Why Values are Always Apt for Technological Disruption

Tweet

A common claim in the philosophy of technology is that technologies can alter social values. That is to say, technologies can change both how we perceive and understand values, and the way in which we prioritise and pursue them. A classic example is the value of privacy. Some argue that the prevalence of surveillance technology, combined with the allure of convenient digital services, has rendered this value close-to-obsolete. When push comes to shove, most people seem willing to trade privacy for digital convenience. Contrariwise, there are those that think that the technological pressure placed on the value of privacy means that it is more precious than ever and that we must do everything we can to preserve it.

But how exactly does technology alter values? It must have something to do with the way technology mediates our relationship with the world around us. In this article, I want to briefly outline some of my own thinking on this topic. I do so in two parts. First, I argue that values exist in conceptual spaces, and that these spaces consist of internal conceptualisations and external conceptual relations. Within these conceptual spaces, values are always, somewhat, under pressure and open to disruption. Second, I argue that technology can have two main effects on these conceptual spaces: they can change the practical achievability of particular conceptualisations of a value, making it harder or easier to obtain; or they can change the priority ranking of a value relative to other external values. This can result in a number of different possible value changes.

The thoughts I elaborate here are distilled from previous work that I published with Henrik Skaug Sætra. We have written two articles on: (i) how technology changes the values of truth and trust and (ii) six different mechanisms of techno-moral change. Both articles are available in open access form. You can read them if you want to follow up on the ideas presented here. I should also add that there are other valuable contributions to the literature on technology and value change. These include the article by Jeroen Hopster and his colleagues on the history of value change, and Ibo van de Poel and Olya Kudina on pragmatism and value change (to name just two examples).

1. Values Exist in Conceptual Spaces

To start with, I suppose we must ask the question: What are values? There are many esoteric philosophical treatises on this topic. I will avoid getting drawn into their intricacies. For present purposes, I define values as things that are, in some sense, desirable, worth pursuing, and/or worth honouring. They are properties of events, states of affairs or people that make these things good and worthwhile. Freedom, pleasure, justice, friendship, sexual intimacy, cooperation, truth, fairness, love, knowledge, beauty are all classic examples of values. There are also disvalues, which are simply the opposite of values: things that are to be avoided, rejected, dishonoured. They are properties of events, states of affairs and people that make them bad.

The metaphysical grounding of values is not something I will discuss in this article. For the purposes of understanding technology and value change, we do not need to take a stance on this matter. For this inquiry, values can be understood in a descriptive and sociological sense: what is it that people say they value? What sort of values are revealed through their behaviour? How do these social facts change over time? Changes in social values may or may not track some ideal sense of values. This does not mean, however, that inquiry into technology and value change is irrelevant for understanding ideal values. It could be that technology pushes people closer to or further away from ideal values. It could also be that technology opens up or allows us to discover new ideal values. It all depends on the effect that technologies have on the conceptual space of values.

That's enough scene-setting. What about this conceptual space of values? This is an abstract space, not a tangible one. It is a bit like the space of possible novels or possible design specs for cars. The conceptual space of values is vast. I suspect it is theoretically infinite insofar as virtually anything could be a value under the right conditions. Nevertheless, for practical purposes, the space is more limited. Most societies have sets of basic values. Many of these basic values are shared, more or less, across most cultures.

There have been various anthropological, psychological and philosophical attempts to identify a core set of basic values. The philosopher Owen Flanagan, for instance, has written about the shared virtues (kinds of value that inhere in individuals) across Western and Eastern cultures. The anthropologist Donald Brown has created a list of 'human universals', based on a cross-comparison of cultures. While this list encompasses many features of human society, it includes values such as play, generosity, responsibility, communality, friendship and so on. More systematically, the psychologist Shalom Schwartz has developed a theory of basic human values. He argues that there are ten individual values shared across most cultures. These values are categorised into four main domains: self-transcendence values, openness to change values, self-enhancement values and conservation values. Schwartz claims that this theory has a good deal of empirical support. The image below summarises his ten values.

In addition to these academic efforts at identifying a core set of basic values, there are projects in the political/legal realm that do something similar. For example, international human rights documents, such as the Universal Declaration of Human Rights, could be viewed as attempts to identify and protect a basic set of human values (insofar as rights typically protect values).

