On Reasonable Avoidability and Lifestyle Diseases

On Reasonable Avoidability and Lifestyle Diseases1
Martin Marchman Andersen, CESEM, Copenhagen University
Work in progress: please do not quote!
Introduction
Epidemiological research is pointing to a number of diseases being associated with particular lifestyle
characteristics, such as smoking, drinking, lack of exercise and over-eating. Within egalitarian and especially
luck egalitarian theory it is therefore obvious to raise questions on whether the individual person with such
lifestyle characteristics, is responsible for the statistically higher risk of getting such diseases and therefore
e.g. should be held cost-responsible for the hospital-expenses they bring about.
In his book Health, Luck and Justice, Shlomi Segall argues for a luck egalitarian approach to distributive
justice in health and health care. More precisely, he suggests that what is unjust is “… for individuals to be
worse off than others due to outcomes that it would have been unreasonable to expect them to avoid”2
(henceforth: reasonable avoidability, RA). He notes that some people have less healthy lifestyles than
others, and by not taking good care of their health, he argues, people with risky lifestyles place unnecessary
burdens on the health care system.3 He maintains therefore that we can reasonably expect individuals to
exercise, not to smoke, drink, overeat, and the like, and justice thus requires individuals with risky
behaviors to bear the related costs. In order not to abandon the imprudent, however, he suggests a model
of meeting basic needs to support the luck egalitarian distribution of health and health care. Through a
model of ex ante taxation, however, alcohol, tobacco, fatty food etc. should be taxed proportionally to
covering the expenses related to the diseases they bring about (hospital expenses etc.).
So how are we to understand this?
Well, more precisely, Segall suggests that we should “… replace “responsibility” with a more plausible
understanding of what constitutes a case of brute luck.” Rather, he maintains, we should understand
“brute luck” as an “…outcome of actions (including omissions) that it would have been unreasonable to
expect the agent to avoid (or not to avoid in case of omissions).”He stresses that “expectation” is to be
1
For valuable comments, recommendations and discussions related to this article or earlier versions of it I would very much like to thank Dan
Wikler, Nir Eyal, Norm Daniels, Alex Gosseries, Xavier Landes, Claus Hansen, Morten Ebbe Juul Nielsen, Kasper Lippert-Rasmussen, Søren Flinch
Midtgaard and not least Shlomi Segall and Nils Holtug
2
Segall, S., 2010, Health, Luck and Justice, Princeton University Press, p. 13
3
Op.cit. p. 1
1
understood as a normative expectation rather than an epistemic one: “We are not inquiring here into what
people are likely to do, but rather what it is that society can reasonably expect of them.”4
As Segall thus suggests replacing “responsibility” with RA I take it to be a misunderstanding to read RA as if
it was another concept of responsibility. Rather, as I understand it, we should replace “responsibility” with
RA as the basis of distribution within luck egalitarianism.
But which outcomes of action, then, would it be unreasonable to expect individuals to avoid?
Unfortunately, Segall is not very informative here. Nonetheless his suggestion calls upon at least two
questions: 1) if we replace responsibility with RA, what principles, then, should guide us in determining
when an act is reasonably avoidable?5 And 2) is RA actually a more plausible basis of distribution than
responsibility?
My aim in this article is therefore twofold and divided in two parts answering these questions respectively.
Part 1: Firstly, I will examine Segall’s actual use of RA. He does not provide us with a general principle to
determine when an act or omission is reasonably avoidable6, but only examples. However, through merely
a process of elimination I will derive two principles of his use.
His use of RA implies (and he claims explicitly) that it is reasonable to expect individuals to avoid smoking
(or to become obese). This is in spite of a non-paternalistic premise and even if smoking (and obesity)
generally speaking is not a financial burden on the health care system. My second step will be to object to
that.
Thirdly, I will further specify Segall’s use of RA and show that one of the principles of his use of RA leads to
counter intuitive results. I will make it probable that Segall either will have to give up his claim that we can
reasonably expect individuals to avoid smoking (and obesity), or to add a paternalistic motivation.
Fourthly, I will show that when a view (like Segall’s), on when inequality in health is unjust, is made
sensitive to individuals’ different efforts in looking after their health, it will lead to contra intuitive results
regarding the distribution of research funds.
4
Op.cit., p. 20
5
Peter Vallentyne also operates with the concept of Reasonable Avoidability. He explains what makes a choice reasonable: “One view is that it is in
the agent’s best interest.” See Vallentyne, P., “Brute Luck, Option Luck, and Equality of Initial Opportunities”, Ethics 112 (2002): 529-57, p. 533. As I
will get back to there is no reason, however, to believe Segall intents the same explanation.
6
This might be a conscious move. Segall acknowledges that RA may lead to indeterminacy. See op.cit., p. 22 and page 16 in the present article.
2
In part 1 I ignore the matter of responsibility and focus only on RA (due to a strict reading of the
replacement of responsibility with RA).
Part 2: I will first try to show that claims on what we reasonably can expect individuals to avoid are
sensitive to our view on responsibility, and that we, Segalls’ intention or not, do not avoid answering the
question of how to understand responsibility. Nonetheless I will secondly support Segall’s claim that we
ought to replace responsibility as the basis of distribution with RA (or something similar), insofar as we
must operate with a “responsibility”-sensitive approach to justice in health and health care.
Finally I will sum up and comment on my findings in part 1 and 2 as seen in its entirety.
