Can grading love and care (and other goods) be an injustice?
It is a widespread intuition that some things in life cannot and should not be measured. For example, quantifying our love for a partner seems problematic. We do not want to rate our affection on a scale of 0-100.* It is an important question, though, whether we can have a complaint of justice about measuring certain goods. Here I consider two lines of argument for thinking that measuring certain things in quantifiable terms can be objectionable.
The first is indirect. It concerns unjust effects of things being measured that were not measured previously. An example is the measurement of the willingness to pay for parking spaces, which Joshua Kopstein recently discussed. Some start-up companies have developed apps through which people bid for spare parking spaces. Kopstein suggests that this system turns a public good into a private good that is allocated according to willingness and ability to pay, thus privileging the rich. This example does suggest that certain kinds of measurement can lead to complaints of justice, if they introduce an allocation mechanism that is not appropriate for the good. But in such cases it is the possibility of wrongful use, not the measuring itself, that can be criticized.
The second way in which measuring could raise complaints of injustice is more direct. Consider a stylized example. Assume that elderly relatives have a legitimate claim to receive some acts of love and care from younger family members. Assume that a start-up company develops an app that evaluates family members, on a score from 0 to 100, on how well their acts deliver care to elderly relatives. And assume that using the app becomes a social trend, such that most people start using it. This might have some beneficial effects. For example, it might become easier to share knowledge about how to cheer up grandma “efficiently” when she is gloomy. But could it also mean that what the elderly relatives receive are not, any longer, acts of love and care, but something else: acts calculated to enhance the wellbeing of elderly relatives? If this is the case, it seems that they could raise a claim of justice. They are denied what they have a legitimate claim to receive. Schematically put, they have a legitimate claim to good X (love and care), but what they receive is good Y (acts that will efficiently enhance wellbeing), because by measuring and quantifying X, it is transformed into Y.
One problem here is whether we can specify a sufficiently clear and plausible account of what good X is and why good Y is different from it.** One possible issue might be that good X is a complex and multi-dimensional good, but by measuring it, we necessarily reduce it to fewer dimensions. Although modern technologies offer increasingly sophisticated ways of measuring things, they still cannot capture all the dimensions of what it means, for example, to have a trusting and loving relationship with someone. Another issue could be that offering good X requires openness to new challenges or a certain degree of spontaneity. Again, these cannot be easily captured in quantitative terms and are, thus, likely to be excluded if one tried to measure X. For example, an important aspect of a loving relationship is that one is sensitive to subtle changes in the other person’s situation, and maybe even that one understands such changes before the person herself fully understands them. It is therefore unclear how they could be included in quantitative measures.
Certain forms of measurement may be simply dysfunctional. In finance, there is Goodhart’s law: “When a measure becomes a target, it ceases to be a good measure.” This might also hold for other areas and make it simply unwise to try to utilise measurements there. But in additional to dysfunctionality, we should not exclude the possibility that measuring certain things may be an injustice. At least in the case of care and love, it seems there is reason to believe that that is the case.
*In Dave Egger’s The Circle there is an episode in which one of the protagonist’s lovers asks for an evaluation of his qualities, on a scale from 0 to 100, directly after the sexual act. The protagonist is somewhat startled, and then resorts to a white lie.
**Aspects of this question have been explored in the debate about limits of the market, where one concern is whether the socially defined “meaning” of goods can be a basis for not measuring goods in market terms. See for example Debra Satz’s discussion of Elizabeth Anderson’s approach in her Why Some Things Should Not Be For Sale.
Criticisms of quantitative fetishism often follow this line: “But you’re not actually measuring what you think you’re measuring!” Which if it holds is usually a pretty powerful/fair critique.
However what if we can accurately measure these things precisely (say with some bio-augmentation)? I think people would still object for, in my opinion, unfair reasons: 1) There is a thought that care should require effort – these devices make it ‘too easy’. 2) There is an objection to recording the data (as it is impersonal?) – all the app does which you outline is record information to allow for some degree of data analysis, if a grandkid can do this in their head without an app, is it still unjust?
Maybe I should put this another way. At the end of the day our brains are devices to record and process data. Apps and other external devices are used because our brains have limitations (behavioural biases, memory limits, computational ceilings).
In an attempt to record and analyse this data externally we are merely trying to do a better job at it than we were doing without any external assistance.
If we are not actually doing this, because we are recording the wrong data, then clearly this is a problem. But in scenarios where we can improve our abilities via schemes like this, it is not clear to me what the objection is – avoiding using these tools just provides an unfair advantage to those who are better (via genetics or socialisation) at performing these tasks without any assistance.