I could go on, but I won't. This small sample of examples suffices for the point I wish to make: although the conceptual space of values is vast, it may be possible to make it more manageable by positing a list of basic values. Some people resist this idea. They think that universalising theories, such as those developed by Brown and Schwartz, simplify the messy complexity of human societies and underplay the real and substantive differences.

Although I lean in favour of universalisation, there is no need to take up a controversial position on the credibility of these theories here. Even if one accepts them, one is not obliged to accept that every society takes the same attitude or stance toward these values. Each theory posits a plurality of values. Societies (and individuals) can vary in how they order and rank these values, and how they deal with tradeoffs between the values. So, for example, if you accept Schwartz's idea that there are ten basic values, and that these values can be ranked in any order, there are (correct me if I'm wrong) 10! possible orderings of value. That's a lot of possible variation (over 3 million). Furthermore, nothing prevents you from adding to this list of basic values with more specific or unique values, relevant to particular societies.

In any event, what really matters for the inquiry into technology and value change is not getting precise about the number of values, nor demarcating the limits of the conceptual space of values. What matter is that values are, or at least are perceived to be, plural (there are many of them). A single value thus exists within a conceptual space of other values. The precise identities of these values matters less than their plurality.

2. How Values Change Within Conceptual Spaces

Turning then to value change, within the conceptual space of values, there are two main dimensions along which a single value can vary:

Internal dimension: How we conceive of or understand the individual value can vary, i.e. our particular conceptualisation of the value can vary.

External dimension: How the value relates to other values can vary, e.g. how it is ranked or traded off against other values within the same conceptual space can vary.

 

Let's consider both of these forms of variation in more detail.

Philosophers will be familiar with the idea that values can have different conceptualisations. Freedom is a value, but the concept of freedom is vague. To render it more precise and meaningful, people offer more detailed conceptualisations of it. On a previous occasion, I highlighted the work of the intellectual historian Quentin Skinner on the different conceptions of freedom that have vied for attention in liberal societies since the 1600s. Skinner, like many others, divided them into two main groups: models of positive freedom (i.e. to be free is to authentically realise oneself) and negative freedom (i.e. to be free is to be free from the influence and control of others). Within the negative freedom conceptualisation, Skinner identifies two further sub-conceptions freedom as non-interference (to be free is to be free from the actual interventions of other people through force or coercion) and freedom as non-domination (to be free is to be free from the dominion or authority of others). Further breakdowns of each of these sub-conceptualisations are possible. For example, Christian List and Laura Vallentini have developed a 'logical space' of different possible conceptualisations of freedom. In essence, this is a two-by-two matrix suggesting that theories of negative freedom vary along two main dimensions, resulting in four different possible conceptualisations of that value. If you are interested, I discussed their model in a previous article.

The example of freedom illustrates the general phenomenon. For each value we care about, there are usually multiple possible conceptualisations of that value. This is true for values such as equality, love, and truth, just as much as it is true for freedom. This variation in the internal conceptualisations of a value matters because when we say that we care about a value, or a society commits to protecting that value in its constitution, it could mean very different things in practice. Which conceptualisation of the value wins out or dominates could vary across cultures and times. Shifting between different conceptualisations of the same value is one form of value change.

What about the external dimension? This is more straightforward. Single values (or their more precise conceptualisations) sit within plural spaces of other values. We are not just committed to freedom; we are also committed to equality. We are not just committed to love and pleasure; we are also committed to hard work and truth. Questions arise as to how we ought to promote or honour all these different values. Is it possible to promote or honour them all with equal zeal? Must we prioritise some over others? Are there genuine cases in which values conflict and we must pick and choose? In the midst of a pandemic, when the risk of death from infectious disease is high, is it okay to prioritise public health and well-being over freedom? Or is this a false dichotomy? Figuring out the answers to these questions is, I believe, an ongoing struggle for all societies. Even when there is temporary agreement reached on how to balance or prioritise values, this can be easily disrupted by changes in circumstances (natural disasters, war and, of course, new technologies).

Once we appreciate the complexity of the conceptual space of values, it is easier to appreciate the phenomenon of value disruption. Most values are subject to internal dispute as to their correct conceptualisations. Furthermore, most values must fit into a wider network of other values, with which they are not always compatible. As such, values are never really in a settled equilibrium state. It's more like there is constant Brownian motion in the conceptual space of values. Disruption is never far away.

