Tuesday, 28 August 2018

On the evolutionary dynamics of social injustice

I've got to blog about this book I read, that isn't even out yet, because it's awesome, and because y'all need to look out for it and go buy it when it's published:

The evolution of inequity by Cailin O 'Connor

is forthcoming from OUP. It is about inequality, especially gender inequality, and why we have it. But its game theory! I think of game theory as being a fairly macho area, the preserve of genius (male) mathematicians, its slang-terminology rife with masculinist metaphors and aggressive undertones. So its wonderful seeing it being applied to problems that are normally feminist territory:

How does social inequality come into existence - i.e. why do we live in a world in which people of one type are systematically disadvantaged compared to people of another type? Why does social injustice persist? More specifically: Why is domestic labour so often divided by gender, and so often unfairly? and What can we do about it?

The book is all the more wonderful for getting written by a woman who, judging from the photos on her website at least, happened to be raising twins at the same time. O'Connor, I salute you!

I'll start my exegesis with a point of which some readers will already be painfully aware. Women get a pretty shitty deal when it comes to the division of household labour. But don’t just take my word for it. According to the Office for National Statistics in Nov 2016, women on average do 26 hours of unpaid household work each week – that’s cooking, cleaning and childcare – while men do just 16 hours. And before you go thinking this difference is all down to women doing fewer hours of paid work, an Australian study found that women in full-time work still spend an extra 6.4 hours on housework each week, compared to full-time working men. Studies in the US have found that women who earn more than their husbands spend even more time on housework than women who don’t earn more! One researcher thinks that high-earning women do this as a way to neutralise the threat of their success on their husband’s masculinity. Hoovering, refilling soap bottles, ordering groceries, meal planning, returning school permission slips, watering the plants, keeping track of homework, writing out thankyou cards, arranging play dates, digging out the winter clothes, checking which clothes fit, making doctor's appointments, interviewing childminders, making pack lunches, disinfecting after accidents, applying the eczema cream, baking birthday cakes.....most men, I conjecture, don't even notice that this stuff gets done. It's unpaid, time-consuming, brain-sapping work. And personally, its not work that feels fulfilling because, with the odd exception, its not work that gets finished, work that delivers any sense of completion. It's work that I do because if it didn't get done life would be worse, and if I didn't do it nobody else will. Admittedly, I do a lot less of it than I might if I had higher standards - I've culled all inessential tasks, such as bothering to make our new home 'nice', doing christmas cards, having a social life, even though i can see that life would be nicer if these things were done (see post Situation vacant). And I have succeeded in pushing some of this work onto my husband - he does a lot of meal planning, for example. But I nonetheless do a disproportionate amount of it, because explaining to him what needs to be done would take more time than just doing it myself. And because a lot of the time I'm just reacting to crises in the moment rather than planning my domestic workload in any organised sense. I'm sorting through winter clothes, for example, because 'shit its time to leave for school and its freezing and where is your coat?!' Not because I have any sensible system in place. There are only occasional days during which I have the spare capacity to think 'It's not fair!'

Feminists have for a long time been debating the question of fairness in the gendered division of labour. So-called 'difference feminists' insist that it is not the nature of domestic work that is the problem, but only that the home-maker role and the work that it involves is denigrated and under-valued. Nurturing children and making a home require distinctive feminine skillsets and can be deeply fulfilling, albeit hard, work, they might say. But Susan Okin pointed out in 1987, even for a woman who adores her role as home-maker and feels valued and appreciated for it, even for her the division of roles by gender is unjust, because it erodes her freedom. A woman who devotes herself to caring for her family has no independent financial resources, but is dependent on other people – usually either the state or a partner. This dependence makes her vulnerable to various forms of exploitation and abuse. A woman who devotes herself to caring for her family has no time, typically, in which to  carry out activities essential to her well being and no time, critically, to participate in the sorts of political activities that are essential to maintaining one’s rights and autonomy. The longer she spends in this role, the worse the problem gets. Time spent working in the home is time not spent acquiring the sorts of education and work experience that can leverage one’s earning power, so her economic vulnerability only increases over time. Her ‘outside options’ (her possibilities for escaping from whoever she is dependent on) get worse and worse. Without serious prospects of finding a fulfilling and financially worthwhile alternative, she is stuck with continuing in the relationship, whatever abuse that might entail. Without serious prospects of enacting political change, she is stuck with the social system that got her into this mess. The gendered division of household labour, according to Okin, just isn’t compatible with a woman’s basic liberties. So however much a domestic goddess might love her side of the domestic division of labour deal, there are good reasons to think it’s just not in her best interests.