Lisa, thanks for the post. What I would be interested to ask you is about the ‘transformation’ effect you mention here. So, in the app case that you mention, I wonder whether the idea is that the required love and care have been ‘replaced’ or ‘supplemented’ (or something else)? That is, when I use the app and select an act to cheer up Grandma efficiently, is it the case that what would have been an act of love and care is: (a) no longer an act of love and care or (b) not *only* an act of love and care? My reason for the question is that, if it were (b), it would seem that elderly relatives still do receive what they are owed (as well as something else). The good, in this case, has multiple simultaneous components, one of which is the owed love and care. The description in your post, thus, suggests that what is happening is (a). That also seems consistent with some of your elaboration in the penultimate paragraph about measurement reducing a good to fewer dimensions. If that is right, I wonder if you could say something more on how that kind of transformation takes place – why it is the case that the good cannot have multiple simultaneous components, some being tracked by the measurement and others not? Is it that measurement makes it logically impossible/difficult to imbue an act with certain features, that we have a common tendency to cease imbuing acts with this feature when we measure, or something else?
As ever, a very interesting post, Lisa. Thanks!
I think that I'd like to hear a bit more about what you mean by 'measure'. When deciding whether to take Grandma out for lunch or to make lunch for her, should I not rely upon some kind of comparative judgement about how much I believe each activity will contribute to her well-being? If I know that Grandma does not like eating out, should I not take this information into account when deciding how to act? These questions are important since its not clear to me how, if at all, comparative judgments differ from measurements.
Moreover, let's now assume that I could acquire further information about Grandma's preferences in order better to judge which activity would most enhance her well-being. Perhaps I could simply ask her or her friends, or perhaps I could check which restaurants she's liked on Facebook. Do I have reason not to appeal to seek out such information? Its not clear to me that I do. The problem, however, is that this cases looks very similar to case that you mention that involves Gradma using a rating app.
As always, there is much more to be said here. Perhaps there is something distinctive about the kinds of measurements that you mention (i.e. those on a scale of 0-100), or perhaps there is something distinctive about the use of numbers. I'd be interested to hear more if you endorse either of these lines of argument.
Alternatively, as Brighouse and Swift argue in their recent book, perhaps there is something distinctively valuable about relationships involving spontaneity, and this is what is objectionable about appealing to these measurements? This, I think, is the most promising version of the argument, but my sense is that it will sanction only very few restrictions on appeals of measurements.
Hi Will, Andrew and Tom, thanks for those great questions, which help me to clarify my views. I'll try to answer them together (please do get back to me if I misunderstood/overlooked something).
– on effort: I think, Will, you are on to something there: the app could make certain things so easy that it would not take any time and care any more to think about them and prepare them. Now, I think it is not so much the effort in itself that is decisive, but what my actions express about my relationship to another person. And if things take no effort at all, this might express lower regard to the person in question than I should show (especially if it's a person from a different generation who thinks that it took me a lot of time to figure these things out – then there could even be an element of deception in it). Of course I could in theory do more of the things that are made easier, to keep the effort constant, but that might not be a realistic scenario in some circumstances.
– I don't think we should be opposed to external devices that help our brains to function better – we do this all the time (and some contemporary philosophers of mind hold that we should include pen and paper, and all kinds of others "scaffolds", including language and culture, into our conception of the mind, as the "extended mind". I think Hegel would have smiled). So I also wouldn't see any problems, as such, with using tools for finding out about, or recording, Grandma's preferences. It's only a matter of degrees whether I remembers them, or whether I write them down, or whether I have some tool for that. The problem is how these tools might affect what else is going on between my grandma and me (see below).
– Data recording might be problematic for completely different reasons, but I hadn't thought about it in the post, and I think these are additional reasons.
– replacement vs. supplementation: if it is supplementation, I don't see major problems. The crux of the argument really is whether or not there is some crowding out effect. Both of the mechanisms you describe, Andrew, could play a role: logical possible, and also simple human tendencies, such as focussing on certain things at the expense of others. One of the things where this seems a real possibility is spontaneity, as you, Tom, mention (haven't read Brighthouse and Swift yet, but this is something I had in mind). In addition to spontaneity, it might be a kind of openness and sensitivity to the other person that you can only realize in unplanned encounters, which might be crowded out by measurements. You are right, Will, that certain kinds of measurements might be particularly bad here, probably because of the way in which they function in focussing our attention – that might be a purely empirical question. With numbers, you often start this game where you want to improve your score, compare it to others, etc. – could be different if we simply had qualitative evaluations in different categories.
– I acknowledge that such tools – and measurement of many other things as well – can also have positive effects (as Will suggests), which would have to be taken into account in an "all-things-considered-judgment". What I am interested in is whether there is something that would be lost as well.
This is a very helpful reply, Lisa.
From the tone of the reply, it seems that you think that things other than spontaneity are also threatened by the use of certain kinds of measurement. Is this correct and, if so, what kinds of things do you have in mind? (I suppose that I share the intuition that you're getting at but, absent any further argument, I'm tempted simply to debunk it.)
