original goodness measure when arriving at the policy
conclusions after all simplifying assumptions is certainly no trivial question,
and it is important to realize that the correspondence between the two may be
quite poor. Still, this is hardly a good argument for economists not to deal with
normative issues, since it is hard to see what could be gained from ignoring a
systematic treatment of all the obviously important issues of what the
I
I
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
285
government should do when more than one single individual is affected.
Furthermore, as history shows, economists will not give up the possibility of
influencing political decisions and it is then far better if this influence is
explicitly built on the only available solid ground, namely on some
fundamental moral values which are then linked to economic theory, however
loose this link may be, rather than if it is built on the sand of some imagined
objective and value-neutral economic science.
Then there is an everlasting process for economists to improve this weak
link by choosing more reasonable assumptions and improving our
understanding of human behaviour. In this process, it is important to be as
explicit as possible concerning basic value judgements and to realize that
humans act in a social context, and hence that we may need to go beyond
reductionist individualism. For example, a well-working economy needs
well-working institutions, which is of course a far from new insight (see e.g.
Veblen 1899; or Knight 1924).37Nevertheless, it is neglected in much welfaretheoretic work, partly due to the obvious fact that modelling evolution and
change of the institutional environment is quite difficult. Again, the
seriousness of this neglect varies largely due to the circumstances. There is
much evidence that the development of proper institutions is particularly
important for the developing world and transitional economies, and hence that
neglecting these issues here is often a very serious simplification.
Furthermore, as proposed by Akerlof (1984) and Hirshleifer (1985), we
need to learn more from our colleagues in social science, regarding
fundamental issues such as preference formation, decision behaviour, and the
creation of values, institutions and social norms. Hirshleifer (1985: 52) argued
that 'good economics will also have to be good anthropology and sociology
and political science and psychology'. Akerlof (1984: 6) 'would like to think
that psycho-socio-anthropo-economics is at the beginning of a period when
many people will be working in this area. Thirty years ago mathematical
economics was probably in a similar stage of development.' Since these words
by Akerlof were written, it is probably fair to say that other social sciences
have actually influenced economic ideas sub~tantially,~~
but hardly
sufficiently.
Finally, even though economics, like all (social) sciences, will always have
to rely on a number of strong and stylized assumptions, there are a number of
very general points which it might be wise to consider (again and again), even
though they may seem trivial:
1
2
There is a trade-off between simplicity and realism (see e.g. Sen 1985).
Sometimes the larger realism is worth the cost in terms of increased
complexity, reflected in more difficult analysis andor less rigour, fewer
unambiguous results, and less obvious and straightforward policy
conclusions.
There is a choice concerning which simplifying assumptions to make.
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
286 Articles
3
This choice should preferably be undertaken so as to maximize the
usefulness, and degree of realism of the analysis, for a given level of
complexity. Hence, it is important not to choose certain assumptions for
dogmatic reasons.
It is important to consider how various simplifying assumptions affect the
results and hence the policy conclusions. This seems particularly
important for assumptions which are often routinely made without much
deeper reflection.
There is certainly no reason for economists to stop dealing with normative
issues and policy recommendations. But in doing so, perhaps sometimes a
little more modesty would not hurt.
Goteborg University
ACKNOWLEDGEMENTS
I have benefited largely from discussions with Daniel Bromley and from his
comments on earlier versions. I have also received valuable comments from
Ramon Lopez, Thomas Sterner, Rick Wicks, the editor and anonymous
referees. Financial support from the Swedish Transport and Communications
Research Board (KFB) is gratefully acknowledged. The usual disclaimer
applies.
NOTES
1 The term 'operational' will be used with this specific meaning throughout the
paper; see Blaug (1992: 87-91) for somewhat different meanings of this term.
2 Unfortunately, the definitions of deontological and teleological theories vary
somewhat in the literature. Rawls (1973: 24) defines an ethical theory as
teleological if 'the good is defined independent from the right, and then the right
is defined as that which maximizes the good'. Howarth (1995) says that an action
that is based on duty (a 'Kantian' motivation), instead of preference fulfilment, is
'deontologically motivated'. Etzioni (1988) seems to use 'deontology' in a
similar way.
3 However, as pointed out by Hammond (1991: 201), when the space of
consequences considered is extended to take everything of ethical significance
into account, consequentialism becomes a tautology.
4 See Blackorby and Donaldson (1992) who apply the ethical view of Singer in an
economic analysis of animal exploitation, and Ng (1995) for an attempt 'towards
welfare biology'.
5 A possible solution to this problem would be to include everything of ethical
relevance, and not just well-being, in the utility function; see Hammond (1996).
Sen (1987: 40, footnote 13) argues against such very broad utility interpretation:
'As a defence of utility-based calculations this is tautologous and adds little to the
discussion.'
6 This seems to be the most common interpretation of the notion B-S SWF.
Unfortunately, sometimes different meanings of this term are used in the
literature.
7 It is not very difficult to extend the analysis to take 'technological' externalities
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
287
into account, which is frequently done in neoclassical economic theory; see e.g.
Baumol and Oates (1988).
Again, this is not to say that happiness and well-being are taken to be the same
thing, but, as argued by Frank (1997), it seems reasonable that happiness is an
important element in most people's well-being.
