Does Can Imply Ought?
Singer on Intuition, Responsibility,
and World Poverty
Joseph Barillari1
14 May 2002
In Rich and Poor[1], ethicist Peter Singer outlines
an approach to world poverty that renders individuals responsible to
give in proportion to their ability to help. Arguing from intuition,
he advocates an overhaul of current moral attitudes toward economic
destitution, and obligates the advantaged to contribute to the
disadvantaged to the extent that they can do so without giving up
anything of ``comparable moral significance.''
In this paper, I argue that Singer's analysis cannot be used to
substantiate such blanket obligations, and that the
thought-experiments Singer uses to buttress his claims are
unconvincing when applied to problems of this scale.
I will begin by outlining a few practical objections to any universal
obligation to donate. I then turn to the philosophical
arguments. After finding Singer's use of intuition to be unpersuasive,
I consider his tacit admission that killing and letting die are
sufficiently different as to be intrinsically distinguishable, and
close with analysis of intuitive objections to Singer's obligation.
Admittedly, the practical is the domain of economists and
policymakers, but, as Singer prefaces his own remarks with empirical
observations, I cannot help but articulate a few of my own.
Singer counters the objection raised by his critic Susan Wolf (that
the obligation to donate will destroy what we view to be the good
life) by claiming first that no right to the ``good life'' can be
recognized in a world that is not a ``world of plenty,'' and second
that we need not sacrifice our human relationships to end poverty, as
they constitute a good in themselves, and are therefore protected by
the criterion of ``comparable moral significance.'' While the latter
claim is ultimately dependent on one's calculus of
significance,2 the former is
based on an incorrect factual assertion that, were it even true, would
be actually aggravated by Singer's advocacy of massive donation.
That is to say, Singer's assertion that ours is not a ``world of
plenty'' directly contradicts his assertion (twelve pages earlier)
that redirecting grain use from raising livestock would make short
work of hunger. We already live in a world of plenty, albeit
one with poor distribution. Distribution is impaired by bad local
attitudes and worse local governments - and unless Oxfam is permitted
to raise its own army, our donations are unlikely to shake
either. Singer mentions that we are under no obligation to donate if
we think our aid will have no long-term value; the causes of poverty
make this the case virtually everywhere: countries are never hungry by
accident. A Niagara of grain shipped into Africa will do nothing to
topple a tyrant who keeps it from his people to starve them into
submission.
Some services -- public safety, social welfare, and education, for
instance -- are supported by collective taxation, not individual
initiative because if they had to rely on individuals, they would be
far less effective. No army supported by donations can be as
effective as one supported by taxes; likewise, effective solutions to
poverty are best implemented by governmental actors, because they can
supplement the purse with the sword. Famine and disease are typically
symptoms of despotic government. Consequently, lasting change is
impossible without military intervention. Our poverty-ending efforts,
therefore, ought not to go into soliciting donations, but into
lobbying our Congressmen.
The feasibility of Singer's thesis hinges on human malleability. Can
people be motivated to work without rewards? Cash can always be
exchanged for virtue: a man unmotivated by money can donate it to
charity. But Singer's plan would leave men with neither cash
nor virtue as compensation: there is no special virtue in
not committing crimes against humanity (which is what visiting
a theater will become, should Singer's thesis become widely
accepted). What has been the mark of a highly moral man will become
obligatory, and traditional forms of internal and external motivation
will be effectively eliminated.
The notion that our world will ever become a world of plenty
through donation-to-the-point-of-comparable-loss is extraordinarily
shortsighted. Plenty follows growth and growth follows spending. A
moratorium on non-essential spending will drive the world faster and
farther from plenty than the most misguided economic planning
imaginable. Even if human beings could be motivated to work ever
harder without compensation, the result of directing an entire
economy's non-essential consumer spending into foreign aid will be an
unmitigated disaster which will forever foreclose the opportunity to
create a world of plenty. Entire sectors of industry will collapse,
prompting staggering unemployment. The notion that spending on luxury
goods can be so easily diverted overseas is premised in a naïve
and atomistic view of economics.3
The expectation4that new sectors will spring up around the business of helping the
impoverished is equally unrealistic: as Singer admits, there is no
difficulty in producing enough food to feed the world, so no
agribusiness boom is likely to answer the luxury bust. Indeed, the
loss of the profitable meat industry alone will more than dwarf the
increased demand for grain for the third world. Nor can we expect to
make short work of poverty and quickly return to luxurious normalcy -
the labor necessary to rebuild impoverished countries will be long and
(for the foreseeable future) economically unrewarding. It will take
decades to build sufficient infrastructure to make these countries
self-sustaining.