Second, how significant do you think spontaneity really is? It seems that I can maintain the level of spontaneity necessary for love and care whilst using quite a lot of measurements. Providing that my conversation with Grandma is spontaneous, it would seem not much to matter, if at all, whether I had non-spontaneously calculated many other of my actions that affect Grandma.
Tom, I’m exactly in the situation you describe: I have the intuition that there is more at stake than just spontaneity, but I find it hard to put my finger on it. I think one way of putting it might be in terms of what your mind is directed at, and what you pay attention to: do you pay attention to the person in question, or do you pay attention to some measurement? It might be a bit similar if you have self-interested motives in an otherwise non-self-interested relationship with someone: it distracts your attention, and your thoughts aren’t where they should be, as it were. Now, this is of course a very hard thing to define, and even to know for yourself (although the other person might feel it very clearly, especially if she knows you well). We are used to thinking about „justice“ in terms of hard facts, i.e. clearly definable, epistemically unproblematic, and hence publicly enforceable things (the distribution of money being the clearest case). So maybe this is part of what may make us reluctant to think about such more subtle issues as matters of justice. But that’s a question of definition, in the end.
I agree that spontaneity can coexist with many forms of measurement. I am not so sure about attentiveness. Also, a third point that occured to me: a loving and caring relationship seems to imply that you are willing to do more than you would normally do if there is a special situation. Many forms of measurement create a tick-the-box-mentality: have I done enough for this week or not? This might also be a problem. But the more I think about it, the more I think that attentiveness is maybe at the core of things. Do you think that captures some of your intuitions, or would you also debunk it?
I share your intuitions in many of these cases, but I am much more reluctant to give these any normative force without more of an argument explaining and justifying their importance. For example, its not clear to me what precisely the problem is if I pay attention to some measurement only because it allows me better to serve the interests of the person in question.
I'm similarly unpersuaded by the box-ticking case. Indeed, to me this looks like great reason to appeal to more measurements! For all kinds of reasons, it is really important to know when I and others have done a fair share, and if this can be served by greater measurement, then great!
I agree that it is tough to find identify our subject matter in a compelling and clearly defined way. My worry, though, is that it we don't make progress with this task, it will be difficult to provide any argument in defence of the claims, and thus we most we can hope for will be intuitionism.
Hi Tom, I'm not sure whether I'd give up so easily. Of course, if by "justice" you mean "what a society should institutionally enforce", it matters that things can be very clearly defined, because they cannot be institutionally enforced if they aren't. If, by "justice" you also means "what duties of justice do I have to other people?", then enforceability is not an issue.
One might say, however, that the latter belongs to questions of character and personal relations, rather than justice. Let's assume that for now. Then there is a further question: should social structures be such that individuals are able to fulfill their duties of justice? And what does this imply for what should or should not be enforced? Etc., you see where I'm getting.
I'm inclined to think that we should try to understand such questions as well as we can, if only in order to defend liberalism against certain conservative or communitarian criticisms that hold that it is hostile to certain forms of personal relationships – or to see to what degree a liberal society can and should do more to make sure that it *isn't* hostile to them. Measurement is one issue where this question becomes relevant…
Thanks, Lisa. I have to say that I find this comment a little puzzling. The comment focuses on 'justice' and 'enforceability', but I didn't (mean to) say anything about either 'justice' or 'enforceability'. Perhaps I have missed something.
On the second point: I agree that any attractive political theory ought not to be hostile to normative significance of certain forms of personal relationships. However, I don't see how this bears upon issues of grading as such. In short, I think more needs to be done to meet the following objection: "If we could accurately measure love and care, then we would have reasons to do so. Since, we can't, don't. Does this mean we ought not to focus on love and care? No."
Hi Tom, maybe I misunderstood the thrust of what you were saying. To take up the second point (which is what I was really after): The way in which you put it here takes it as given that we cannot accurately measure love and care. If we assume that this is the case, then there are follow up questions that concern practices and institutions, for example in health care (and this is one of the places where enforceability comes in). I think the constellation is often as follows: we have certain institutional reasons to measure things, for example accountability or a fair distribution of burden (for example among employees in a hospital). But there are also reasons to think that by introducing measures, we are actually introducing proxies, and then individuals start following the proxies instead of pursuing that which they are supposed to pursue, and to which the recipients have a claim of justice.
Maybe private family relations are not the best example. But one can think of many other areas where this problem might occur.
Thanks, Lisa. Again, this is a very helpful clarification.
Am I right in thinking that you're objecting not to the use of grading and measuring *as such*, but instead to the use of proxies that risk tracking the wrong thing? If this is the claim, I am in agreement. That is, I agree that, if we are to use grading and measuring, we have a weighty reason to ensure we track the things that we are actually interested in. However, if this is the claim, is it not just an objection to the bad use of measurement rather than measurement as such? (Of course, this is not to deny that it is important to point out that we must be highly attentive to the potential misuse of measurements.)
You are right that there is an additional premiss (which you seem to share, but it would not be inconsistent not to share it): that love and care are indeed such that we would end up measuring proxies. So this, together with the argument that we should not measure if we can only measure bad proxies (and there are no other reasons for why we should measure), means that in this case it is wrong to measure.
I think there are many things for which measurement is unproblematic, or maybe: for which it is better, all things considered, to measure than not to measure. The question comes down to: what are the things where we end up measuring the wrong thing? It seems to me that there is a general tendency, at the moment, to want to measure *everything*, simply because there are so many technical possibilities (the internet, apps, etc.). This is why I think the issue deserves attention.
This is a fascinating topic – sorry to come late to the discussion. I think I agree both with you, Lisa, that attempts to measure love and care is wrong-headed and with you Tom that the problem is not, as such, with measuring (except in cases when measuring undermines spontaneity and spontaneity is a constitutive part of love&care, hence by measuring we destroy the thing that we deem desirable in the first place.)
Since agreeing with both of you seems inconsistent, then perhaps there are two different meanings of measurement here, the first (Lisa's) more precise and technical. The first and most obvious reason, as far as I can see, against measuring love is that we don't really know what love is. But assume it was possible to get a measure of how loved people are by simply asking them if they feel loved. I find it plausible that such a 'measurement' can be done without undermining spontaneity, intimacy or attentiveness, all of which I think are essential to the endurance of love (that thing whose nature eludes me). And suppose we know how to increase the likelihood of people *spontaneously* forming relationships based on love: By eliminating poverty (such that people are not exceedingly worried with survival and hence have resources for disinterested interactions with each other). By creating jobs in economically backwards areas (such that not all youth migrates, leaving the old behind). By creating social clubs, where everyone can go play social games, chat, dance, cook and eat with others. Through people-oriented urban planning (again to create social spaces where people can meet and have fun together). And so on.
Then you (the state) wait a bit and return to see how people are doing – are they still lonely? Depressed? (You may even want to ask them direct questions about the quality of their relationships.)
This approach to thinking about close relationships in relation to justice, which I think is the dominant one, seems to me both sensible *and* using some non-objectionable measurements. Measurement without numbers, that is. It won't be problem-free: you cannot, I assume, measure if the good in question is fairly distributed (precisely because you can't come up with numbers), but only if everybody has enough of it. And you can't do this kind of stuff, most likely, from a position of state neutrality.
This comment has been removed by the author.
Anca, thanks for bringing up this interesting scenario! Much to chew on in it! I guess one issue here might be a worry about perfectionism. But let's assume that a liberal state can be justified in providing settings that make various conceptions of the good life livable, and that the kind of social relations you talk about are part of some conceptions of the good life (in other words, let's assume that "state neutrality" does not mean that the state should just keep its hands out of everything, but rather frame social institutions in ways that give various conceptions of the good life an opportunity to thrive). Then it might well make sense to have forms of measurements, even for vague things such as "love". In such a case, the only thing I'd probably want to say is: let's be really careful in what kind of measurements we use, and what incentives this might create, and whether, if we anticipate some distortions through the process of measurement, it still makes sense to do it that way.
This may seem like an obvious and not very relevant point, but there are areas in real-life where this really matters, for example with regard to public education. For example, if you measure "quality of schools" by the grades students get, you might get a grade inflation without any real improvements – and the students might end up not getting something they are entitled to get. One might also ask whether GDP per capita, as a measure of the overall economic well-being of a society, is a good measure, or a proxy that can actually distort things…
Dear Lisa, thank you for your post and sorry for being REALLY late to the discussion. Just a quick thought on the points that you and Anca raised in your discussion. I believe that the issue of proxies and the question how to select the right indicators is a vital one which is often really overlooked. Indeed it seems very difficult to measure love directly. However, we may look for indirect ways, as Anca has pointed out. Poverty is a good example: it hampers your abilities and opportunities for forming loving relationships in many ways. Thus, relieving poverty may be one way to give people the opportunity for relationships without being overly perfectionist. Of course, hot measure poverty is another issue which I do not want to go into here, but still I think we should also think along the lines of proxies and interpretation of measurement data…
You are right – there can be cases in which there is a positive overlap, as it were, between different things, some of which we can measure, and some of which are harder/impossible to measure. I mainly focused on cases in which the relation is more problematic. But a full account of normative issues of social measurement would also have to include the cases in which there actually are very good proxies (and these can sometimes be very surprising – I once heard a talk about how the distribution of body height in a population can be used as a measure for equality…).