For relation to optimal taxation, see Boskin and Sheshinski (1978), Tuomala
(1990) and Persson (1995). See also Kocherlakota (1996) for a recent discussion
in relation to 'the equity premium puzzle', i.e. why the rate of return on stocks has
been that much larger compared to that on bonds. He discusses a utility concept
which depends on relative consumption in the time dimension, i.e. that utility
today is a function of both current consumption and yesterday's consumption.
Zamagni (1995) provides a comprehensive collection of papers on economics
and altruism. See e.g. Bergstrom (l982), Johansson (1992) and Jones-Lee (1992)
for applications to cost-benefit analysis. Charity and private provision of public
goods are considered by among others Andreoni (1989, 1990), Bergstrom et al.
(1986) and Sugden (1982, 1984). Altruism within the family is considered by
Becker (1976), Bergstrom (1989, 1995) and Bernheim and Stark (1988). Why
altruism occurs in an evolutionary environment and altruism in relation to
sociobiology have been studied by Bergstrom (1995), Bergstrom and Stark
(1993), Simon (1993) and Stark (1995).
However, it should be noted that utilitarianism is not a subset of B-S SWF. For
example, we may perfectly well think of utilitarianism where the utility functions
are not independent of each other. Furthermore, utilitarianism can also
incorporate animals; see e.g. Singer (1979). The order in which additional
restrictions are imposed in this paper is thus not unambiguous, but rather made
for illustrative purposes.
An often used SWF is characterized by constant social inequality aversion (in
terms of utilities) so that W = uil"l(l - cp), where cp is the social inequality
In (ui).) It is
aversion. (When the inequality aversion is equal to one, W =
straightforward to see that this expression simplifies to classical utilitarianism
when cp = 0 and to a so-called Rawlsian maxi-min SWF when cp goes to infinity.
Note that we have dealt with distribution of utilities, not consumption. For
example, if both individuals have an identical utility function u = dc then in state
I Allan receives 100 consumption units and Boris 25 units with probability 0.5,
and vice versa with probability 0.5. In state 11, both Allan and Boris receive 49
consumption units each.
See, however, Broome (1991a, 1996c) for a useful reinterpretation of the
theorems by Harsanyi.
Alternatives to utilitarianism include the commonly discussed contract theories
by Rawls (1973) and Nozick (1974).
Some utilitarians would argue that these kinds of expressed views or social norms
do not reflect fundamental, but rather instrumental, values. For example, a
situation where the difference in the marginal utility of income is estimated
between different persons may per se result in individual suffering, and hence
reduced overall utility.
Consider the Al-and-Bob example. Starting with the betterness relation, we
assume that it is not 'good' to redistribute money from A1 to Bob. But if we apply
a utilitarian SWF (either weighted or unweighted) it would increase social
welfare to do so. If, on the other hand, we impose an additional simplification that
we will count their marginal utility of consumption as equal (if their incomes are
equal) then a redistribution will not increase social welfare, which is in
accordance with the betterness relation.
In the economics of altruism, this type of commitment is sometimes termed
xi
xi
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Articles
'genuine altruism' (Edwards 1992; Johansson 1997), as opposed to the more
common pure altruism (altruism with regard to others' utilities) and paternalistic
altruism (altruism with regard to some specific component of others' utilities)
where the altruistic concern is modelled inside the utility function.
They provide similar arguments for other cases which may seemingly be
incompatible with the assumption of exogenous and stable preferences, including
the addiction to drugs.
However, one needs to be careful when interpreting Becker and Stigler since they
use notions such as 'commodities' with a somewhat different meaning compared
to most economists. As expressed by Ackerman (1997: 661): 'The reader who
lacks an English-to-Becker dictionary must remember that what Becker calls
commodities are what others would call experiences or satisfactions, while the
commodities visible to the rest of us are, for Becker, inputs purchased by
households in order to produce commodities.' Furthermore, as noted by Akerlof
(1997), Becker has in later writings adopted a less rigid position towards
changing preferences.
The recent contribution by Hanemann and Kristrom (1995) on preferenceuncertainty seems to be a step in that direction in environmental economics.
Frank (1992) provides a thorough review of this comprehensive work.
Dawkins (1980) provides an overview of the valuation of animal welfare.
The revealed preference assumption is not useful for explaining the choice as
such, since the choice is simply assumed to reflect the preferences (Bromley
1989). However, the assumption is useful to explain preferences, such as whether
people seem to prefer white or brown bread.
Still, it may be a both useful and reasonable approximate assumption, but that is
of course a fundamental difference.
With the same type of 'logic' we can 'prove' that Bangladesh is a part of Africa
(which, of course, it is not). If Africa is defined as the continent which consists of
the poor countries of the earth, and Bangladesh is considered a poor country
(which it is), then Bangladesh must, by necessity, be a part of Africa!
The Pareto criterion, as it is most often interpreted, says that an action should be
undertaken if a Pareto improvement is achieved. It does not say that an action
should not be undertaken if a Pareto improvement is not achieved.
The criteria are defined somewhat differently by Hicks and by Kaldor. Acommon
description of these criteria is that an economic allocation X is superior to an
allocation Y if and only if it is possible to reach an allocation Z through
redistribution fromX such that Zis preferred to Y according to the Pareto test. One
problem with this statement is that, when relative prices are affected by the
decision, it may be that two separate allocations can be potentially Pareto
dominated by each other; see e.g. Scitovsky (1941) or Samuelson (1950).