3. How Technology Disrupts Values

Finally, we come to the question of how technology disrupts values. If you accept the argument to this point, you accept that values are always somewhat apt for disruption. They are never as stable nor settled as they may seem. Consequently, it may not take much for technology, particularly new technologies to unsettle the present equilibrium. I believe that there are two main ways in which this can happen.

The first is that technology can affect how salient and practically achievable a particular conceptualisation of a value is. This is a form of internal value disruption.

An example might be the affect of artificial intelligence on different conceptualisations of the value of equality. Broadly speaking, there are two main conceptualisations of equality: equality of outcome and equality of opportunity. These conceptualisations are not necessarily incompatible, but they are, in practice, in tension with one another. Equality of outcome is about making sure that everyone gets a roughly equal (or equitable) share of some social good, e.g. wealth, healthcare, housing. Equality of opportunity is about giving everyone a roughly equal chance to compete for desirable social opportunities (jobs and education being the most obvious two). Equality of opportunity is sometimes favoured over equality of outcome on the grounds that forcing everyone to have an equal share of some good is either impractical or impinges upon other rights and entitlements. At present, most societies pursue a mix of policies aimed at both types of equality, but most liberal societies tend to prioritise opportunities over any strict form of equal outcomes.

How might AI affect this? In a recent paper, I looked at a number of different scenarios, focusing in particular on the impact of generative AI on how we think about equality. On the one hand, I argued that recent empirical tests of the impact of generative AI on workplace productivity suggest that the technology is of greater benefit to less-skilled, less educated and less experienced workers (read the paper for a fuller review of the current empirical basis for this claim, with some important caveats). This suggests that generative AI might assist with the practical achievability of equality of opportunity, providing for a way to level the playing field at work over and above (and complementary to) existing forms of education and training. Thus, equality of opportunity is likely to be seen as more desirable and more practically feasible. That said, I also argued that this could be a short run disruption, particularly if AI comes to replace, rather than augment, many human workers. If that happens, then equality of opportunity (in its current form) may become less practically achievable because technology starts to suck up all the desirable workplace opportunities. This may force a reconceptualisation of equality of opportunity (i.e. focus on other kinds of desirable opportunity) or a switch to equality of outcome as the more practically achievable version of the value.

The second way in which technology could affect values is by changing the priority ranking of one value relative to another. This would be a form of external value disruption.

An example might be the effect of digital technologies, including AI, on the value of truth vis-a-vis other values such as tribal loyalty or individual psychological welfare. This is an example that Henrik and I discuss in more detail in our paper on the technological transformation of truth and trust. Very briefly, we argue that the pursuit of truth is an important social value and that truth is valuable for both instrumental and intrinsic reasons. But truth is always a somewhat unstable value since their can be uncomfortable truths (e.g. climate change) that conflict with individual and group preferences. Powerful people often have a reason to manipulate our perceptions of the truth to suit their own agenda; we often deceive ourselves by pursuing information that confirms our existing biases; and we often want to share the beliefs and opinions of our peers so as not to step out of line.

One danger with digital technologies is that they make it easier to manipulate and control the flow of information in a way that makes it more difficult to access the truth. Examples of this abound, but the obvious ones are the ways in which social media and the algorithmic curation of information push people into 'bubbles' and information silos, along with the ways in which AI facilitates the creation of hard-to-detect fake information (e.g. cheapfakes and deepfakes). These developments can make it harder for people to commit themselves to the pursuit of truth and more willing to trade the value of truth off against other more easily attainable values such as displaying tribal loyalty, virtue signalling to others and making themselves feel good by limiting their exposure to uncomfortable ideas. This could lead to a down-ranking of truth relative to other values.

These examples of value disruption are somewhat speculative, even though there is evidence for both of them. Nevertheless, I provide them here not because they are necessarily accurate but because they illustrate the general idea of how technology can disrupt the conceptual space of values.

Subscribe to the newsletter

Are Seatbelt Mandates Unduly Paternalistic?