Decoding Segall’s Use of RA
Segall provides us with a number of examples to show why RA as a replacement of “responsibility” is
attractive. Here are two rather obvious ones:
1) Women are generally said to be responsible for getting pregnant. If responsibility (in this more
direct sense) is to be the basis for distribution, it would mean that we should deny women medical
treatment for pregnancy and delivery. RA, in contrast to responsibility, gives us justice-based
reasons to provide treatment to pregnant women, as ”… it would be unreasonable to ask women
(or couples for that matter), as a rule, not to get pregnant.”7
And:
2) “Firefighters are responsible for incurring burns because they have chosen, by their own free will,
to enter a burning building.” Segall maintains that even though we might have non-justice based
reason to care for injured firefighters, we might also have justice-based reasons for doing so, since:
“… it would be unreasonable to expect everyone in society not to become a firefighter, as there is
clearly still a need for that line of work.”8
Even though these points seem to have a strong intuitive appeal, Segall, unfortunately, is not very
informative in explaining more precisely what should guide us in determining when some action or
omission is reasonably avoidable.
7
Op.cit., p. 21
8
Op.cit., p. 22
3
One suggestion can be derived from example 1 and 2: it is unreasonable to expect A to avoid X, if it is
unreasonable to expect everyone to avoid X.9 Naturally, however, this requirement, though reasonable, is
not sufficient to guide us in all matters where we wish to claim some act or omission to be or not to be
reasonably avoidable. Consider another example provided by Segall:
3) “People residing in certain parts of California are (often) responsible for choosing to live on a
geographical fault line. However, it would be unreasonable to expect them not to settle in these
areas owing to the slight chance of being hit by an earthquake.”10
This, Segall continues, is in contrast to the matter of smoking:
4) “… a geographical choice of residence is different from the decision to smoke, for example (as the
latter does involve what most people would consider a legitimate trade-off between prudence and
pleasure)”11
Here, obviously, it does not seem similarly plausible to claim that it would be unreasonable to expect
everyone not to settle in these certain parts of California. The idea in example 1 and 2 is clearly that society,
in general, would be left with a huge problem if no woman had children, or if no one would take the job as
a firefighter. If no one settled in these risky areas of California, society would not similarly have a problem.
Therefore a rather specific meaning lies in the claim that it is unreasonable to expect everyone to avoid X.
Stating that smoking is a trade-off between prudence and pleasure, where (implicitly) a geographical choice
is not, might at first appear to be another potential guidance in determining when some act or omission is
reasonably avoidable. But such division does not seem to be there: if smoking is a trade-off between
prudence and pleasure, in which relevant sense is a geographical choice not a trade-off between prudence
and pleasure? Given even the slight chance of being hit by an earthquake, it does involve an element of
prudence to reside in certain parts of California, which then might be traded in for the pleasure of residing
there. Our intuition might agree that there is a relevant difference between a geographical choice and the
choice of smoking, but the difference, it seems, can hardly be that the latter is a trade-off while the former
is not. We will therefore be in need of further specification.
So if we share the intuition that there is a relevant difference, what might it be? We might suggest that
there somehow is a difference in how much value a Californian will have to give up by leaving her home
9
What it is that it is unreasonable for everyone to avoid can be a variable. I will get back to this in part 2.
10
Op.cit., p. 20
11
Op.cit., p. 21
4
compared to how much a smoker will have to give up by quitting the tobacco. But whether that would be
due to a subjective or an objective parameter, the suggestion seems to be in conflict with another of
Segall’s examples. Consider the following:
5) “… suppose that a woman is warned by her physician of a high risk to her health should she choose
to get pregnant. If she does opt for the pregnancy, then it would not be unreasonable, in my view,
to see her as forfeiting a claim to treatment on grounds of egalitarian distributive justice.”12
So to suggest that it somehow matters, in Segall’s use of RA, how much value the individual is giving up by
avoiding X, does therefore not seem very obvious: I guess it would be unusually controversial to suggest the
value of pregnancy to be relevantly lower than the value of residing where your home is.
There might of course be a hidden variable: it could be the case that the size of the value an individual is
giving up does matter, but that Segall in example 5 finds it irrelevant because the risk is too high. But
Segall’s general use of RA does not clearly support this possibility, and moreover he stresses that RA is not
paternalism:
“In that sense imposing the costs of treatment on the imprudent is not a paternalistic policy. Rather,
according to my proposal those costs are passed on to the imprudent not because they have failed to take
good care of themselves as such, but rather because they have avoidably burdened the public health care
system.”13
He can therefore not justify any particular claim on what is reasonably avoidable by referring to the
improvement of the welfare of the individuals at stake.14 E.g. he cannot justify smoking as being reasonably
avoidable by claiming smokers to be better off not smoking. As this requirement of non-paternalism does
not seem to contradict any of his particular claims on what we reasonably can expect individuals to avoid, I
take it that the size of value an individual is giving up by avoiding X does not matter in Segall’s use of RA.
One last suggestion of how we are to determine when some action or omission is reasonably avoidable is to
evaluate and compare the different risks at stake. It is conceivable that Segall finds it reasonably avoidable
to smoke, but not to leave certain part of California, because the chance of being hit by an earthquake is
only slight, while the chance of falling sick due to smoking is comparatively higher. As this is further
12
Op.cit. p. 22
13Op.cit.,
14
p. 78. For this reason in particular there is no reason to believe Segall is sharing Vallentyne’s idea of reasonable avoidability.
For conditions on paternalism, see http://plato.stanford.edu
5
supported by the fact that he in example 5 is operating with a high risk to health in case of pregnancy, I
believe we have good reasons to accept it.
Again, Segall’s use of RA is not perfectly clear. But to sum up, I have found two principles for determining
whether an act or omission is reasonably avoidable: i) it is unreasonable to expect A to avoid X, if it is
unreasonable to expect everyone to avoid X, and ii) the size of the risk attached to X somehow affects
whether an act or omission is reasonably avoidable, in the sense that the higher the risk attached to X, the
more we should be inclined to find X reasonably avoidable. So far we may say that it is unreasonable to
expect A to avoid X, if the risk attached to X is lower than Y.