Boadway (1974) has shown that similar inconsistencies may occur with the
related aggregate willingness-to-pay welfare criterion. There is by now a very
large literature on the aggregation problems associated with various
compensation criteria, i.e. arising from the fact that we rule out direct
interpersonal comparisons of utility.
See Hubin (1994) for a recent critique of this view.
Little (1957) proposed an alternative modification to take distributional concern
into account. An action should then be undertaken if any of the Hicks-Kaldor
criteria and the Scitovsky criterion is fulfilled and at the same time the change in
income distribution is considered positive. However, as discussed, e.g. by Nath
(1969), the Little criterion too many have problems with inconsistency and
contradictions.
Alternatively, one may argue that under some conditions there are other possible
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
32
33
34
35
36
37
38
289
measures such as income taxes which are better suited to deal with distributional
issues. Christiansen (1981) and Boadway and Keen (1993) derive sufficient
conditions when the basic Samuelson rule may be applied in the presence of an
optimal non-linear income tax. Of course, the existence of perfect differentiated
lump-sum taxes is also a sufficient condition (Atkinson and Stiglitz 1980), but in
the public finance literature this very unrealistic assumption is often considered
to be of limited policy relevance.
See the discussion between Rosenbaum (1995, 1996) and Davis (1991, 1996) on
whether, and in what sense, interpersonal comparisons of utility are positive,
normative or value-laden. Elster and Roemer (1991) provide a comprehensive
treatment of several aspects of interpersonal comparisons of utility.
During quantity changes, these measures are sometimes known as
'compensating' and 'equivalent surplus', respectively (Freeman 1993).
There is still an on-going discussion about whether the large disparities typically
found between willingness-to-pay and willingness-to-accept measures in
surveys for public goods can be explained within the framework of conventional
economic theory, or whether non-conventional phenomena such as endowment
effects play an important role; see Kahneman and Knetsch (1992a, b) and Smith
(1992).
The fact that uncertainty and imperfect information have played a minor role for
a long time in economic theory should, however, not be taken as an indication
supporting the view that most economists were unaware of this strong limitation;
see e.g. Knight (1921). Hirshleifer and Riley (1992) provides a good textbook
treatment of economics and information.
From Jensen's inequality we have that E(C(X)) > C(E(X)) if C(X) is a convex
function, where E denotes expected value, Cis the social cost, and X is a variable
describing the outcome in physical terms (e.g. temperature increase due to global
warming). Hence, a sufficient condition is that C is increasing in X at an
increasing rate (i.e. that both the first and the second derivatives are positive).
See Bromley (1989) and Hodgson (1988) for good textbook treatments of
institutional economics, and North (1990) for an example of the new institutional
economics which is generally more closely related to conventional neoclassical
economics. Hodgson et al. (1994) provides a good overview of published work in
institutional and evolutionary economics.
See Goodwin et al. (1997) and Ackerman et al. (1997) for surveys of this
promising development.
REFERENCES
Abramovitz, M. (1979) 'Economic growth and its discontents', in Michael Boskin
(ed.) Economics and Human Welfare, New York: Academic Press, pp. 3-21.
Ackerman, F. (1997) 'Consumed in theory: alternative perspectives on the economics
of consumption', Journal of Economic Issues 3 l(3): 65 1-64.
Ackerman, F., Kiron, D. Goodwin, N.R., Harris, J.M. and Gallagher, K. (1997)
Human Well-Being and Economic Goals, Washington DC: Island Press.
Akerlof, G.A. (1984) An Economic Theorist's Book of Tales, Cambridge: Cambridge
University Press.
(1997) 'Social distance and social decisions', Econometrica 65(5): 1005-27.
Akerlof, G.A. and Dickens, T.W. (1982) 'The economic consequences of cognitive
dissonance', American Economic Review 72: 307-19.
Aldred, J. (1994) 'Existence value, welfare and altruism', Environmental Values 3(4):
381402.
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
290 Articles
Alexander, R.D. (1987) The Biology of Moral Systems, New York: Aldine de
Gruyter.
Andreoni, J. (1989) 'Giving with impure altruism: applications to charity and
Ricardian equivalence', Journal of Political Economy 97(6): 1447-58.
(1990) 'Impure altruism and donations to public goods: a theory of warm
glow giving', Economic Journal 100: 464-77.
Amott, R. (1994) 'Reflections on optimal tax theory', in J.M. Quigley and E.
Smolinsky (eds) Modem Public Finance, Cambridge, Mass.: Harvard University
Press, pp. 257-89.
Arrow, K.A. (1951) Social Choice and Individual Values, New York: John Wiley &
Sons.
Atkinson, A.B. and Stiglitz, J.E. (1980) Lectures in Public Economics, Singapore:
McGraw-Hill.
Ayres, R.U. (1998) Turning Point: The End of the Growth Paradigm, London:
Earthscan.
Azar, C. (1995) Long-term Environmental Problems, Economic Measures and
Physical Indicators, PhD thesis, Institute of Physical Resource Theory, Chalmers
University of Technology, Goteborg, Sweden.