Tweet

A while back, I interviewed the philosopher Jessica Flanigan about her views on teaching and research. The interview was part of my series on the ethics of academia. In the midst of that interview, I asked her a question that I asked many of my guests: do you really believe in the ideas/arguments you develop in your research? The particular prompt for that question, in Flanigan's case, was an article she wrote about seatbelt mandates. Seatbelt mandates are laws that make it illegal not to wear a seatbelt. Flanigan, writing from a broadly libertarian perspective, argued that they were unduly paternalistic and, hence, should be revoked. I wondered if she really hoped that policy-makers would take that argument seriously. She said she did.

I didn't ask the question because I doubted her sincerity. I asked it because I sometimes lack confidence in my own arguments and ideas and so I quietly hope that others don't take them too seriously. It was refreshing to talk to someone that apparently backed themselves. Anyway, since I spoke to her, I have used Flanigan's article about seatbelt mandates in several of my classes. I think it is a thought-provoking piece and it's interesting to see what students make of it since the conclusion it defends is so contrary to their own preconceptions and the moral status quo. In my experience, very few students agree with her, but it can be hard for them to pinpoint exactly where her argument goes wrong.

In what follows, I want to analyse and evaluate Flanigan's argument. I will outline its basic structure and consider how Flanigan defends each of the main premises. I will also offer some critical questions and reflections on these defences. This article is based on notes I have prepared for my classes so it may be a bit more disjointed and less conclusive than some of the other pieces on this blog. I tend to raise questions for discussion rather than offer firm opinions of my own. You have been warned.

1. The Basic Structure of the Argument

Flanigan sets out the main argument with admirable clarity. It is syllogistic in form. The major premise (1) states a general normative principle; the minor premise (2) makes a factual claim relating to the conditions stated in the normative principle; the conclusion follows. Here it is:

• (1) If a person’s autonomous choice is not harmful to others, then public officials should not paternalistically use force or threats of force to prevent him/her from making that choice

• (2) Riding in an automobile unbelted is an autonomous choice that is not harmful to others

• (3) Therefore, public officials should not paternalistically prevent people from riding unbelted.

The major premise echoes the Millian Harm Principle -- that the only moral justification for using force or the threat of force is to prevent harm to others -- but Flanigan offers a more elaborate and theoretically pluralist defence of it. She then evaluates three objections to premise (1) and two objections to premise (3). 

Before considering each aspect of her defence, it is worth briefly noting that in defending this argument Flanigan has an ulterior motive. She is honest about this in the introduction to her article. As already noted, she writes from a broadly libertarian perspective. This means she is generally opposed to government interventions and regulatory overreach. The argument against seatbelt mandates is a kind of Trojan horse argument. In other words, Flanigan knows that most people agree with seatbelt mandates and that opposing them will seem contrary to common sense. But if she can convince people that they are unjust, then this will make it easier to make the case for other limitations on regulatory interventions. As she puts it herself if "even seatbelt mandates are unjust", whither other forms of paternalistic intervention?

2. The Positive Defence of Premise (1): Against paternalistic intervention

Flanigan offers a pluralistic defence of premise (1). She suggests that it can be supported from different theoretical grounds (both deontological and consequentialist). You can count it up in different ways but, by my reading, she offers a least six lines of argument in support of premise 1. 

The first supporting argument is based on Kantian respect for autonomous choice. The idea is that we should each be given the authority to use our autonomous capacities to make decisions about how to live our own lives. This moral principle creates a ‘very strong presumption’ against interference. 

Of course, many people don't accept the Kantian view. And, even if they do, they tend to accept that there are limits to the scope of our own autonomy (Flanigan concedes this). The question then becomes: what are they and do seatbelt mandates fall inside or outside those limits? 

The second argument is a more constrained version of the first based on respect for autonomous choices relating to bodies. The idea here is that even if you reject the general Kantian view, you probably accept that we have a strong presumptive right to decide what happens to our own bodies. The argument then is that riding unbelted is a decision relating to what happens to our own bodies.  Flanigan fleshes this argument out by using some analogies such as base jumping and drinking to excess. If we allow those activities then why not allow riding unbelted? She also uses a thought experiment which is repeated or echoed various times in the article: Imagine someone threatened you for refusing to exercise. Would you accept such a threat? Surely not. So why accept it in the case of a threat if you don't wear a seatbelt.  