While dire forecasts are subject to suspicion, if only because they
are premised in counterfactual claims (in this case, the widespread
adoption of the Singer ethic), it is unreasonable to believe that the
fallout from the collapse of non-essential industries is so trivial as
to be dismissable as a short-term transition cost.
The arguments above are are not objections to Singer's obligation
because they claim that widespread donation would fail to do any good,
or that we would sacrifice something of comparable moral significance
(our own sustenance, for instance) in being compelled to
donate. Let us suppose that triage or inaction would not be better
solutions, that government-led intervention is ineffective, and that
by `saving' a child we do not mean `staving off starvation and disease
for six months, only to have the child killed in a civil war or by a
new famine,' but something more permanent. Let us suppose that each
and every one of the practical objections above is ungrounded. Even if
they are (which is unlikely; they are moderate and supported by basic
economic reasoning), I argue that Singer does not meet his burden of
proving a moral obligation for giving rooted in the status of having.
Singer assigns responsibility on the basis of one's ability to help
without ``sacrificing anything of comparable moral significance.'' He
teases the notion of culpability for inaction from our intuitions
using the famous well-dressed man/drowning child thought experiment:
it would be unfathomably callous for the man to fail to rescue the
child, given that he could easily do so. To ignore the child is to
bring about its death. This is the kernel of his argument about
poverty: to choose luxuries over donation is to choose death for
children in absolute poverty. Singer claims that our responsibility
for the deaths of these children is intrinsically (though obviously
not extrinsically) no different from our responsibility if we had
murdered them.
No longer can those of us who do not work for de Beers safely say that
we have done nothing to cause the deaths of African children. Is this
what we intended when we intuitively levied moral sanctions against
the man who would not ruin his shoes to save a drowning child? I
argue that it is not, and that Singer has tried to magnify our
intuitions to a scale at which they are useless.
Singer employs some rhetorical sleight-of-hand to establish
responsibility: he showcases a simplistic single-actor example (the
drowning child) to substantiate his views on the remote, complex,
multi-actor problem of world poverty.
This thought-experiment draws upon intuition that does not scale to
situations with more than one actor. Consider a different spin on the
drowning-child scenario: what if the victim-to-be is an inebriated
Princeton student who passed out in the fountain outside Robertson
Hall on Thesis Day, with dozens of his classmates cavorting in the
water alongside him? If they all neglect to drag him to safety
(they're far too busy gamboling), is each one as responsible for his
death as if he were the only one present? Do we feel quite the same
about them as we did about the lone well-dressed man?
Should they all be considered murderers? Obviously not. Intuitively,
the proliferation of potential actors reduces the moral culpability of
each individual actor. A former classmate of mine has referred to this
exculpation as the ``multiple agency'' exception. The addition of
potential actors to these thought-experiments changes the dynamics
sufficiently to render the single-actor-single-victim analogy useless.
Intuition says that each may be responsible, but not to the same
extent as if each were alone. If an individual is less culpable
because his fellow bystanders didn't do anything either, then (by
extension) each Westerner is several orders of magnitude less culpable
for third-world death-by-starvation by virtue of being in good
company. Alternatively, if the presence of others does not make
a difference, then we can say that the situation does not change when
we add actors, and by extension, that those actors would make no
difference if they did not exist. This should exculpate them
(for the first man therefore bears the guilt), except that Singer says
that no salvation can come from counterfactual positions. In this
case, Singer has to claim that individually, each and every one of
those bystanders is as guilty (recall that we're supposing that guilt
is unaffected by the number of actors) as the man who couldn't be
bothered to save the drowning child (and thus aggregate guilt
increases without bound as we add more bystanders). Now intuition
starts to work against him: are we really as culpable for the drowning
child as we are for the African millions? Put another way, are we all
guilty of dozens of murders every minute?
Intuition would say no.
Let us test our intuition a bit further with another
thought-experiment: what if the notorious Man with the Bugatti had an
outing in a trainyard, with dozens of friends (whose savings were
similarly invested in their respective cars), each standing by a
switch?5 When the speeding trolley came along, if any
individual threw his switch, he would destroy his own car and save the
child. If none of them did so, the child would die - but would we
consider them all murderers? Does failure to make the sacrifice
necessary to save the child make one a murderer, if any one of a group
of people might have shouldered it instead?
Intuition says no. Singer's teasing-out of intuitions does not scale to
these larger problems.
Crucial to all of these scenarios is the element of acute crisis and
immediacy. World hunger, by contrast, is a chronic problem, insolvable
by a single sacrifice. The intuition that Singer uses to lead us to
the sacrifice-everything-of-lesser-moral-significance thesis, even if
it scales up to multi-actor problems (and I have argued that it does
not), is not applicable long-term problems.
Consider a new spin on the thought-experiment above. Suppose that the
child above was not stuck on a track, but in a small cottage, with a
large sign on the front lawn, informing the passer-by that the child
would die of a chronic illness if his family did not receive a
specified sum every month for his treatment. Each of the men drives by
and recognizes that sum to be equivalent to the market value of his
car, plus enough of his pension to cut his spending down to the bare
minimum needed to survive.
None of the car-owners could miss the sign. If each of them merely
shakes his head with pity and drives past, do we make them all
murderers? If so, should we stay in our houses to avoid seeing such
things and becoming culpable for them? Let us take it even closer to
reality: suppose that the sign admits that there is some small (but
nonzero) chance that governmental funding may be secured to save the
child, but there's no way for the car-owner to know that it's been
secured until it's too late to save the child.
At this point, intuition has either failed us completely (in which
case, refer to the argumentation in the next few sections), or led us
to the conclusion that the mere ability-to-assist does not create an
obligation to assist, for the alternative would obligate us to deposit
our savings at every legitimate donation-place, collecting-tin, or
charitable organization we happened to see. That is exactly Singer's
argument, and I do not believe it is possible to substantiate it with
mere intuition.
Obviously, the notion that we have some duty to the poor may be
widespread, but few people act upon it - whereas most people would
act upon the intuitive obligation to save the drowning child. Why the
difference? It seems to indicate either that our intuitions, when
scaled, no longer motivate us into direct action (a problem in human
psychology and motivation), or perhaps that there is some difference
- namely, the scaling factors6 - between the situations (drowning
child and world poverty) that makes intuition for one inapplicable to
the other.
We can't accept the sacrifice-until-the-loss-of-something-comparable
obligation simply because we're browbeaten into support for it by the
intuitive immorality of allowing the drowning child to die. Just as
our geometric intuition is of no use to us in higher-dimensional
spaces,7 our small-scale, single-actor moral intuition is no substitute
for careful analysis in situations with multiple potential actors and
victims, analysis with which Singer does not provide us.
Singer writes that a reasonable policy should expect people to donate
only ten percent of their income, with some flexibility based on
ability to give. In setting a guideline for ``minimum morality,'' much
less admitting that this number is arbitrary, he destroys the
legitimacy of the claim that the acts/omissions distinction is
meaningless. If there were truly no intrinsic difference between
buying a stereo set and shooting children, then both ought to be
intolerable, and a ten-percent compromise regarded as an insult to
humanity. By contrast, if indeed we must fulfill only a portion of
our moral obligation to stop people from dying, then we should
recognize the same degree of leeway in cases of active murder:
perhaps, one must only abstain from murder in ninety percent of cases,
for morality in the remaining ten percent might be more than we should
expect from human beings.
Even if we admit that Singer gave the ten-percent figure as a
public-policy suggestion, such a suggestion would be absurd if we
recognized no relevant intrinsic distinctions between passive and
active murder. The only way to avoid the absurdity is to recognize
that the acts/omissions distinction cannot be so easily brushed aside.
By admitting that we can set an approximate donation amount by which
one can be regarded as being ethically average, Singer tacitly
concedes this point. If letting-die were truly equivalent to murder,
it would be intolerable to permit people to contribute only a fraction
of their potential donation.
While my presentation of objections to erasing the killing/letting die
distinction is not meant to be exhaustive, a second,
better-known8 objection underscores the
difficulty of erasing it: some argue that culpability cannot be
equivalent because those who ``kill'' by omission generally lack
murderous intent. Singer answers this objection by considering
speeding motorists, who are prosecuted when they strike pedestrians
despite a lack of intent to kill. I posit that this example does not
map cleanly onto failure to donate, where the killer and victim are
not so readily identifiable, and blame not so easily assigned. Without
the usual direct interaction between the parties, assigning blame for
large-scale inaction is non-trivial.
Singer claims that ``if a consequence of my spending money on a luxury
item is that someone dies, I am responsible for that death.'' We want
to avoid having that death on our conscience, so hypothesize that
instead of buying a stereo, we send our money to an NGO who uses it to
save five people. If the price of a stereo set was the maximum we
could sacrifice, we have acted morally.
However, if there was some chance we could have donated more --
perhaps by cutting back on luxuries, perhaps by earning more, we have
(according to Singer) murdered those whom we could have but did not
save. Until an economist sets a cap on human productivity, one can
always theoretically earn more. This means that one can choose
no employment other than that which is most likely to end hunger,
typically the most lucrative. Some may try to escape this conclusion
by claiming that a man is not culpable for worldwide deaths-by-hunger
if he has done all that he can to prevent them.10 Singer implicitly casts this
objection aside in a footnote: he claims that one is obligated to
retain a high-paying job, even if it requires spending a portion of
one's income on the trappings of the position (clothes, a car, etc.),
if a lower-paying job would leave less net profit left over to donate.
Consequently, we ought to direct our lifestyles to maximize our
earnings, so a man who fails to work hard enough to earn a salary
increase is guilty for the deaths that the money he could have earned
would have prevented. Thus doing ``all that you can'' means absolutely
that. No man is permitted to be a painter if he could earn (and
donate) more as a graphic designer, no man may become a scholar if he
would save more lives as a corporate lawyer.
This destruction of preferences is inevitable. The ``comparable moral
significance'' exception is of no help if we want to write poetry
instead of advertising copy. Are the the dozens of lives lost
comparable to the freedom to write a few books of verse?
While the short-term sacrifices that Singer explicates are intuitively
obligatory, intuition may not stretch far enough to compel us to
reject all lifestyle choices except those that are most likely to save
children in poverty. Nor does it seem intuitive that we are morally
required to ignore virtually all of our personal preferences. Is this
what we intended when we condemned the well-dressed man?
Singer's obligation to give fails in several respects: first, it is
impractical. Second, it is primarily substantiated by unscalable
thought-experiments and misapplied intuition. Third, he admits that
our policies regarding letting die can be lax in relation to those
concerning killing, which is a tacit admission that the erasure of the
intrinsic distinctions between the two -- a crucial component of
Singer's argument -- is impossible: if we truly regarded murder and
failure to save as equivalent, a policy that required only donating
ten percent of one's income would be an insult to humanity.
Finally, even if we accept the
give-until-something-of-comparable-moral-significance-is-sacrificed
obligation as having withstood these criticisms, we realize that that
no human being is capable of acting ethically within it without
sacrificing not only money, but virtually any hope for
fulfillment. Foreclosed are all opportunities except those expected to
save the most children, effectively destroying individual agency - a
highly counterintuitive stance. With those objections, I find Singer's
obligation to give to be unworkable.
- 1
-
Peter Singer.
Rich and poor.
Hinman Collection.
This paper contains approximately 3387 words (not counting
title, author, date, this paragraph, or the References
section). This paper represents my own work in accordance with
University regulations.
Does Can Imply Ought?
Singer on Intuition, Responsibility,
and World Poverty
This document was generated using the
LaTeX2HTML translator Version 2K.1beta (1.55)
Copyright © 1993, 1994, 1995, 1996,
Nikos Drakos,
Computer Based Learning Unit, University of Leeds.
Copyright © 1997, 1998, 1999,
Ross Moore,
Mathematics Department, Macquarie University, Sydney.
The command line arguments were:
latex2html -split 0 -nonavigation poverty.tex
The translation was initiated by Joseph Barillari on 2002-05-27
Footnotes
- ... Barillari1
- Florian Becker
provided advice on the construction of this paper. Anne Caswell
Klein of the Writing Center provided advice on the revision of
this paper.
- ...
significance,2
- I posit that by Singer's reasoning it would be
very difficult to justify assembling the resources necessary
to beget and raise one's own children, as few things are as
significant as the certain deaths of poor children.
- ... economics.3
- Ironically, Singer
criticizes Locke and Nozick for advocating a naïve and
atomistic (but nevertheless widely accepted) view of human rights.
- ... expectation4
- Florian Becker mentioned this objection.
- ...
switch?5
- Florian Becker discussed similar examples in a
meeting on May 6.
- ... factors6
- The number of actors and the
duration of the obligation.
- ...
spaces,7
- Professor R. Sedgewick, CS226 lecture notes, Spring
2002
- ...
better-known8
- Singer discusses it.
- ... responsibility9
- Florian
Becker suggested improvements to this section.
- ... them.10
- I
discussed this with a classmate.
Joseph Barillari
2002-05-27