Baumol, W.J. and Oates, W.E. (1988) The Theory of Environmental Policy,
Cambridge: Cambridge University Press.
Becht, M. (1995) 'The theory and estimation of individual and social welfare
measures', Journal of Economic Surveys 9: 53-87.
Becker, G.S. (1976) The Economic Approach to Human Behavior, Chicago:
University of Chicago Press.
Bergstrom, T.C. (1982) 'When is a man's life worth more than his human capital?', in
M.W. Jones-Lee (ed.) The Value of Life and Safety, Amsterdam: North-Holland,
pp. 3-26.
(1989) 'A fresh look at the rotten kid theorem and other household
mysteries', Journal of Political Economy 97(5): 1138-59.
(1995) 'On the evolution of altruistic ethical rules for siblings', American
Economic Review 85(1): 58-81.
Bergstrom, T.C., Blume, L. and Varian, H. (1986) 'On the private provision of public
goods', Journal of Public Economics 29: 2 5 4 9 .
Bergstrom, T.C. and Stark, 0 . (1993) 'How altruism can prevail in an evolutionary
environment', American Economic Review 83(2): 149-55.
Bemheim, B.D. and Stark, 0. (1988) 'Altruism within the family reconsidered: do
nice guys finish last?', American Economic Review 78(5): 1034-45.
Blackorby, C. (1990) 'Economic policy in a second-best environment', Canadian
Journal of Economics 23: 748-7 1.
Blackorby, C. and Donaldson, D. (1984) 'Social criteria for evaluating population
change', Journal of Public Economics 25: 13-33.
(1991) 'Normative population theory: a comment', Social Choice and
Welfare 8: 261-7.
(1992) 'Pigs and guinea pigs: a note on the ethics of animal exploitation',
Economic Journal 102: 1345-69.
Blaug, M. (1992) The Methodology of Economics, or How Economists Explain,
second edition, Cambridge Surveys of Economics Literature. Cambridge, New
York and Melbourne: Cambridge University Press.
Boadway, R. (1974) 'The welfare foundations of cost-benefit analysis', Economic
Journal 84: 926-39.
Boadway, R. and Keen, M. (1993) 'Public goods, self-selection and optimal income
taxation', International Economic Review 34(3): 463-78.
Boskin, M.J. and Sheshinski, E. (1978) 'Optimal income redistribution when
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
29 1
individual welfare depends on relative income', Quarterly Journal of Economics
92: 589-602.
Boulding, K. (1969) 'Economics as a moral science', American Economic Review
59(1): 307-19.
(1991) 'What is evolutionary economics?', Journal of Evolutionary
Economics l(1): 9-17.
Brekke, K.-A. (1997) Economic Growth and the Environment, Cheltenham: Edward
Elgar.
Bromley, D. (1989) Economic Interests and Institutions: The Conceptual
Foundations of Public Policy, Oxford and New York: Blackwell.
(1992) 'The ideology of efficiency: searching for a theory of policy
analysis', Journal of Environmental Economics and Management 19: 86-107.
Broome, J. (1991a) Weighing Goods, Cambridge, Mass.: Basil Blackwell.
(1991b) 'Utility', Economics and Philosophy 7: 1-12.
(1991~)'A reply to Sen', Economics and Philosophy 7: 285-7.
(1992) Counting the Cost of Global Change, Cambridge: The White Horse
Press.
(1996a) 'The value of life and the value of population', Journal of
Population Economics 9(1): 3-18.
(1996b) 'The welfare economics of population', Oxford Economic Papers
48(2): 177-93.
(1996~)'Economic analysis and the structure of good: transferring utility
theory from preferences to betterness', in F. Farina, F. Hahn and S. Vannucci (eds)
Ethics, Rationality, and Economic Behaviour, Oxford and New York: Clarendon
Press.
Chow, G.C. (1997) 'Challenges of China's economic system for economic theory',
American Economic Review 87(2): 321-7.
Christiansen, V. (1981) 'Evaluation of public projects under optimal taxation', Review
of Economic Studies 48: 447-57.
Cole, H.L., Mailath, G.J. and Postlewaite, A. (1998) 'Class systems and the
enforcement of social norms', forthcoming in Journal of Public Economics.
Coleman, J.S. (1990) Foundations of Social Theory, Cambridge, Mass.: Harvard
University Press.
Conlisk, J. (1996) 'Why bounded rationality?', Journal of Economic Literature 34
(2): 669-700.
Dasgupta, P. (1993) An Inquiry into Well-Being and Destitution, Oxford: Clarendon
Press.
Davidson, D. (1986) 'Judging interpersonal interests', in J. Elster and A. Hylland
(eds) Foundations of Social Choice Theory, Studies in Rationality and Social
Change Series, Cambridge: Cambridge University Press, pp. 195-2 11.
Davis, J.B. (1991) 'Interpretation of interpersonal utility comparisons: positive,
normative or descriptive?', Journal of Income Distribution 1: 73-90.
(1996) 'Value-ladenness in economics: reply to Rosenbaum', Journal of
Economic Methodology 3(1): 121-5.
Dawkins, M. (1980) Animal Suffering: The Science of Animal Welfare, London:
Chapman and Hall.
Diamond, P.A. (1967) 'Cardinal welfare, individualistic ethics, and interpersonal
comparisons of utility: comment', Journal of Political Economy 75: 765-6.