One danger with this argument is whether it is substantially different from the first one. Since we are embodied beings, virtually all decisions we make affect our bodies in some ways. If so, then it isn't clear that this sets a principled limit that is distinct from general Kantian view? More importantly, however, there is the important question as to whether the decision to ride unbelted is only about what happens to our own bodies. Aren't other people affected by that decision too? In class discussions, this tends to be the biggest stumbling block for students trying to accept Flanigan's argument. For them, the decision to ride unbelted poses risks to others and that's a major reason for seatbelt mandates. Flanigan addresses this counterargument later in her paper and we will consider it in more detail below.

The third argument is that paternalistic policies of this sort are unlikely to promote people’s well-being, at least in the long run. In making this argument, Flanigan offers a major concession: she accepts that seatbelt mandates would probably reduce overall deaths, maybe by up to 3,000 deaths per year in the US (according to a study she cites). But she counters that avoiding death is not the only thing that matters. Individuals are better than public officials to weigh the risk of death against other risks/benefits. Mandates take this choice away from people. 

This is a fairly standard libertarian argument and, in many instances, I agree with it. I am not sure that individuals are always the best judge of what is in their own interests, but I agree that in many cases they are no worse than other officials (etc). Still, the plausibility of this argument all boils down to how we draw the lines. Are there not some limits to an individual's determination of what is in their own interest. Do we have a strong right to self-harm?  How would the logic here apply to other cases? For example, should we have a right to sell ourselves into slavery? Or should there be a public prohibition against this? If so, then there are limits to what individuals can decide for themselves. And if so, it remains to be determined whether seatbelt mandates fall inside or outside those limits. 

The fourth argument is that there are often excessive costs to paternalistic interventions. The potential benefits of seatbelt mandates must be weighed against their likely costs, e.g. the waste of policing resources on enforcement; the administrative costs of the punishment system, and so on.

In my experience, students are often reluctant to accept this argument on the grounds that they are reluctant to put a price on the value of a human life (let alone 3000 of them). But since we do this all the time, there is some merit to the argument. If the economic value of a life is around 10 million dollars (which was a figure I saw cited for 2020) then the seatbelt mandate system must cost less than $30 billion for it to be cost effective (assuming 3000 lives saved). I'm not sure if that is the case, but I suspect it probably is. One danger, as well, with calculating costs in this area would be the risk of double-counting costs. The reality is that seatbelt mandates are likely to be enforced as part of a whole package of road traffic laws (speeding, drink driving, routine stops and checks for licences and registration etc). It might be hard to disentangle the costs and benefits of seatbelt mandates from the costs and benefits of these other laws.

The fifth argument is that the enforcement of such a paternalistic intervention would be inegalitarian. This is, perhaps, the most interesting of Flanigan's arguments. Her claim is that police are more likely to enforce the law on racial minorities and could use the law as a pretext for other interventions (Flanigan cites examples of how fatal shooting of black people arose from routine seatbelt stops). The punishments also impose a disproportionate cost on poor people. 

Students in my classes tend to find this argument the most persuasive but then wonder whether this an argument that works only in the USA. Racial discrimination in road traffic enforcement is (as far as I know) not a major problem where I live. Furthermore, fatal shootings in traffic stops are uncommon in most other jurisdictions. It is really the US police culture that is singled out by that argument. That said,  other forms of inequality arise in other jurisdictions and it's possible that these are reflected in how the laws are enforced. One final point, which is related to Flanigan's argument, is that if we assume seatbelt mandates are enforced through fines, then all flat rate fines impose disproportionate costs on poorer people. But is this a reason to abandon this form of punishment? Furthermore, it is possible to create a system of progressive fines for road traffic violations. Some European countries, for instance, have a system of fines calculated as a percentage of income.

The sixth argument is that paternalistic laws of this sort treat citizens like children and it is insulting and disrespectful to be treated like a child. Flanigan uses some analogies to make the case. She argues that it would be humiliating for a teacher to force a student to correct an addition mistake; it would be degrading for doctors to force patients to accept treatment. Her point is that there is an important moral distinction between pointing out someone’s error and forcing them to correct that error. It might be okay to point out to people that they are taking excessive risk by not wearing a seatbelt, but forcing them to wear one is going to far. She also argues that there is also some hypocrisy and inconsistency in the enforcement of such laws as they assume citizens are incapable of balancing risks and making compromises when public officials do this all the time, e.g. motorbikes are much more dangerous than cars and yet are routinely licenced.