With this in mind I will now provide a first objection before I will further specify Segall’s use of RA.
First Objection
Studies show that the ultimo lifetime health care costs are higher for ‘healthy’ people than for smokers
(and for obese).15 So why then would we pass on the costs related to smoking to the smokers? How can we
say smokers have avoidably burdened the public health care system, if they de facto are not a burden?
In fact Segall considers this objection:
“Suppose, however, that it does turn out that smokers save the health care system money, and that those
pursuing healthier lifestyles are the ones who burden the system. Would the implications be embarrassing
for the luck egalitarian health care system?”
Segall does not think so:
“The justification for passing on the costs of treatment to the imprudent was based on the premise that
they unnecessarily burden the health care system.”16
So we might say that whenever someone enters a hospital with a need for treatment, then the system is
burdened. And burdens stemming from behaviors that we reasonably can expect individuals to avoid are
unnecessary burdens. As such the price is irrelevant. If your need for treatment stems from behavior that it
would be unreasonable to expect you to avoid, then it is a welcome burden.
15
Se e.g. van Baal PHM, et al., 2008, Lifetime medical costs of obesity: prevention no cure for increasing health expenditure. Plos Med 5(2): e 29.
Doi:10.1371/journal. Pmed.0050029. Note also that this is a study in lifetime medical costs. The findings might therefore not apply to e.g. the
American health care system as there can be interruptions in the matter of who risks pools with whom. This might occur when individuals shift from
employer financed health insurances to Medicare. However, as such system is not a system of universal health care it is not relevant for the present
discussion.
16
Segall, ibid., p. 83
6
Segall can also argue that this is not paternalism. Rather, he can say, it is unjust for the rest of the society to
pay for burdens that stem from imprudent behavior. As he asks elsewhere: “Why should the rest of us pay
for their avoidable reckless behavior?”17
But is that really convincing? Is it plausible to tax smokers even though smoking generally leads to savings?
Not for the best of those who smoke (given the premise of non-paternalism), but because it is somehow
unfair to let others pay for particular instances of imprudent behavior?18
As smoking often seems to be associated with various negative aspects, it might be helpful to consider
another behavior, like dangerous sport activities – a behavior which Segall equally finds it reasonable to
expect individuals to avoid.19
Snowboarding
Suppose a group of people are very much into snowboarding in a way that involves great risks. They find a
place around a ravine between two mountains over which they exercise their activity. The activity gives
them great pleasure, even though, or perhaps because, it involves great risks. Now some of the group
members choose to quit the activity after some years, as they get other concerns and interests and
therefore no longer find the activity worth the risk. Other group members continue. For them this activity is
a considerable part of what brings them value in life. Except for this activity the members of the group
differ in no relevant way from the rest of the population. They do their jobs and so on, but in the weekends
they go to this ravine and snowboard.
Now the activity leads to quite a few number of deaths. The thing is that if a snowboarder falls down the
ravine, the chances of surviving is only 10%. For this reason the average cost of these group members’
burdens on the health care system is a good deal lower than in the rest of the society. 90% of the instances
of potential health care needs simply never turn into burdens because the snowboarders simply die. In 10%
of the accidents, however, the snowboarders hit the edge in a way so they do not fall all the way down the
ravine, and they most often get away with only a broken leg.
So do we have justice-based reasons to treat these snowboarders when they get to the hospital? Segall’s
use of RA implies that we do not. The snowboarders’ broken legs are simply outcomes of action that it is
17
Op.cit., p. 1
18
This is not to deny that smoking would count as option luck in an option/brute luck distinction. The point is rather that such distinction is
implausible when applied on health matters, which e.g. Segall’s example on earthquake risks in California already illustrates.
19
Op.cit., p. 81
7
reasonable to expect individuals to avoid. Due to his meeting basic needs-model he would naturally say
that we have other reasons to treat them, but that, insofar it is possible, it would be just to pass on the
costs to the snowboarders themselves.
I’m not very sure this matches our intuitions. We might naturally have various reasons to worry about this
activity, e.g. to be sure the snowboarders are properly informed about the risk (or if we have very strong
reasons to believe it is a very bad trade-off between prudence and pleasure, we might even forbid it). But
that, I believe, would be due to our concern for the snowboarders, and hence for paternalistic reasons,
rather than because we find it unjust for the rest of the society to pay for an instance of accidence
stemming from a behavior we somehow find “imprudent”, and – most importantly – a behavior that in
general terms does not lead to higher health care costs, but to de facto savings.
In responding to the objection that smokers may not be a burden at all, Segall contrasts his position to
Dworkin’s “resourcist” view:
“The latter is premised on the thought that each individual is allotted an equal share of resources, which
she is then free to do with as she pleases.”20
Indeed this is quite a different view, which I do not intend to defend. But I see no clear reason why a critic
of Segall’s view, i.e. that justice requires passing on expenses stemming from imprudent behavior even
though the behavior generally is not a burden, should be committed to accept this. Rather, Segall’s view
could be modified, saying that we have reasons to pass on the expenses stemming from a behavior X, if
a) X, generally speaking, leads to higher health care expenses than non X, and
b) We do not have other relevant reasons to find it unreasonable to expect individuals to avoid X
Further Qualification of Segall’s Use of RA
Now one further detail regarding Segall’s use of RA has been found: the size of the general costs brought
about by a certain behavior is irrelevant.
I will now go on and further specify how to understand Segall’s use of RA. I will offer two cases to illustrate
why his use of RA is not sufficiently attractive and hence must be modified.