Many more women have entered the work force since Okin was writing in 1987. Not many households can afford to have just one breadwinner any more. But the signs are that rather than switch housework for paid work, women now just do both (see Hochschild's The Second Shift). Women are more present in the work force, but they're still the ones running most households. And we can adapt Okin's argument, arguably, to fit someone like myself: Someone who works full time but still finds themself doing a disproportionate share of domestic labour. I see my career slipping and backsliding under the strain of all this extra stuff that fills my brain. I see my paycheck falling behind that of my husband as the months of maternity leave and the years of exhaustion take their toll on my rate of promotion. All this is slowly but surely eroding my 'outside options' - the realm of possibilities that would be open to me if I were to need to end my marriage. So Okin's 'cycle of vulnerability' is turning - the longer a woman spends in an inequitable domestic partnership, the more she will come to be dependent on her partner for her self -esteem, financial security, social standing, political representation and so on. In a nutshell, the more time you spend devoting yourself to family work, the harder it is to do anything else.

Feminists have debated a variety of different solutions to this problem: We must abolish the traditional family and move to a system of artificial gestation and state orphanages (Shulamith Firsetone); We must abolish the capitalistic expectation of the 'ideal worker' who is able to devote most of their time to paid work (Joan Williams); We should pay women a proper salary for domestic work (Okin); We should oblige friends and relatives to participate in systems of extended social childcare (Sarah Blaffer Hrdy).

These are fascinating issues but there is an important question that underlies them and that has serious potential to affect the legitimacy of any answer: How on earth did we get into this mess?
That is, why are women in this disadvantageous position where they get penalised, in effect, for doing this work that holds the fabric of society together and guarantees the future of the species?

A traditional answer is biologically deterministic and functionalist: Things are they way they are because men and women specialise in the things they're good at, the things they're biologically well suited for. It just so happens that men are stronger, and lots of the important work requires strength, while women tend to be too busy with babies to do any important stuff. In other words, women do the house stuff because nipples and wombs and stuff. It's natural for women to be with the kids, because they breastfeed them. It's better if they stay in one place because women who are pregnant/carrying babies have limited mobility, so its easier for men to do stuff that requires going around a lot, plus men are stronger and more expendable so they should take care of the heavy or dangerous stuff like hunting and defending the camp from enemies. While women are sitting by the fire they may as well mend the cloaks and cook the mammoth etc.

One problem with a narrative like this is that rather than being founded in empirical evidence it  looks like a straightforward transposition of the modern home-maker/bread-winner distinction onto an imaginary prehistoric past - its about as sophisticated as the Flintstones. Some evidence suggests that Wilma spent most of her time moving around foraging for food with her baby strapped to her back, rather than scrubbing any sort of hearth. Fred did indeed spend large amounts of time hunting large dangerous animals such as woolly mammoths. But, according to some views, this was largely an exercise in sexual selection, rather than an activity which provided any meaningful source of protein to the family diet (further reading). On this latter perspective, Wilma was a largely self-sufficient single mum (albeit one who depended extensively on other females, especially post-menopausal women) and hunting was quite a lot like modern sports: Largely pointless but if a man gets really good at it he'll probably get to shag loads of women.

Feminist anthropologists and archaeologists have claimed that mainstream views about human evolution are horribly corrupted by the sexist lens through which they've mostly been generated (See Sally Slocum 1975). But alternatives seem, just as much as their androcentric rivals, to assume from the outset the existence of two complementary genders that carry out different tasks: one role for the men, another for the women .  Why not discard that assumption, and ask why we don't have a society in which every human is a jack of all trades, sometimes hunting, sometimes babysitting, and so on. Or a society in which there are three roles, or four or five, and each role is filled with males or females as they choose. Or any number of other ways of doing things. There are two sorts of questions: First, why divide labour by gender in the particular, unfair way that is typical (at least in middle-class, white, families)? But also, more generally, why divide labour by gender at all?
The biological determinist position is compelling, *if* you assume that there are only two roles that work can be divided into, and that one of them involves static childcare while the other involves heavy and motile work.   But why assume all that?

O'Connor has some answers for us. First up, How did we end up believing that people come in two varieties: men and women? Gender is, on her view, a socially constructed category of person which piggybacked on innate but arbitrary sex tags to solve a coordination problem. To unpack that, we might imagine a time in which humans were undifferentiated. In any situation where they had to divide labour ('you go and alert the others, i'll stay here and keep watch' or 'you light the fire while I butcher the carcass') they would have to spend time deciding who would do what, and possibly arguing about it. They would have occasions in which the task of allotting the roles failed, and they'd have all parties staying to watch but noone getting help, with subsequent disaster.