There are several questions that can be raised about this argument. One is whether offensiveness/insult is an appropriate guide or restraint on legitimate criminalisation? Is it always insulting and offensive to be treated like a child? Are we not sometimes grateful when people force us to correct our errors? Furthermore, public administration is always complex and sometimes contradictory (e.g. drug laws versus alcohol laws). Is apparent contradiction/hypocrisy a good reason to object to a particular law (if that law is otherwise well justified)?

3. The Negative Defence of Premise (1): Can Paternalism be Justified?

Flanigan also offers a negative defence of premise (1) by considering potential objections to the premise. These arguments each try to offer justifications for paternalistic interventions in some cases.

Th first objection to premise (1) is to claim that we should be less worried about paternalistic interventions when the choices affected by them are relatively trivial or unimportant. The argument then would be that the preference for riding unbelted is trivial (not an important freedom) and so less worthy of respect.

Flanigan has three rebuttals to this argument. First, she says it doesn’t matter if it is trivial; the interference required to enforce it is non-trivial and we have a non-trivial reason to reject such interference. Second, she argues that some people may highly value the freedom to make the choice, even if they choose to buckle up all the time. Why discount or presume their subjective preference to be trivial? Who are we to make that judgment? Third, she argues that maybe we have developed an adaptive preference for riding belted (like being brainwashed) and if more people exercise the right to ride unbelted we wouldn’t see it as trivial anymore. 

To my mind, this third rebuttal is the most interesting and it is the one I often get students to reflect on in class. It is always worth asking whether we are victims of status quo bias when it comes to a tolerance for certain laws or interferences. What seems trivial or well-justified to us right now might not seem so trivial or well-justified to future generations.

The second objection to premise (1) is that if the paternalistic intervention ultimately promotes autonomy, it can be justified (we tolerate the intervention if it serves the greater good of autonomy overall). The argument then would be that seatbelt mandates do this by protecting us from a potentially deadly choice.

Flanigan has two rebuttals to this. First, she highlights an inconsistency potentially inherent in it. She points out that if we think people have a right to die, or to choose to die, we should think they have a right to ride unbelted. How persuasive this is depends on whether you think people have a right to die. Second, and perhaps more convincingly, she uses a slippery slope argument. She says that if this argument worked, it would also justify other, seemingly excessive, interferences with choice, e.g. mandated diet and exercise, visits to the doctor etc. I sympathise on the slippery slope point, but I wonder whether there is there an important difference between genuinely self-regarding decisions, such as right to die, and driving without a seatbelt which, given the nature of driving, always has the potential to harm others. Again, this potential for harm to others seems to be the main stumbling block for Flanigan's argument. 

The third objection to premise (1) is that people can consent to paternalistic interventions if, on balance, they promote long-term, higher order interests over short-term immediate preferences. The idea then is that seatbelt mandates to promote those long-term higher order interests by keeping us alive (or reducing the risk of death).

Flanigan offers two responses to this. The first is that, even if we did consent to such interventions, we would have a right to change our minds. She uses an analogy with joining a Weight Watchers programme. Imagine if we forced someone to continue with a weight loss programme that they really didn’t like. Second, she points out the obvious: we typically don’t get to consent or opt-in to seatbelt mandates. They are forced upon us. So this argument seems moot (of course, this leads to its own slippery slope since we don't consent to most laws).

4. Defending Premise (2): Don't Seatbelt Mandates Prevent Harm to Others?

Finally, we move on to Flanigan's defence of premise (2). As already noted, this is where, in my class discussions, most students tend to object. She considers two potential objections to that premise. The first is slightly weaker, in my view, but still interesting. The second is where most people have problems.

The first objection is that the decision to ride unbelted is not fully autonomous because it is irrational. Why so? Because it is shortsighted and not based on a proper understanding of the risks involves. The idea then is that irrational decisions need not be respected and can be paternalistically overruled. 