Recall the two principles I derived from Segall’s examples on what is and what is not reasonably avoidable.
Taken together we get:
20
Op.cit., p. 83
8
It is unreasonable to expect individuals to avoid X, if
a) it is unreasonable to expect everyone to avoid X, or
b) the health risk attached to X is lower than Y
Due to Segall’s non-paternalism the justification for these requirements, as we have seen, cannot be the
concern for individuals imposing risk to their own health, but that it is unjust for the rest of the society to
pay for an instance of accidence stemming from behaviors that it is reasonable to expect individuals to
avoid.
Therefore, more explicitly, b should read:
b*) the risk of burdening the health care system, due to X, is lower than Y
Getting this straight it seems we can derive a further implication. Now recall my analysis of Segall’s example
(3) and (5), respectively, that it is unreasonable to expect individuals not to settle in certain parts of
California due to the slight chance of being hit by an earthquake21, and that a woman who gets pregnant in
spite of being warned by her physician of a high risk to her health forfeits a claim to treatment on justicebased grounds.22 I concluded that it is difficult to believe that Segall’s use of RA involves comparative
considerations on how much value an individual is giving up by avoiding X.
I believe therefore that we should read b* very explicitly: what matters for RA is the risk of burdening the
health care system, not what the risk regards:
Emil
Imagine Emil, who loves to cut little wooden Indians in his garage when he comes home after a hard day of
work. By doing this he risks cutting his finger at two levels: 1) sometimes he gets away with just a scratch,
but 2) he also runs the risk of cutting his finger more seriously – in fact to a degree where he will be in need
of treatment to get his finger sewed. Now 2 has the exact same accidence-frequency as an average
smokers risk of getting sick, because of smoking, to a degree where she will be in need of treatment.
If my reading of Segall’s use of RA is correct, it will imply that we do not have justice-based reasons to treat
Emil when he gets to the hospital with his needy finger, since 2 has an accidence-frequency higher than Y,
21
Op.cit., p. 20
22
Op.cit., p. 22
9
and (I take it not to be the case) it is not unreasonable to expect everyone to avoid cutting little wooden
Indians.
The point is that insofar the justification for RA is that it is unjust for the rest of the society to pay for an
instance of accidence stemming from behaviors that it is reasonable to expect individuals to avoid, then the
gravity of what an individual risk to herself does not seem to matter. A risk higher than Y of a small burden
is reasonably avoidable just as a risk higher than Y of a big burden.
This seems contrary to a more personal risk evaluation. We might take e.g. smoking to be a bad trade off
because we do not find the pleasure (if it is a pleasure) to be worth the combination of the risk of falling
sick, and the gravity of what the risk regards (lung cancer etc.). Reversely, we might find Emil’s hobby to be
rather harmless; even though the accidence-frequency is just as high as for smokers, the risk of being
sewed in the finger seems like a small price to pay.
In fact, if b* is correct, then it leaves us with an even more remarkable situation:
Cigrays
Imagine a new product entering the market. It’s a kind of a gum – addictive, just as smoking, which soon
20% of the population comes to like. As it turns out after some years 99% of those who have been enjoying
Cigrays simply die when they are in their late sixties and their early seventies. They simply dies from what
looks like a very effective heart attack, and they therefore never arrive at the hospital with any needs. The
remaining 1%, however, gets terribly sick – equally in their late sixties and their early seventies – and will
thus be in need of treatment.
Now as it turns out the fact that 99% of those enjoying Cigrays simply die, the risk of burdening the health
care system due to Cigrays is lower than Y, and the use of Cigrays then cannot be said to be a reasonably
avoidable behavior. Segall would simply have to say that it is unreasonable to expect individuals to avoid
Cigrays due to the slight chance of burdening the health care system.
To sum up I find it hard to see why we reasonably can expect individuals not to smoke (or eat too much) if
smoking (or obesity) does not in fact lead to higher health care costs.23 When paternalism is not an option
23
It might be suggested that not only health care costs should be considered, but also matters such as loss in productivity and other forms of wealth
creation. I disagree. I believe such considerations go beyond what we owe each other in a liberal society. Say Peter (e.g.) has a well paid job and
therefore benefits the society through taxes. If he wishes to retire early, we would not say – on justice-based grounds – that he is not allowed to do
so, even though it would lead to fewer taxes. Similarly, we would not say that he is not allowed to over-eat or smoke because he thereby runs a risk
of an early dead or an early retirement, and thus will pay fewer taxes, even though this very well might be a loss for the society. If we nonetheless
do decide to take non-health care budgets into account, I am not sure it would change the conclusion. The consequences of living an unhealthy life
often occur late in life, and we therefore might gain significantly due to a drop in pension expenses. And, not least, lifestyle diseases are tellingly
10
we namely cannot justify that due to our concern for the smokers. We can only argue that it is unjust for
the society to pay for a particular instance of imprudent behavior. In generalizing Segall’s use of RA I have
found that a risky behavior is reasonable insofar as the risk of burdening the health care system is lower
than Y, which has lead to the conclusion that it is unreasonable to expect individuals to avoid Cigrays, but
reasonable to expect individuals like Emil not to cut little wooden Indians. I believe these findings to be
contra intuitive – also for Segall, who avoids paternalism, but states that:
“Pursuing the kind of lifestyle that results in a long and healthy life is precisely what any health care system
is supposed to promote.”24
Due to these findings I therefore take it that Segall must either say that it is unreasonable to expect
individuals to avoid smoking (and obesity), or give up his premise of non-paternalism.25
Inequality in Health
Segall’s luck egalitarianism does not only address the matter of access to health treatments. It matters
similarly for our judgment of when inequality in health is unjust. On this matter Segall favors luck
prioritarianism:
“Fairness requires giving priority to improving the health of an individual if she has invested more rather
than less effort in looking after her health, and of any two individuals who have invested equal amounts of
effort, giving priority to those who are worse off (health-wise)”26
I now want to show that when a view on when inequality in health is made sensitive to individuals different
effort it leads to unattractive consequences in regard to our priority of medical research.
In justifying this luck prioritarianism Segall objects to the claim that inequality in health is always unjust. It is
namely implausible, he maintains, that the inequality between a jogger and a smoker is unjust. In defense
of his position he therefore states:
more widespread in lower socio-economic groups, who by definition are the groups that contribute less to the society. All this of course must be
made subject for an empirical investigation.
24
Op.cit., p. 83
25
It might be argued that smoking is simply bad for the society. But I see no liberal reasons why that should be the case in a way that is not
reducible to particular individuals for whom smoking is god or bad.
26
Op.cit., p. 112
11
“It not only avoids the smoker/jogger objection but appears, in general, less paternalistic than equalizing
health as such. For society, according to the ideal of equal opportunity for health, only seeks to allow
individuals to be as healthy as they themselves choose to be.”27
Now the view that inequality in health is always unjust, does not necessarily lead to paternalism. We might
simply say that the inequality between the smoker and the jogger is unjust, but since we do not wish e.g. to
forbid the smoker to smoke, and find it even more unjust to level the jogger down, there might simply be
nothing we can do about it.
But there might potentially be something we can do about it, and our view on when inequality in health is
unjust will therefore matter for our political decisions hereupon:
Second Objection
Imagine a society in which 20 % of the population has between 10% and 20% reduced sight. They are thus
in need of either glasses or lenses. The reason these people have a reduced sight, let us further imagine,
stems from genetic dispositions. Another 20 % of the population are smokers, who smoke to a degree to
which they in average suffer a loss of six years of life expectancy (and X healthy years of life) compared to
the rest of the society.28Imagine furthermore that public funded research takes place in the society, and a
pot of money is to be decided to be spent on one of two research projects:
a) An optical drill that is 100% effective in operating up to 20% reduced sight. The operation has no
side effects and we have reason to believe that at least half of the people with reduced sight will
ask for it.
b) A 100%-effective pill that removes both the physical and mental desire to smoke. The pill will have
no side effects. Due to a formerly made opinion investigation and the existing marked for smoking
rehab we have strong reason to believe that at least half of the smokers will take it.29
The chances of success, let us finally assume, are for both projects equally high.
Now it seems that Segall, due to his luck prioritarianism, will find only justice-based reasons to opt for the
former project, as the people having reduced sight have invested more effort in looking after their health,
27
Op.cit., p. 100
28
Find valid reference
29
In fact, an American study shows that 79% of those who smoke would like to give it up. See: http://www.gallup.com/poll/7270/most-smokerswish-they-could-quit.aspx
12
than has the group of smokers. And it seems hard to believe that there, due to Segall’s theory, can be other
reasons to opt for the latter – after all this is not a case of meeting basic needs. In contrast to smokers,
people with reduced sight, even though only slightly reduced, are, when stemming from genetic
dispositions, worse off than others due to outcomes that it would have been unreasonable to expect them
to avoid.
As such it is irrelevant that we have all reasons to believe the latter project to result in tellingly more
QALY’s, or that we know – due to the social gradient in health – that smokers are overrepresented in the
lower social groups, and thus are worse off in general, in terms of welfare or resources (as currencies of our
egalitarian aim).
Now it is possible to read Segall as if his most primary aim is to show that (contrary to e.g. Daniels30) a
“responsibility”-sensitive universal health care system is possible (i.e. without abandoning the reckless).
Given his model of meeting basic needs and ex ante taxation of reckless behavior it seems that he does
show that (even though, as I hope to have showed, his use of RA is less plausible). But regarding the
distribution of research funds this does not apply, and the premise of “responsibility”-sensitiveness
therefore seems to block the more general health efficiency – whether we aim for egalitarianism,
prioritarianism or utilitarianism.
Part 2: RA and Responsibility
I now want to move on to the second part of this article and examine whether RA is a more plausible basis
for distribution than merely “responsibility”.
As I stated earlier Segall suggests that we should “… replace “responsibility” with a more plausible
understanding of what constitutes a case of brute luck.” Rather, he maintains, we should understand
“brute luck” as RA. He stresses that “expectation” is to be understood as a normative expectation rather
than an epistemic one: “We are not inquiring here into what people are likely to do, but rather what it is
that society can reasonably expect of them.”31
30
Se e.g. op.cit., p. 30-34 and Daniels, N., Just Health: Meeting Health Needs Fairly, Cambridge University Press, 2008
31
Segall, ibid., p. 20
13
Again, as Segall suggests replacing “responsibility” with RA I take it to be a misunderstanding to read RA as
if it was another concept of responsibility. Rather, as I understand it, we should replace “responsibility”
with RA as the basis of distribution within luck egalitarianism.32
In part 1 of this article I have therefore ignored the matter of responsibility and examined and criticized
Segall’s conclusions in terms of RA. Here in part 2, I will first argue that we can hardly operate with RA
insensitive to some view of responsibility. Secondly I will suggest that we consider RA to carry a double
meaning: It both equals a rather specific idea of what we, more literally, reasonably can expect individuals
to avoid, and an umbrella concept, serving as the basis of distribution, respecting matters of responsibility,
incentive regulation and reasonable avoidability in the literal sense.
So in his justification of RA it seems that Segall is assuming a rather common sense understanding of
responsibility, namely in a sense that implies agents, in terms of merely themselves, to be able to do
otherwise.33 For instance he is assuming Californians to be responsible for residing in (earthquake) risky
places, and therefore suggests that we, in spite of that, should see it as unreasonable to expect them to
avoid it.
But it might be that Segall’s idea of this replacement is more radical. In a footnote to his replacementsuggestion he is namely referring to Ripstein’s article “Equality, Luck, and Responsibility.”34
Now in this article Ripstein argues against what he calls the naturalization of responsibility. He states that:
“In purely naturalistic terms, there is no obvious place to stop in tracing back the causes of behavior.”35
A naturalization of responsibility is thus most likely to come to mean that responsibility is impossible. He
therefore maintains that:
“The appropriate response (…) is not to give an anti-naturalistic metaphysical account of responsibility, nor
to naturalize the notion. It is rather to change the subject, and make responsibility a question that is – to
borrow a phrase from Rawls – political, not metaphysical.”36
32
Throughout Health, Luck and Justice Segall seems to be using RA and responsibility interchangeably, but I take his use of responsibility to mean
RA.
33
See e.g. op.cit., p. 14.
34
Op.cit., p 177, note 25
35
Ripstein, A., 1994, “Equality, Luck, and Responsibility”, Philosophy and Public affairs , Vol. 23, no.1, pp. 3-23, p. 5
Op.cit., p. 10
36
14
Ripstein thus argues for a purely political concept of responsibility, which it is possible that Segall lends
support from in his concept of RA.
If this is the case then it nonetheless seems that we, in our reasoning on what is reasonably avoidable,
necessarily share at least some intuitions in regards to our reasoning on when the proper conditions for
responsibility are met. Therefore I do not see that RA escapes the legacy-problems of responsibility.
Imagine the following (by borrowing from and reframing a Frankfurt-case37):
Jones is a smoker who considers quitting rather often. Black however, who is an excellent brain surgeon,
wants for some reason Jones to continue smoking. Therefore, without Jones’ awareness, Black implants a
mechanism in Jones brain so that he can supervise and intervene in his brain-processes. The next day Jones
in fact decides to quit smoking, but shortly after Black activates the mechanism in Jones brain, so that Jones
decides to undo his decision and hence lights up another cigarette.
Now, in this case I do not think our intuitions support the claim that we reasonably can expect Jones not to
smoke. So regarding the possibility to do otherwise (an often raised requirement in theory of
responsibility38) it seems RA is just as open to metaphysical objections as responsibility is. If we have
reasons to believe that determinism (hard or soft) is true, then I cannot see how it would not affect our
view on when an act or omission is reasonable to expect individuals to avoid.
It seems therefore that Segall in his claims on what we reasonably can expect individuals to avoid
presupposes one understanding of responsibility and disregards others: it is unreasonable to expect
firefighters to avoid risks even though they are responsible, and it is unreasonable to expect Californians
not to settle in California even though they are responsible, and so on. If responsibility, in the sense Segall
presupposes, is correct, then it does seem more plausible to operate with the concept of RA. Therefore,
given this reservation, the claim should read:
If firefighters are responsible for a higher risk of incurring burns, then it would nonetheless be
unreasonable to expect them to avoid it.
But what if it turns out that responsibility (including compatibilism) is merely implausible? Would that be an
argument against RA as a more plausible basis of distribution?39 Well, I do not think it would, even though it
37
Frankfurt, H. G., 1969, “Alternate Possibilities and Moral Responsibility”, Journal of Philosophy, 66
38
Se e.g. Hurley, S., 2003, Justice, Luck, and Knowledge, Harvard University Press
39
We need only think of responsibility in the sense that affects distributive justice. I read e.g. Scanlon’s distinction between attributability and
substantive responsibility as a distinction between what affects us being open for moral criticism and what affects distributive justice. See Scanlon,
T.M., 2000, What We Owe to Each Other, Harvard University Press, p. 251
15
naturally would be an argument against quite a few particular claims on what precisely we reasonably can
expect individuals to avoid. I will try to explain why.
RA as an Umbrella Concept
As it seems obvious that what we reasonably can expect individuals to avoid is sensitive to our view on
responsibility, I suggest we consider RA also as an umbrella concept.
Segall considers the objection that it is unclear what more precisely we reasonably can expect individuals to
avoid, and that RA therefore leads to indeterminacy. In fact he accepts that, but insists on seeing it as a
source of strength:
“The strength of the “reasonable avoidability” criterion is that it can give due consideration to the changing
circumstances of each case. (…) We could envisage some extreme circumstances under which something
like China’s “one child” policy is a reasonable one to pursue.”40
This would namely mean that RA also captures aims of incentive regulation. Suppose responsibility turns
out to be merely impossible and we, like China, faces a problem of increase in population (assuming it in
fact is a problem). And let us further assume that incentive regulation in this respect in fact changes
couples’ behavior. Would it be just pro tanto to say that we reasonably can expect couples not to get more
than one child? I think we would have to say yes. Therefore changes in the concept of responsibility do not
rule out the plausibility of RA, though it naturally affects the plausibility of particular claims on what we
reasonably can expects individuals to avoid – namely the claims that stem from matters of responsibility.
And being able to capture matters of incentive regulation in matters of distributive justice might in various
cases turn out to be valuable:
Suppose our answer to the equality of what question is welfare, and Thomas, through no fault of his own,
ends up with a rather expensive taste. Only a high consumption of vintage Champagne and caviar can make
him as well off (welfare-wise) as others.41 Now, luck egalitarianism commits us to compensate him, and if
no better option exists, then to provide him with the Champagne and the caviar that his welfare requires.
RA, on the other hand, can more easily escape this conclusion, since also incentive regulation is captured in
the concept, and we (I assume) do not wish to encourage such a behavior.