Dixit, A. and Norman, V. (1978) 'Advertising and welfare', Bell Journal ofEconomics
9(1): 1-17.
Dixon, H.D. (1997) 'Controversy - economics and happiness', Economic Journal
107: 1812-14.
Duesenberry, J.S. (1949) Income, Saving, and the Theory of Consumer Behavior,
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
292 Articles
Cambridge, Mass.: Harvard University Press.
Easterlin, R.A. (1974) 'Does economic growth enhance the human lot? Some
empirical evidence', in Paul A. David and Melvin Reder (eds) Nations and
Households in Economic Growth: Essays in Honor of Moses Abramovitz, Palo
Alto, Calif.: Stanford University Press, pp. 89-125.
(1995) 'Will raising the incomes of all increase the happiness of all?',
Journal of Economic Behavior and Organization 27: 3 5 4 7 .
Edwards, S.F. (1992) 'Rethinking existence values', Land Economics 6: 120-2.
Elster, J. (1982) 'Sour grapes - utilitarianism and the genesis of wants', in A.K. Sen
and B. Williams (eds) Utilitarianism and Beyond, Cambridge University Press,
pp. 219-38.
(1989a) 'Social norms and economic theory', Journal of Economic
Perspectives 3(4): 99-117.
(1989b) The Cement of Society, Cambridge: Cambridge University Press.
Elster, J. and Hylland, A. (eds) (1986) Foundations of Social Choice Theory, Studies
in Rationality and Social Change, Cambridge University Press.
Elster, J. and Roemer, J.E. (eds.) (1991) Interpersonal Comparisons of Well-Being,
Studies in Rationality and Social Change, Cambridge University Press.
Etzioni, A.K. (1988) The Moral Dimension: Toward a New Economics, New York:
The Free Press, Macmillan.
Frank, R.H. (1985a) 'The demand for unobservable and other nonpositional goods',
American Economic Review 75: 101-16.
(1985b) Choosing the Right Pond. Human Behavior and the Quest for
Status, Oxford University Press.
(1988) Passions within Reason: The Strategic Role of the Emotions, New
York: WW Norton.
(1991) 'Positional externalities', in R. Zeckhauser (ed.) Strategy and
Choice, Cambridge, Mass.: MIT Press.
(1992) 'Melding sociology and economics: James Coleman's foundations of
social theory', Journal of Economic Literature 30: 147-70.
(1997) 'The frame of reference as a public good', Economic Journal 107:
183247.
Freeman, A.M. I11 (1993) The Measurements of Environmental and Resource Values,
Washington: Resources for the Future.
Friedman, M. (1947) 'Lerner on the economics of control', Journal of Political
Economy 55, pp. 405-16.
(1994) [I9531 Essays in Positive Economics, Chicago: University of
Chicago Press. Reprinted as chapter 9 in D.M. Hausman (ed.) (1994) The
Philosophy of Economics -An Anthology, second edition, Cambridge University
Press.
Fuchs, V.R. (1996) 'Economics, values, and health care reform', American Economic
Review 86: 1-24.
Galbraith, J.K. (1991) The Afluent Society, fourth edition, London: Penguin Books.
Goodwin, N.R. (1991) Social Economics: An Alternative Theory, Volume One,
London: Macmillan.
Goodwin, N.R., Ackerman, F. and Kiron, D. (1997) The Consumer Society,
Washington DC.: Island Press.
Hamlin, A.P. (1996) Ethics and Economics, The International Library of Critical
Writings in Economics 63, Cheltenham: Edward Elgar.
Hammond, P.J. (1988) 'Consequentialist demographic norms and parenting rights',
Social Choice and Welfare 5(2/3): 127-45.
(1991) 'Interpersonal comparisons of utility: why and how they are and
should be made', in J. Elster and J.E. Roemer (eds) Interpersonal Comparisons of
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
293
Well-Being,Cambridge University Press, pp. 200-54.
(1996) 'Consequentialist decision theory and utilitarian ethics', in F. Farina,
F. Hahn and S. Vannucci (eds) Ethics, Rationality, and Economic Behaviour,
Oxford and New York: Clarendon Press, pp. 92-118.
Hanemann, W.M. (1991) 'Willingness to pay and willingness to accept: how much
can they differ?', American Economic Review 81(3): 63547.
Hanemann, W. M. and Kristrom, B. (1995) 'Preference uncertainty, optimal designs
and spikes', in P.-0. Johansson, B. Kristrom and K.-G. Maler (eds) Current Issues
in Environmental Economics, Manchester and New York: Manchester University
Press, pp. 58-77.
Harrod, R.F. (1938) 'Scope and methods of economics', Economic Journal 48:
383-412.
(1948) Towards a Dynamic Economics, London: St. Martins Press.
Harsanyi, J.C. (1955) 'Cardinal welfare, individualistic ethics, and interpersonal
comparisons of utility', Journal of Political Economy 63: 309-21.
(1975) 'Non-linear social welfare functions: do economists have a special
exemption from Bayesian rationality?', Theory and Decision 6: 31 1-32.
(1977) 'Non-linear social welfare functions: a rejoinder to Professor Sen', in
R. Butts and J. Hintikka (eds) Logic, Methodology and Philosophy of Science,
Dordrecht: Reidel, pp. 293-6.
Haslett, D.W. (1990) 'What is utility?', Economics and Philosophy 6: 65-94.
Hausman, D.M. (1994) The Philosophy of Economics -an Anthology, second edition,
Cambridge University Press.
Hausman, D.M. and McPherson, M.S. (1993) 'Taking ethics seriously: economics
and contemporary moral philosophy', Journal of Economic Literature 31(2):
671-731.
(1994) 'Economics, rationality, and ethics', in Hausman (1994), pp. 252-77.
(1996) Economic Analysis and Moral Philosophy, Cambridge University
Press.
Hausman, J.A. (1981) 'Exact consumer's surplus and dead-weight loss', American
Economic Review 7 1: 662-76.
Helliwell, John E and Putnam, R.D. (1995) 'Economic growth and social capital in
Italy', Eastern Economic Journal 21(3): 295-307.
Hicks, J.R. (1939) Value and Capital, Oxford University Press.
(1943) 'The four consumer's su luses', Review of Economic Studies 11:
'P
31-41.
Hirsch, F. (1977) Social Limits to Growth, Cambridge, Mass.: Harvard University
Press.
Hirshleifer, J. (1985) 'The expanding domain of economics', American Economic
Review 75: 53-68.
Hirshleifer, J. and Riley, J.G. (1992) The Analytics of Uncertainty and Information,
Cambridge Surveys of Economic Literature, Cambridge: Cambridge University
Press.
Hodgson, G.M. (1988) Economics and Institutions, Oxford: Polity Press.
Hodgson, G.M., Samuels, W.J. and Tool, M.R. (eds) (1994) The Edward Elgar
Companion to Institutional and Evolutionary Economics, Aldershot: Edward
Elgar.
Howarth, R.B. (1995) 'Sustainability under uncertainty: a deontological approach',
Land Economics 7 l(4): 417-27.
Hubin, D.C. (1994) 'The moral justification of benefitlcost analysis', Economics and
Philosophy 10: 169-94.
Johansson, 0. (1997) 'Optimal Pigouvian taxes with regard to altruism', Land
Economics 73(3): 297-308.
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
294 Articles
Johansson, P.-0. (1992) 'Altruism in cost-benefit analysis', Environmental and
Resource Economics 2: 605-13.
Jones-Lee, M.W. (1992) 'Paternalistic altruism and the value of a statistical life',
Economic Journal 102: 80-90.
Kahneman, D. and Knetsch, J.L. (1992a) 'Valuing public goods: the purchase of moral
satisfaction', Journal of Environmental Economics and Management 22(1):
57-70.
(1992b) 'Contingent valuation and the value of public goods: reply', Journal
of Environmental Economics and Management 22(1): 9 M .
Kahneman, D. and Tversky, A. (1979) 'Prospect theory: an analysis of decision under
risk', Econometrica 47: 263-9 1.
(1984) 'Choices, values and frames', American Psychologist 39: 341-50.
Kahneman, D. and Varey, C. (1991) 'Notes on the psychology of utility', in J. Elster
and J.E. Roemer (eds) Interpersonal Comparisons of Well-Being, Cambridge
University Press, pp. 127-63.
Kaldor, N. (1939) 'Welfare propositions and interpersonal comparisons of utility',
Economic Journal 49: 549-52.
Kaplow, Louis (1995) 'A fundamental objection to tax equity norms: a call for
utilitarianism', National Tax Journal 48: 497-5 14.
Kemp, J. (1997) 'New methods and understanding in economic dynamics? An
introductory guide to chaos and economics', Economic Issues 2(1): 1-26.
Knight, F.H. (1921) Risk, Uncertainty and Profit, New York: Houghton Mifflin.
(1924) 'The limitations of scientific method in economics', in R.G. Tugwell
(ed.) The Trend of Economics, New York: Alfred Knopf, pp. 229-67.
Kocherlakota, N.R. (1996) 'The equity premium: it's still a puzzle', Journal of
Economic Literature 34: 42-71.
Kreps D.M. (1997) 'Intrinsic motivation and extrinsic incentives', American
Economic Review 87(2): 359-64.
Layard, R. (1980) 'Human satisfaction and public policy', Economic Journal 90:
737-50.
Lerner, A.P. (1944) The Economics of Control, London: Macmillan.
Lindbeck, A. (1995) 'Welfare state disincentives with endogenous habits and norms',
Scandinavian Journal of Economics 97(4): 477-94.
(1997) 'Incentives and social norms in household behavior', American
Economic Review 87(2): 370-7.
Little, I.M.D. (1957) A Critique of Welfare Economics, second edition, Oxford
University Press.
Mirrlees, James A. (1971) 'An exploration in the theory of optimum income taxation',
Review of Economic Studies 38: 175-208.
Mitchell, R.C. and Carson, R.T. (1989) Using Surveys to Value Public Goods: The
Contingent Valuation Method, Washington, DC: Resources for the Future.
Myrdal, Gunnar (1969) Objectivity in Social Research, New York: Pantheon.
Nath, S.K. (1969)A Reappraisal of Welfare Economics, London: Routledge & Kegan
Paul.
Ng, Y.-K. (1981a) 'Welfarism: a defense against Sen's attack', Economic Journal 91:
527-30.