Flanigan offers four responses to this argument. The first is that the decision to ride unbelted may be undergirded by a rational preference for liberty/to be free from legal penalties. The second is that we may have rational grounds for favouring the decision to reserve for ourselves the right to decide irrationally. This is a convoluted idea, but Flanigan offers a simple analogy: falling in love. Love is often irrational yet we value the freedom to choose who we love. The third response is to appeal to another slippery slope. If we accepted the right to overrule irrational choices, this would justify many other, seemingly excessive, interferences, e.g. dietary preferences are often irrational yet no one thinks you should be forced not to eat another slice of cake. The fourth response is that, paradoxically, the apparent effectiveness of mandate laws is evidence for the autonomous nature of the choice to ride unbelted (b/c it shows people are swayed, rationally, by incentives).

Each of these rebuttals raises a host of issues that are worthy of further discussion. There isn't the space or time to do so in this post. One obvious point I would raise is whether the slippery slope really applies in this scenario. Are there not important differences between the levels of autonomy/irrationality involved in different decisions? Also, are there not important difference between activities that are wholly or largely self-regarding (e.g. diet) and activities that necessarily entail some interaction with or potential harm to others (e.g. driving on a public road)?

The leads us, finally, to the main objection to Flanigan's argument. This objection maintains that the decision to ride unbelted is harmful to others and so falls outside the scope of the normative principle stated in premise (1). In particular, the argument is that the decision to ride unbelted imposes harm, or a significant risk of harm, on (i) other passengers; (ii) traumatised bystanders (of road traffic accidents); and (iii) society at large due to medical costs and so on.

This is, obviously, a complicated argument since there are different kinds of harm to others at stake. In my view (and the view of most students I have discussed this paper with) the harm to other passengers tends to be the most significant objection to Flanigan's argument. Many years, in Ireland, there was a famous road safety ad which depicted, graphically, the potential impact of an unbelted rider on other passengers in a road traffic accident. You can watch it here. This ad has been imprinted on my memory and so probably biases my view on this matter.

But an ad is not reality. What is that the actual increased risk to other drivers? Flanigan cites figures from one study suggesting that the increased risk (of death) is only 20%. She then argues that this increased risk is comparable to (and perhaps less than) other risks that we do not legally protect, e.g. the risk of driving with an inexperienced driver, or driving late at night. She also suggested that, in many cases (but not all) the passengers consent to the additional risk by agreeing to ride in a car with an unbelted driver. 

I find these arguments unpersuasive and most people I discuss the argument with tend to agree. For one thing, it is not clear that a 20% increased risk of dying is trivial, nor that this figure is the correct one. You can find other studies suggesting that the risk is much higher, and that death alone is not the only risk to be factored in: there is also an increased risk of other kinds of injury. It is probably true that we are not consistent in how we legally protect people from the risks associated with driving. But, as noted above, inconsistencies are rife in legal policy: their presence, alone, does not mean that a particular law is unjustified if good arguments can be offered in support. There is also the relative ease of stipulating and enforcing a seatbelt mandate vis-a-vis other kinds of mandate (e.g. not driving late at night), and the different costs and benefits associated with those other potential mandates. Finally, on the consent point, it seems an implausible reading of human psychology to suggest that people can easily refuse to drive with an unbelted passenger. For young men, for instance, peer pressure and a sense of machismo might make it hard for someone to act on their true preferences. The mandate provides a useful fallback for such people. Similarly, there are many people who cannot easily choose who they ride with: children, elderly people and so on (Flanigan acknowledges this, to be fair).

Flanigan's objections to the relevance of the other two categories of harm are more plausible to me. She argues that potential psychic distress to third parties others is not a good reason to interfere with an autonomous choice. That seems right to me. And on the healthcare costs, she argues that if people have a right to healthcare, then they cannot be denied this right simply as a consequence of making bad choices that make it more likely they will need to exercise that right. By way of analogy she argues that protestors do not waive their right to police protection simply because they protest the police. Furthermore, if people do not have a right to healthcare, then we have no duty to pay for their treatment, so they do not impose any additional cost on us. If we treat them, it is a supererogatory act. 

That's the end of this summary of Flanigan's paper (along with some critical commentary of my own). As I say, the paper is thought-provoking. It raises a number of interesting issues in moral and legal theory and thus provides good fodder for classroom discussions, even if most people don't agree with it.

Subscribe to the newsletter