O'Connor points out that there is an efficiency saving, in these situations, if we can use some arbitrary convention to help us coordinate. Perhaps the rule can be that the tallest person gets help, while the shortest stays to keep watch. That way we can be sure that both roles get filled, and there are no arguments. The best kind of convention is universal - everyone knows the rule - and unambiguous - there won't be arguments about who is tallest etc.

Sex characteristics enter as a possible basis for a convention: they are visible and universal markers which can be used to sort people into two types. And unlike many other differences upon which people can be sorted (age, eye colour, etc) they will tend to divide most human groups roughly into halves, so they provide a decent basis for solving any coordination problem that requires two roles. Note that sex characteristics don't divide people into two groups very neatly - they're not discrete characters, and many people will fall in an intermediate position. They can be turned into a much more useful basis for solving coordination problems, however, if the people involved are encouraged to adopt stereotyped dress and behaviour, to make it easier for everybody to see immediately which type everyone falls into, and avoid arguments about which members of the group are male and which are female. So gender presentation evolves, very plausibly, as a way to make it even easier to use sex types as a basis on which to solve coordination problems.

O'Connor talks also about bundling (p.81): it might be that gender starts out as a convention that solves one particular coordination problem, who should take first watch at night, say, and has nothing to do with other sorts of work. But once the types exist, in the popular imagination, they become available to support other conventions too. If everyone gets used to the idea that the people with penises take first watch at night, and they subsequently need to decide who should go catch a mammoth and who should keep the fire going, they might find it natural to think of the penis or no-penis distinction as a useful way to divide people into the hunting/staying home roles, even if possession of a penis is of no relevance at all to the issue of who should go hunting, any more than it was to the question of who should take first watch. The gender categories have simply become salient as types of people. As these sorts of conventions are created, so is the idea that a human group contains one type of person who has a penis, takes first watch at night, and hunts mammoth. This distinctions become important things for any member of the group to learn, even if there is no non-arbitrary connection between having a penis and hunting mammoths.

The all-important step in O'Connor's thesis is the move from a society in which types are used to solve coordination problems, to inequality, ie a situation where one type is systematically worse off than the other. I was initially a bit confused at this step in the argument, for it seemed to me that O'Connor was assuming that there were two roles available to be filled, and that one of them left the occupants worse off than the others. I think  she does argue that, if this were the situation, people would nonetheless fill the inferior role *if* refusing to do so made them even worse off. And I think she does argue that women are in this position in some ways. We resent wiping the table, name-labelling the school uniforms, cleaning the toilet, but we do it anyway because the alternative is that it doesn't get done at all. Or we stay married to the takes-us-for-granted man-child, because divorce would leave us struggling with the mortgage, doing bed time alone, forced to give up foreign holidays. Thus is solved what Clarke and Blake 1994 call “the central paradox of political life:|Why people cooperate with their own subordination and exploitation in non-coercive circumstances".

However, if this were all O'Connor were doing, it would involve substantial and questionnable presuppositions: Why assume that family work has to be divided in two, and why does it have to be divided in such a way that one of the roles is worse than the other? Surely its not inevitable that domestic work is shared between only two people, or that only one of the two roles receive respect and material reward, or that the workplace be organised around people who have no domestic responsibilities. These are all cultural inventions, so if this were all that O'Connor could show, then the main part of the explanatory task of answering 'How did we get here?' would remain.

Upon a second reading, however, I think O'Connor does more. I think her argument is that once the gender types are in place, as a consequence of the need to solve some equitable two-party coordination problem, then they exist as general social resources. All we need to do to arrive at inequality is assume that there is some coordination problem which involves unequal roles, that just cannot be shared fairly, that leaves one party better off than the other, if only slightly. O'Connor sketches some different ways in which a role might be superior: it might be straightforwardly more pleasant. It might create a smaller reduction in other options. It might leave one with a larger amount of control over some thing. For example, if you live in a place where food can only be grown after fields have been ploughed, rather than a place where fields can be tilled using a hand axe, then everyone is dependent for food upon the person who operates the plough (p.35). The ploughman automatically holds power over his wife, as long as there is no possibility for her to plough for herself.