40
Segall, ibid., p. 22. This contradict Segall’s earlier point that it is unreasonable to expect every woman not to get pregnant. But again, what it is
that is unreasonable for everyone to avoid is a variable.
41
Acknowledging a well-known Dworkin-case. Find
16
However, I do not mean to suggest that welfare-egalitarianism necessarily would be inconsistent in
concluding the same. The point is rather that RA as such is closer to an implementable principle of
distribution. Comparatively, we might see responsibility (or lack thereof) as being too narrow to capture
what we generally find to owe each other.
RA, however, cannot just be reduced to a combination of responsibility and incentive regulation (assuming
both to be variables here). If that were the case then it would not make sense to claim that Californians
only unreasonably can be expected to avoid residing in the certain areas of California. If they are
responsible for it, as Segall suggests, then we would only have incentive regulative reasons to say that we
cannot reasonably expect them to avoid these areas, which would not make any sense. They are
responsible, and we can hardly find reasons to encourage people to reside in these areas (ceteris paribus),
yet it is unreasonable to expect them to avoid it.
Segall therefore does provide new meaning to our basis of distribution. So there is a factor of reasonable
avoidability.
Given these considerations I thus find it more convenient to consider RA to carry a double meaning: It both
equals a rather specific idea of what we, more literally, reasonably can expect individuals to avoid (though
sensitive to whatever responsibility commits us to) and an umbrella concept, serving as the basis of
distribution, respecting matters of responsibility, incentive regulation and reasonable avoidability (and
these matters naturally affect each other).
Even though we then have (at least42) three uncertain variables to determine, it does seem that such a
framework potentially can tell us more accurately when to distribute.
So to sum up, it seems that RA is to be determined in a combination of responsibility, incentive regulation
and what more literally is reasonable to expect individuals to avoid.
The Space for Reasonable Avoidability in the Literal Sense
Now Segall suggests that even though e.g. Californians are responsible for residing in risky areas, then it
would be unreasonable to expect them to avoid it. That, I believe, is plausible, but needs elaboration:
If Californians are responsible for residing in risky areas, then it would nonetheless be unreasonable to
expect them to avoid it.
42
I mean, there might be more to consider
17
This seems to suggest the space for RA in the literally sense (so that RA cannot be reduced to only incentive
regulation and responsibility). It is less obvious, however, whether the possibility of responsibility (including
compatibilism) will turn out to be a necessary condition for the space of RA in the literally sense. If, say, the
woman who opts for pregnancy in spite of being warned by her physician is not responsible for getting
pregnant, and there is no present need for incentive regulation, then how can it nonetheless be plausible to
claim it reasonable to expect her to avoid it?
Looking forward there seems therefore to be a task in providing cases where we, if responsibility is
impossible, and if we have no present reasons for incentive regulation, nonetheless plausibly can claim
reasonable avoidability. I find it doubtful that such a case can be found.
All this said I would like finally to suggest that if we opt for a luck egalitarian approach to justice in health
and health care, then we actually need a replacement of responsibility with RA or something similar.43 This
is for a very simple but (presumably) missed reason: If responsibility is the basis of distribution, and it is
true that e.g. smokers are responsible for smoking, then it would also follow that non-smokers ceteris
paribus are responsible for not smoking. And if non-smoking, or more generally living a healthy life, is
associated with a higher risk of being in need of various late-in-life operations, like e.g. hip operations, then
it seems that healthy individuals are responsible for that higher risk, which, admittedly, seems
unreasonable. Luck or responsibility are namely neutral in regard to value. Contrary, with a distribution
model of RA, Segall may be able to succeed in adding the premise that it is unreasonable to expect an
individual to avoid living a healthy life, since it would be unreasonable to expect everyone to avoid living a
healthy life.44
This, however, should not (and does not) commit us to say that it is reasonable to expect individuals to
avoid living an unhealthy life.
Final Comments
In this article I have examined Segall’s concept of reasonable avoidability (RA) in two levels. In part 2 of the
article I have first argued that RA appears to be sensitive to our view on responsibility. I have therefore
43
For alternatives see e.g. Eyal, N., “Egalitarian Justice and Innocent Choice”, Journal of Ethics and Social Philosophy, Vol. 2, No. 1, 2007.
44
There is still a problem that this approach does not avoid: Segall opposes to Daniels’ rawlsian theory partly because it fails to provide justicebased reasons to treat people after they turn 75 (Segall ibid., p. 33). Segall suggests the alternative that it is unjust for individuals to be worse off
than others due to outcomes that it would have been unreasonable to expect them to avoid (op.cit., p. 13). Now it might be a controversial matter
to decide when someone is worse off than others, but at some point we will find elderly individuals, whom we do not have justice-based reasons to
treat, since they have already lived for so long that we cannot plausibly say that they are worse off than others.
18
argued that we are better off considering RA as an umbrella concept capturing (at least) matters of
responsibility, matters of incentive regulation and matters of reasonable avoidability in the literal sense.
I have argued that Segall makes a strong point in suggesting a replacement of responsibility in favor of the
broader concept of RA as the basis of distribution. In fact I have argued that a replacement of responsibility
is needed in order for luck egalitarianism to apply plausibly on matters of justice in health and health care.
The value of that replacement, however, is sensitive to our view on responsibility, since if responsibility is
merely impossible (including compatibilism), then the space for arguing that something nonetheless is
reasonably avoidable (in the literal sense) seems rather limited, if not entirely absent. Of course, if
responsibility is impossible, then RA, in contrast to responsibility as the basis of distribution, has the
advantage of capturing matters of incentive regulation, if e.g. we are facing a problem like China’s of
increase in population. Being able to capture matters of incentive regulation, however, is hardly epochmaking, since other versions of luck egalitarianism probably too are open for matters of incentive
regulation. The gain in this respect is only to capture two concepts in one.
Therefore the advantage of RA lies in the space where responsibility is possible (and even more in the sense
Segall presupposes), since it then enables us to compensate individuals for something, even though they
are responsible for it. As Segall has suggested, it then becomes possible to provide medical treatment for
pregnancy and delivery and help to Californians in case of an earthquake – and both for justice-based
reasons.
In part 1 I have examined Segall’s particular use of RA without taking responsibility into account. I have
accepted the premise of replacing responsibility with RA, even though I in part 2 deny that we can operate
with a responsibility-insensitive idea of RA.
I have examined Segall’s particular claims on what we reasonably can expect individuals to avoid and
concluded that his use of the concept implies that it is unreasonable to expect individuals to avoid a
behavior X, if
a)
it is unreasonable to expect everyone to avoid X, or
b*) the risk of burdening the health care system, due to X, is lower than Y
These premises lead to Segall’s claim that we reasonably can expect individuals not to smoke for nonpaternalistic reasons, and even though the health care costs associated with smoking are lower than those
associated with non-smoking.
19
I have found this claim to be implausible, since it can hardly be unjust for the rest of the society to pay for a
particular instance of accidence, if the cause of that accidence generally leads to lower health care costs.
To support that claim I have showed that the premise b* also leads to the conclusion, that we do not have
justice-based reasons to treat Emil when he too often comes into the hospital with the need of getting his
finger sewed, but that we do have justice-based reasons to treat users of Cigrays. Insofar the empirical
study is sound (and smoking and obesity therefore leads to health care savings) I believe therefore Segall
must either say that it is unreasonable to expect individuals to avoid smoking (and obesity), or give up his
premise of non-paternalism.45
In part 1 finally, I have showed that when a view on justice in health is made sensitive to individuals
different efforts, then it seems to lead to rather ineffective and hence contra intuitive results when it
comes to the distribution of medical research projects.
If we combine the findings in part 1 and part 2, we might ask whether e.g. Segall’s claim, that we
reasonably can expect individuals not to smoke, is somehow strengthened if we include the assumption
that smokers are responsible for smoking. They might well not be a burden, but if they are responsible,
then – if they fall sick due to smoking – they are worse off, but due to matters of their own responsibility,
and it follows from strict luck egalitarianism that they forfeit their legitimate claim to compensation.
But the point in using RA rather than responsibility is exactly that it can be unreasonable to expect
individuals to avoid X even though they are responsible for X. I find it plausible to believe that partly due to
Segall’s own cases – e.g. that it is unreasonable to expect individuals to avoid residing in California. I will
therefore suggest that even if smokers are responsible for smoking, then insofar as smoking cannot
generally be said to be a health care burden, it cannot non-paternalisticly be reasonable to expect them to
avoid it. Whether an individual smokes or not, she will most likely become a burden to the health care
system sooner or later. By choosing to smoke (or to become obese) smokers are simply offering the society
a discount. And whether we ought to intervene in smoking (and over-eating) is therefore a matter of
paternalism (or, technically speaking, whether the smokers (and the obese) wish for it46).
It is less obvious that this answer avoids my second objection (that we have justice based reasons to fund
the optical drill, but not the pill that removes the desire to smoke).
45
This is not to say that we, if it turned out that smoking and obesity leads to higher costs, necessarily have justice-based reasons to tax them. But
we might say it gives us a pro tanto reason to do so.
46
If smokers ask for the tax it would not be paternalism. See also: Gruber and Mullainathan, “Do Cigarette Taxes Make Smokers Happier”, The B.E.
Journal of Economic Analysis and Policy, Vol. 5, 2005
20
Even if it is unreasonable to expect individuals to avoid smoking, then it still seems to be the case that
smokers ceteris paribus have invested less effort in looking after their health, than has the group of
reduced sight individuals.
But note that this mismatch also affect Segall’s claims: even though it is unreasonable to expect individuals
not to settle in certain parts of California, it is still the case that individuals not residing in these parts of
California, ceteris paribus, are investing more effort in looking after their health than those who do reside
there…
Not least therefore, political philosophy will still be left with the highly important matter of finding the right
understanding of responsibility.
Literature:
Daniels, N., Just Health: Meeting Health Needs Fairly, Cambridge University Press, 2008
Eyal, N., “Egalitarian Justice and Innocent Choice”, Journal of Ethics and Social Philosophy, Vol. 2, No. 1, 2007
Frankfurt, H. G., 1969, “Alternate Possibilities and Moral Responsibility”, Journal of Philosophy, 66
Gruber and Mullainathan, “Do Cigarette Taxes Make Smokers Happier”, The B.E. Journal of Economic Analysis and
Policy, Vol. 5, 2005
Hurley, S., 2003, Justice, Luck, and Knowledge, Harvard University Press
Ripstein, A., 1994, “Equality, Luck, and Responsibility”, Philosophy and Public affairs , Vol. 23, no.1, pp. 3-23
Scanlon, T.M., 2000, What We Owe to Each Other, Harvard University Press
Segall, S., 2010, Health, Luck and Justice, Princeton University Press
Vallentyne, P., “Brute Luck, Option Luck, and Equality of Initial Opportunities”, Ethics 112 (2002): 529-57
van Baal PHM, et al., 2008, Lifetime medical costs of obesity: prevention no cure for increasing health expenditure.
Plos Med 5(2): e 29. Doi:10.1371/journal. Pmed.0050029
http://plato.stanford.edu
http://www.gallup.com/poll/7270/most-smokers-wish-they-could-quit.aspx
21