(1981b) 'Bentham or Nash? On the acceptable form of social welfare
functions', Economic Record 57: 23 1-50.
(1983) 'Some broader issues of social choice', in P.K. Pattanaik and M.
Selles (eds) Social Choice and Welfare, Amsterdam: North-Holland.
(1986) 'Social criteria for evaluating population change: an alternative to the
Blackorby-Donaldson criteria', Journal of Public Economics 29: 375-81.
(1987) 'Relative-income effects and the appropriate level of public
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations
295
expenditure', Oxford Economic Papers 39(2): 293-300.
(1995) 'Towards welfare biology - evolutionary economics of animal
consciousness and suffering', Biology and Philosophy lO(3): 25-85.
(1997) 'A case for happiness, cardinalism, and interpersonal comparability',
Economic Journal 107: 1848-58.
Nordhaus W.D. (1991) 'To slow or not to slow: the economics of the greenhouse
effect', Economic Journal 101: 920-37.
(1994) Managing the Global Commons: the Economics of Climate Change,
Cambridge, Mass.: MIT Press.
North, D.C. (1990) Institutions, Institutional Change and Economic Pe$ormance,
Cambridge, Mass.: Cambridge University Press.
Norton, B.G. (1987) Why Preserve Natural Variety? Princeton, NJ: Princeton
University Press.
Nozick, R. (1974) Anarchy, State and Utopia, New York: Basic Books.
O'Neill, 0. (1997) 'Environmental values, anthropocentrism and speciesism',
Environmental Values 6(2): 127-42.
Osmani, S. (1993) 'B.M.S. van Praag: the relativity of the welfare concept,
commentary', in M.C. Nussbaum and A. Sen (eds) The Quality of Life, Oxford
University Press, pp. 386-92.
Ostrom, E. (1990) Governing the Commons: The Evolution of Institutions for
Collective Action, Political Economy of Institutions and Decisions Series,
Cambridge University Press.
s s economic performance', Economic Journal 107:
Oswald, A.JT (1997) ' ~ a p ~ i n eand
1815-31.
Persson, M. (1995) 'Why are taxes so high in egalitarian societies?', Scandinavian
Journal of Economics 97: 469-76.
Pollak, R.A. (1978) 'Endogenous tastes in demand and welfare analysis', American
Economic Review 68: 374-9.
Putnam, H. (1993) 'Objectivity and the science-ethics distinction', in M.C.
Nussbaum and A. Sen (eds) The Quality of Life, Oxford University Press,
pp. 143-57.
Putnam, R.D., Leonardi, R. and Nanetti, R.Y. (1993) Making Democracy Work: Civic
Traditions in Modern Italy, Princeton, NJ: Princeton University Press.
Rarnsey, F. (1928) 'Amathematical theory of savings', Economic Journal 138: 543-9.
Randall, A. and Stoll, J.R. (1980) 'Cdisumer's surplus in commodity space',
American Economic Review 71: 449-57.
Rappaport, S. (1996) 'Abstraction and unrealistic assumptions in economics',
Journal of Economic Methodology 3: 215-36.
Rawls, J. (1973) A Theory of Justice, Oxford University Press.
Robbins, L. (1932) An Essay on the Nature and Significance of Economic Science,
London: Macmillan.
Robinson, J. (1962) Economic Philosophy, London: Watts.
Romer, P.M. (1999) Changing Tastes. How Evolution and Experience Shape
Economic Behaviour, Oscar Morgenstern Memorial Lectures, Cambridge
University Press, forthcoming.
Rosenbaum, E.E (1995) 'Interpersonal comparison of utility: positive, normative or
value laden?', Journal of Economic Mehodology 2(2): 239-57.
(1996) 'Value-ladenness in economics: a rejoinder to Davis', Journal of
Economic Mehodology 3(1): 127-9.
Samuelson, P.A. (1938) 'A note on the pure theory of consumer's behaviour, public
expenditure', Economics 5 : 61-7 1.
(1950) 'Evaluation of real national income', Oxford Economic Papers NS2:
1-29.
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
296 Articles
(1954) 'The pure theory of public expenditure', Review of Economic Studies
36: 387-9.
Samueison, W. and Zeckhauser, R. (1988) 'Status quo bias in decision making',
Journal of Risk and Uncertainty l(1): 7-59.
Scheffler, S. (1982) The Rejection of Consequentialism: a Philosophical Investigation
of the Considerations Underlying Rival Moral Conceptions, Oxford University
Press.
Scitovsky, T. (1941) 'A note on welfare propositions in economics', Review of
Economic Studies 9, pp. 77-88.
(1974) 'Are men rational or economists wrong?', in Paul A. David and
Melvin Reder (eds) Nations and Households in Economic Growth: Essays in
Honor ofMoses Abramovitz, Palo Alto, Calif.: Stanford University Press: pp. 22335.
(1992) The Joyless Economy, Oxford University Press.
Sen, A.K. (1973) On Economic Inequality, Oxford: Clarendon Press.
(1976) 'Welfare inequalities and Rawlsian axiomatics', Theory and Decision
7: 243-62.
(1977a) 'Non-linear social welfare functions: a reply to Professor Harsanyi',
in R. Butts and J. Hintikka (eds) Logic, Methodology and Philosophy of Science,
Dordrecht: Reidel, pp. 297-302.
(1977b) 'Rational fools: a critique of the behavioural assumptions of
economic theory', Philosophy and Public Affairs 6: 317-44.
(1979) 'Personal utilities and public judgements: or what's wrong with
welfare economics?', Economic Journal 89: 537-58.
(1981) 'A reply to "Welfarism: a defence against Sen's attack"', Economic
Journal 91: 53 1-5.
(1985) 'Goals, commitment, and identity', Journal of Law, Economics and
Organization l(2): 341-55.
(1986) 'Social choice theory', in K.J. Arrow and M.D. Intrilligator (eds)
Handbook of Mathematical Economics, 111, ch. 22, Amsterdam: North-Holland,
pp. 1073-1181.
(1987) On Ethics and Economics, Oxford and New York: Blackwell.
(1991) 'Utility - ideas and terminology', Economics and Philosophy 7:
277-83.
(1992) Inequality Reexamined, Cambridge, Mass.: Harvard University
Press.
(1993) 'Capability and well-being', in M.C. Nussbaum and A. Sen (eds) The
Quality of Life, Oxford University Press.
Simon, H. (1993) 'Altruism and economics', American Economic Review 83(2):
156-61.
Singer, P. (1975)AnimalLiberation: A New Ethics for Our Treatment ofAnimals, New
York: Avon.
(1979) Practical Ethics, Cambridge University Press.
(1980) 'Animals and the value of life', in T. Regan (ed.) Matters of Life and
Death: New Introductory Essays in Moral Philosophy, New York: Random House.
Slovic, P. and Lichtenstein, S. (1983) 'Preference reversals: a broader perspective',
American Economic Review 73(4): 596-605.
Smith, V.K. (1992) 'Arbitrary values, good causes, and premature verdicts:
comment', Journal of Environmental Economics and Management 22(1): 7 1-89.
Solow, R.M. (1994) [I9711 'Science and ideology in economics', The Public Interest
(23): 94-107. Reprinted as chapter 13 in Hausman (1994), pp. 239-59.
Spash, C.L. (1993) 'Economics, ethics, and long-term environmental damages',
Environmental Ethics 15: 117-32.
Downloaded By: [Göteborg University Library] At: 09:33 11 December 2007
Ethics and economic policy recommendations 297
Spash, C.L. and Hanley, N. (1995) 'Preferences, information and biodiversity
preservation', Ecological Economics 12: 195-208.
Stark, 0. (1995) Altruism and Beyond, Cambridge University Press.
Stigler, G.J. and Becker, G.S. (1977) 'De gustibus non est disputandum', American
Economic Review 67: 76-90.
Stiglitz, J.E. (1985) 'Pareto efficient and optimal taxation and the new new welfare
economics', chapter 15 in A.J. Auerbach and M. Feldstein (eds) Handbook of
Public Economics 1, Amsterdam: North-Holland.
Sugden R. (1982) 'On the economics of philanthropy', Economic Journal 92:
341-50.
(1984) 'Reciprocity: the supply of public goods through voluntary
contributions', Economic Journal 94: 772-87.
(1986) The Economics of Rights, Cooperation and Welfare, Oxford: Basil
Blackwell.
(1989) 'Spontaneous order', Journal of Economic Perspectives 3(4): 85-97.
Thaler, R.H. (1980) 'Toward a positive theory of consumer choice', Journal of
Economic Behavior and Organization 1: 39-60.
(1992) The Winner's Curse: Paradoxes and Anomalies of Economic Life,
Princeton, NJ: Princeton University Press.
Thurow, L. (1996) The Future of Capitalism, London: Nicholas Brealey.
Tuomala, M. (1990) Optimal Income Tax and Redistribution, Oxford: Clarendon
Press.
Tversky, A. and Griffin, D. (1991) 'Endowment and contrast in judgements of wellbeing', in R. Zeckhauser (ed.) Strategy and Choice, Cambridge, Mass.: MIT Press.
Tversky, A. and Kahneman, D. (1974) 'Judgements under uncertainty: heuristics and
biases', Science 185: 1124-3 1.
Tversky, A,, Slovic, P. and Kahneman, D. (1990) 'The causes of preference reversal',
American Economic Review 80(1): 204-17.
Varian, H.R. (1984) Microeconomic Analysis, New York: W.W. Norton.
Veblen, T. (1899) The Theory of the Leisure Class, New York: Macmillan.
Vickers, D. (1997) Economics and Ethics: An Introduction to Theory, Institutions, and
Policy, Westport, CO and London: Praeger.
Viscusi, W.K. (1992) Fatal Tradeoffs, Public and Private Responsibilities for Risk,
Oxford University Press.
Weston, S.C. (1994) 'Toward a better understanding of the positive/nomative
distinction in economics', Economics and Philosophy 10: 1-17.
Weymark, J.A. (1991) 'A reconsideration of the Harsanyi-Sen debate on
utilitarianism', in J. Elster and J.E. Roemer (eds) Interpersonal Comparisons of
Well-Being, Cambridge University Press, pp. 255-320.
Willig, R. (1976) 'Consumer's surplus without apology',American Economic Review
66: 589-97.
Zamagni, Stefano (ed.) (1995) The Economics of Altruism, Cheltenham: Edward
Elgar.