It is plausible to me that such a scenario will arise in most cases - that it would take an inhuman amount of effort and foresight to avoid any arrangement that left one party better off. Once again, we don't need to make any assumptions about how such a coordination problem is solved. It could be a coin toss that decides which of the types is going to do the superior role. We don't need to assume that one side is better at the superior role, or that one side is willing or able to use a strength discrepancy to secure for themselves the superior role. We only need to assume that the roles cannot always be dished out perfectly fairly. Once such a role is filled by convention then one of the types has a small advantage over the other. and here is where the game theoretic logic really becomes powerful: Once one type has a small advantage the rest is history, because the other side will never catch up. "Once inequity emerges in these models, it takes very little for it to persist indefinitely."p.5 The small discrepancy in bargaining power means that the initial advantage will get larger and larger, no matter how noble the intentions of the players. This happens because the discrepancy gives the two types of player asymmetric 'disagreement points' (section 5.1) or outside options - i.e. one of them needs the deal more badly than the other does. They have the weaker hand. So as long as each side behaves in such a way that they always seek to get themselves the best deal for themselves that they can (and lets not pretend it is a small matter to call such behaviour 'rational') an initial advantage will snowball until the two sides are locked in a situation where one type is very much worse off than the other. "When we see persistent inequity between groups, the explanation need not always appeal to anything in particular about those groups. Random chance can advantage one group over another, and the dynamics of bargaining conventions can sustain that advantage indefinitely.” (p.107)

So to pick the original questions back up: How did we get into the traditional gendered division of labour, in which humans fulfill one of two mutually excluding roles: home maker/breadwinner, and one of these roles invites glory, wealth, autonomy and leisure time, while the other attracts physical and emotional abuse? According to O'Connor's narrative, all this happened because humans found it useful to divide each other into two types, and somewhere along the line one of these two types - men - found themselves with an advantage that they have subsequently leveraged to the point where, in the UK in the 1950s everybody  expected that being a man goes along with doing a paid job outside the home, attracting material rewards along the way, while being a woman means working for one of them and doing all the things he doesn't want to do but that have to be done.

It is important to note that the traditional division of labour I described is traditional only in a particular white middle-class context, and that labour has been divided in very different ways in different contexts. O'Connor's account isn't limited to any context, however: she predicts that all societies will conditionalise on sex-types, and that type-conditioning will always lead to inequality, but the details about how roles are distributed are contingent and expected to vary.

O'Connor is at pains to point out what an important corrective her account is to a narrative which conceives of inequality simply as a moral failing to be fixed. The unjust outcome does not depend on one of the types intending to oppress the other, or using a size advantage to get what they want, or on the losing type having some sort of unconscious desire to be oppressed or whatever. Which is not to say that in real life groups haven't consciously tried to oppress each other (O'Connor is clear throughout that her models are simplifications that leave out many real world factors such as bias and stereotyping) - its just that even if they didn't, inequity would still emerge. Its simply a straightforward outcome of any bargaining situation that the side with the worst hand will most likely lose. It didn't have to be women that had the worst hand - but it was always going to be somebody.

Another primary fruit of O'Connor's application of game theoretic reasoning to problems of social justice is the realisation that inequitable arrangements are likely to become normal even if there are no differences in ability or power between the two types of people. If this message became widely accepted, it would dismantle the standard privileged assumption that the world is essentially fair, that people who do well generally deserve it, and replace it with a picture in which unfairness is the default, a picture in which the moral imperative is not to let nature run its course but to continually act to compensate for a natural tendency towards injustice.

What about the final and perhaps most important question: What should we do about all this? How do we fix it? O'Connor makes a heroic effort in the final chapter to prevent her account from being overwhelmingly depressing, but you shouldn't come to this book looking for silver bullets or even for reassurance. O'Connor argues that inequality is fairly inevitable: that is, it will emerge in a wide variety of contexts given a very minimal set of assumptions. She says not only will it be enormously difficult to eradicate inequality, but if we were to succeed it would be extremely difficult to prevent inequity from re-emerging.  We shouldn't think of the battle for social justice as something that could be conclusively won and finished with. It has to be an ongoing process in which we continually act against the reemergence of inequity, because The battle for social justice is against a hydra that grows a new head each time it is cut off” (163)

O'Connor's account neatly disarms biological determinism: the view that we oughtn't try to change things, because the traditional division of labour is optimal given our natural biological proclivities. O'Connor shows us that biological determinism is a red herring, it doesn't matter. We would have ended up in an unfair world whether or not its true that women are innately more suited to childcare than are men. Things would be stacked against *someone* no matter what our true nature is or implies. There is a potential lesson for men of the Alt-Right here: it could so easily have been you! In fact, given O'Connor's argument that all societies will have a tendency to return to an inequitable equilibrium even if we succeed in interrupting the current patriarchal one, it could be you next time! Perhaps this insight could be used to harness everybody's self-interest to the cause of preventing inequality.

No comments: