The trouble with feminism

In this post, I want to explain why I’m not a feminist and why I think you shouldn’t be either. Of course, I cannot discuss everything I think is wrong with feminism, so I’m just going to focus on what I take to be the most important problems with it. Before I do that, however, I must say what I mean by “feminism”, since this term clearly doesn’t mean the same thing for everyone and it is only in some of those senses that I’m not a feminist and think that feminism is mistaken.

In the past, “feminism” was used to describe the belief that men and women should have equality of opportunity and, while the details of this claim need to be worked out (there may be some occupations, for instance, where there is a good reason to prefer men or women), I don’t disagree with that. For instance, I agree that women should have the right to vote or shouldn’t be barred from studying at university, but I claim that “feminism” increasingly means something a lot stronger than that.

In fact, I think that many feminists today, if not most, are in practice supporting policies that would grant women more opportunity than men. For instance, when they support affirmative action policies designed to give preference to women in hiring, they are in effect supporting policies that are incompatible with equality of opportunity between men and women. Of course, they would deny that and claim that such a preference is only meant to compensate for the undue advantage men would otherwise have because of implicit bias or whatnot, but I think they are wrong because I believe that, in many cases where affirmative action in favor of women is implemented, men would not in fact enjoy any unfair advantage in the absence of affirmative action.

Thus, if feminism was just the belief that men and women should have equality of opportunity, I would agree with it. But I don’t think that’s what feminism is anymore. Philosophers of language disagree about what determines the meaning of a term, but they agree that, whatever it is exactly, it’s sensitive to the way in which that term is used. If a few decades from now, people stop using “cat” to refer to cats and instead use it to refer to dogs, it will no longer mean cat anymore. Similarly, while “feminism” used to be used to talk about the view that men and women should have equality of opportunity, I don’t think that is true anymore and therefore I think “feminism” no longer means that.

There is plenty of evidence for the claim I just made about how people use the word “feminism” and its cognates today. In particular, several polls in both the US and the UK found that only a minority of people identify as feminist, even though the overwhelming majority of them agree that men and women should have equal opportunity. For instance, a poll for Vox in 2015 found that only 18% of respondents identified as feminist, yet 85% of them said they believed in equality for women. Perhaps even more striking, a British poll conducted in 2016 on 8,000 people found that only 6.7% of the population self-described as feminist, while 60.7% believed in equality for men and women but didn’t self-describe as feminist. The figures for women were respectively 9.2% and 64.9%.

Even in polls that find a majority of people identify as feminist, there is usually a much larger proportion of people who agree that men and women should have equality of opportunity, which means that many are in favor of this equality but don’t identify as feminist. For instance, a poll conducted in France in 2014 found that 50% of respondents and 58% of women identify as feminist, even though 97% of them agree that men and women should be paid the same for the same job and 84% that women do most of the work at home.

I take this kind of facts to be evidence that “feminism” no longer means that men and women should have equality of opportunity. While the overwhelming majority of people agree with that view, a much smaller proportion, often a minority, identify as feminist. Some of this could be a result of social desirability bias, if people believe that feminists have a problem of image (which they do, as polls also show), but it’s unlikely to explain the whole discrepancy. I am not claiming that there is no ambiguity in the meaning of “feminism” and that many people don’t use it to mean something like equality of opportunity. On the contrary, it is clear that many people do, including feminists who defend much stronger views.

Indeed, it’s common for feminists who spend most of their time defending much stronger views to object to people who say they are not feminist that such a claim is hard to contenance, given that feminism is just the view that men and women should have equality of opportunity. But whether feminists realize it or not, feminism is no longer just that. It also comes with a variety of views, both descriptive and normative, that are not entailed by the proposition that men and women should have equality of opportunity. For instance, one could agree that men and women should have equality of opportunity, while insisting that, in the West at least, they already do.

Even if one does not think so, and the polls I cited above suggest that most people do not, there are a variety of ways to cash out what it means exactly for men and women to have equality of opportunity. Many people who agree that men and women should have equality of opportunity disagree with the way in which feminists understand it. Thus, while the shift of meaning I am talking about is not complete, I believe it is already well on its way. We have already reached a point where, given how the term “feminism” and its cognates are used, it is no longer possible to maintain that it just means the view that men and women should have equality of opportunity.

Some people who believe that men and women should have equality of opportunity but typically disagree with feminists are trying to reclaim the word. (I’m thinking of people such as Christina Hoff Summers, who I usually agree with, except when it comes to the question of whether people who hold the kind of views we do should use the word “feminism” to describe themselves.) I believe this is a mistake, but this is a political view, not a view about what “feminism” in fact means. I shall not defend that view here, I’m only trying to argue that, at this point, “feminism” doesn’t just mean equality of opportunity between the sexes. I think even people who agree with me that feminism, by which I mean what most people are talking about today when they use the term “feminism”, is misguided should acknowledge that such a shift of meaning has largely occurred, even if unlike me they think we should reclaim it.

It may seem that I have belabored this point, but I think it’s the source of many avoidable misunderstandings, so I thought it would be useful to discuss it at some length. Now that I have said what feminism is not, I must say what I think it is. This task is more difficult, because feminists hold a variety of views and often disagree with each other, yet I think it’s still possible to identify something they have in common and plays a central role in the feminist movement. I think a good candidate is the claim that, even in contemporary Western societies, women are oppressed. Marilyn Frye, a philosopher who wrote perhaps the most influential account of oppression among feminists, calls that “a fundamental claim of feminism”. I think a commitment to this claim also does a good job of distinguishing feminists from people who merely believe that men and women should have equality of opportunity.

I’m not claiming one can’t find people who qualify as feminists even though they reject this claim, but given how vague the boundaries of the feminist movement are, trying to find necessary and sufficient conditions for someone to be a feminist is probably futile anyway. Nor am I saying that one couldn’t find another claim that does just as good a job of setting apart feminists from people who merely believe that men and women should enjoy the same rights, only that no other claim will do a significantly better job. While the vast majority of feminists are committed to the claim that women are oppressed, they don’t mean exactly the same thing by that, so in a sense it is not exactly the same claim to which they are committed. Feminists hold a variety of views about the nature of oppression, but I will focus on Frye’s account, which as I noted above is probably the most influential. Beside, I think I could say very similar things against other feminist accounts of oppression, so I don’t think focusing on Frye’s version is really a problem.

I think it’s clear that, in the ordinary sense of the term “oppression”, women in contemporary Western societies are not oppressed. For instance, according to Merriam-Webster dictionary, oppression is the “unjust or cruel exercise of authority or power”. When people talk about oppression, they usually have in mind things such as the way in which black people were treated in the US when slavery was legal, the treatment of Jews by the nazis under the Third Reich or perhaps the severe legal and social restrictions that, even in the West, women faced up until a few decades ago . Now, whatever obstacles you think women still face in contemporary Western societies, it surely doesn’t rise to that level.

Thus, in the ordinary sense of the term, women are not oppressed. Feminists may think that the ordinary concept of oppression is defective, and they may even be right about that, but even if they are it doesn’t change the fact that, given how the term “oppression” is currently used by most people, women in contemporary Western societies are not oppressed. If feminists manage to convince people to use the word differently, perhaps this will change, but so far I don’t think they have succeeded. No matter how much they would like to, activists and intellectuals don’t have the power to change the meaning of words by fiat.

Indeed, although she never admits that she is redefining the term, Frye spends a lot of time in her influential paper anticipating objections from people who are using “oppression” in the sense this word has in ordinary language. In order to do so, she famously used the analogy of a bird cage:

If you look very closely at just one wire in the cage, you cannot see the other wires. If your conception of what is before you is determined by this myopic focus, you could look at that one wire, up and down the length of it, and be unable to see why a bird would not just fly around the wire any time it wanted to go somewhere. … It is only when you step back, stop looking at the wires one by one, microscopically, and take a macroscopic view of the whole cage, that you can see why the bird does not go anywhere; and then you will see it in a moment.

According to Frye, if people are denying that women in contemporary Western societies are oppressed, it is because they only attend to the obstacles they face individually, without stepping back to consider the structure they form. This is why people, including many women, don’t notice that women are oppressed, even though they are.

But in the ordinary sense of the term, there is no such thing as invisible oppression, which again shows that Frye is using “oppression” in a technical sense. To be clear, as long as one is clear that it is what one is doing, there is nothing wrong with that, provided that one’s concept is analytically useful. Indeed, as Rudolf Carnap famously argued in his book about the foundations of probability, the concepts of everyday language are vague and can be made more precise in a variety of ways. Sometimes it can even be useful to clarify the concept in a way that is at odds with the concept of everyday language one is making precise. Thus, while I think Frye’s account of oppression is not particularly useful and can’t do the job she wants it to do, I do not criticize it for being at odds with the ordinary concept of oppression.

If I thought her account of oppression was illuminating, instead of being a source of confusion, I would have no problem with using the word “oppression” to talk about the phenomenon she describes in her paper. But since I don’t think her account is really useful, I believe it’s important to resist using that word to talk about the kind of things she is talking about. Of course, in a sense, this is just a semantical issue. One could always stipulate that one is using “oppression” to talk about the kind of things Frye is talking about, even though it’s not the same thing most people are talking about when they use that word in everyday language. But semantics often matters, if only because people don’t react in the same way depending on what words are used, even if they are used to say the same thing. Activists understand this and that’s why they always try to bend the meaning of words. Unless you agree with their goals, you shouldn’t let them.

In another paper, Frye defines oppression as “a system of interrelated barriers and forces which reduce, immobilize and mold people who belong to a certain group, and effect their subordination to another group (individually to individuals of the other group, and as a group, to that group).” Women are systematically disadvantaged in a variety of ways because they are women. They are more often victims of sexual assault, have less political and economic power, are seen as mentally unstable when they exhibit anger, have to do most of the domestic work, etc. While taken in isolation these disadvantages may not seem like a big deal, Frye claims they reinforce each other and are part of the same structure that keep women in check, which is why together they amount to oppression. Moreover, they often have the property of putting women in a bind, by leaving them with no good options as to what they should do. For instance, if a woman complains because a man opens the door for her, she will be seen as angry and unreasonable. But if she does not, then she is forced to consent to this ritual, which according to Frye sends the message that women are incapable. Finally, not only are these various obstacles targeting women because they are women, but they also operate to men’s advantage.

The examples Frye gives are right out of the catalogue of injustices that, according to feminists, women continue to endure. Now I don’t deny that, even in contemporary Western societies, women are systematically disadvantaged in a variety of ways and that in many cases it’s at least partly the result of wrongful discrimination, although the precise mechanism is often not what feminists think. The problem is that, in many other cases, feminists are alleging that women face disadvantages when it’s not actually the case or, at least, the evidence they adduce does not show that.

For instance, as we have seen again during the last US presidential election, feminists often claim that female politicians face significant gender bias. They claim that women who run for office are less likely to be seen as strong leaders, that newspapers are more likely to comment on their appearance, that people are less likely to vote for them, etc. These claims have become so ubiquitous that even people who don’t identify as feminists assume they are true. Yet in a recent book on the issue, Daniel Hayes and Jennifer Lawless have shown that, despite what most people think, gender bias was almost non-existent in politics. Female politicians are no less likely than men to be seen as strong leaders, newspapers are no more likely to comment on their appearance, people are no less likely to vote for them, etc. Perhaps future research will invalidate some of their findings, but the important point here is that feminists constantly make that sort of claims without offering a shred of evidence, because that politics is rife with gender bias is so obvious to them that it doesn’t even occur to them that non-anecdotal evidence might be necessary to show it.

Even when women really are disadvantaged, feminists often claim that it’s because of wrongful discrimination, when in fact it’s really unclear. For instance, it’s true that, even in contemporary Western societies, women are paid less on average than men. Thus, in that sense, they are disadvantaged. But feminists often claim that it’s because employers have a preference for men just because they are men. Yet the economic literature on the gender pay gap does not support that claim. It’s plausible that such a preference on the part of employers explains some of the gap, but I think it’s fair to say that most economists would agree that, even if it does, it can only explain a relatively small part.As you can see on this graph from a review of the literature on the gender pay gap, once you control for number of hours worked, experience, occupation, etc., most of the gap disappears.

While the residual is often assumed to be the result of a preference for men just because they are men, we can’t actually make this inference, because it could just reflect unmeasured difference in productivity between men and women. (Of course, it’s also possible that unmeasured productivity favors women, in which case the residual underestimates the effect of discrimination. But as far as I know, studies that attempt to measure productivity, such as this study on lawyers, always find that doing so reduces the gap, often to the point that it’s no longer statistically significant, so that’s very unlikely.) Thus, while it’s plausible that some of the residual is due to a preference for men just because they are men on the part of employers, it’s not clear that a large part of it is.

Smart feminists are aware of this and that’s why they offer a more sophisticated argument for the view that discrimination plays a large role in gender pay gap. They correctly point out that, even if outright discrimination by employers doesn’t have much to do with the gender pay gap, it doesn’t follow that unfair discrimination against women is not a big part of the story. For instance, a large factor of the gender pay gap is that men work longer hours than women, but this has a lot to do with the fact that women still do most of the work at home and spend more time taking care of children. According to feminists, this is the result of discrimination against women, though not by employers. A disproportionate share of domestic tasks and child rearing still falls on women because not only men but also women themselves still hold sexist beliefs about women. Similarly, while the fact that women tend to choose different occupations than men explains a relatively large part of the gender pay gap, this fact is not innocent. Feminists claim that it’s because women are socialized to prefer low-paying jobs such as nursing instead of high-paying jobs such as software development. In other words, while outright discrimination by employers doesn’t explain much, more subtle discrimination in society at large does.

There is some merit to this argument. For instance, it’s indisputable that a disproportionate amount of domestic tasks still fall on women, which is no doubt partly because sexist attitudes persist. It’s also clear that children have a major impact on the gender pay gap. But despite what feminists often claim, it is really doubtful that, even if we could totally eliminate gender roles, most of the gap would not remain. As we shall see later, abilities and interests are not distributed identically in men and women. Sophisticated feminists do not deny that, but claim that it’s almost entirely the result of socialization and gender stereotypes. In general, for any psychological difference between men and women, they tend to assume that it’s entirely or almost entirely cultural. However, not only are there strong theoretical reasons to expect this not to be the case, but it’s not also not what the evidence says.

First, even before you have a look at the evidence, your prior that some of the differences we observe between men and women are natural should be pretty high. Indeed, humans are a product of evolution, just like other species. But we know that important behavioral/psychological differences exist between the sexes in other species, including among those which are closest to us in evolutionary terms. It would quite extraordinary if humans were the exception. Indeed, thanks to the progress of molecular genetics, thousands of genes that are expressed differently in men and women have recently been identified. There is no reason to expect that such differences can only occur for genes that have to do with what is going on below the neck and, in fact, some of the genes that were identified by this analysis affect brain tissue.

The evidence that bears specifically on the extent to which observed psychological differences between men and women are the product of socialization doesn’t support the feminist position either. If these psychological differences were mostly the result of the fact that boys and girls are socialized to conform to gender stereotypes, you would expect them to be smaller in societies where gender equality is higher. Indeed, as Alice Eagly and her co-authors, who can hardly be suspected of hostility toward feminism, wrote a few years ago, the “demise of many sex differences with increasing gender equality is a prediction of social role theory”. But as David Schmitt noted, this is not what the evidence suggests, on the contrary. Most sex differences in psychological traits are larger in societies where sex role socialization is more egalitarian and sociopolitical gender equity is greater. Moreover, when you control for a bunch of possible confounders, this typically remains the case and sometimes the effect grows even stronger.

This pattern is difficult to reconcile with the view that psychological differences between men and women are the product of socialization in the way feminists claim. But it’s consistent with the view that, to a significant extent, they are a product of evolved differences between men and women. It is true that, for most psychological traits, the difference between men and women varies a lot across places and times, but this doesn’t really threaten the view I just mentioned. First, because there are genetic differences between human populations, there is no reason to expect that genetic differences between men and women are the same everywhere. Moreover, the claim that evolution produced differences in psychological traits between men and women doesn’t mean that such evolved differences are insensitive to environmental factors. (In fact, as Schmitt points out, the way in which they are sensitive to the environment could even have been selected.) On the contrary, nobody doubts that genetic factors interact with environmental factors to produce whatever psychological differences between men and women exist in a given society, but it doesn’t mean that some of them aren’t natural.

We sometimes hear the suggestion that, because genes interact with the environment to produce a phenotypic response, the nature/nurture distinction can’t even be drawn. The argument starts with the observation that, when you say that a difference between groups or individuals for that matter is natural, what you mean is that it would still exist in a possible world where, unlike in the actual world, such and such environmental factors are present/absent. Conversely, when you deny that a group difference is natural, you are saying that it would not exist in such a possible world. This means that whether a group difference is natural depends on the kind of possible world at which you are evaluating the counterfactual, which is where the fact that genetics factors interact with environmental factors enters the argument.

Indeed, in the right kind of environment, virtually any genetics factors can be causally related to virtually any trait. For instance, it may seem that genes that determine eye color are unrelated to educational achievement, but they would not be in a society where green-eyed individuals are barred from going to school. Similarly, even though men are currently more interested in software engineering than women pretty much everywhere in the world, it’s likely that in a society where boys are systematically beaten up every time they even get close to a computer, not only would men not be more interested than women in software engineering, but the opposite would be true. Drawing on that observation, one might claim that the question of whether men are naturally more interested in software engineering than women does not even make sense, because whether men or women are more interested in software engineering depends on the environment. More generally, the fact that genes interact with the environment means that, for virtually any psychological difference between men and women, one can always think of a society in which this difference would not exist or even be reversed.

However, I think this argument is confused, because while I agree that when you say that a difference between groups or individuals is natural, what you mean is that it would still obtain in a different possible world, there are constraints on what this possible world can be like. In particular, it must be a world in which the groups or individuals being compared are treated in the same way with respect the relevant trait, which rules out a world in which boys are systematically beaten up every time they get close to a computer if we’re trying to assess whether men are naturally more interested in software engineering than women. Instead, we are thinking of a world in which boys and girls are not socialized to conform to gender roles, but which is otherwise pretty similar to the actual world.

The problem is that there is more than one way to treat groups or individuals in the same way with respect to any given trait. For example, we could make both girls and boys study mathematics for 3 hours a week, just as we could make them both study it for 6 hours a week. Now suppose that, when both girls and boys are made to study mathematics 3 hours a week, boys are better at it, but that when both are made to study it for 6 hours a week, it’s the opposite. In that case, the truth-value of the counterfactual will depend on the possible world at which it’s evaluated, but there doesn’t seem to be any objective reason to prefer one to the other.

This point generalizes, because for any difference between groups or individuals, even with the constraint that the groups or individuals in question must be treated in the same way with respect to the relevant trait, there are always infinitely many possible worlds at which the counterfactual could be evaluated and it’s doubtful that it has the same truth-value in every one of them. But I don’t think it means that we should conclude that claims about whether a difference is natural are nonsensical and that we should reject the nature/nurture distinction. I think it just means that, to some extent, the nature/nurture distinction is relative to a set of values and interests, which determine what region of the space of possible worlds is used to evaluate the counterfactual.

For instance, given our values and interests, it’s likely that both a world in which boys and girls are made to study mathematics 3 hours a week and a world in which they are made to study it 6 hours a week will be part of the region in question, but not a world in which they are made to study it 20 hours a week or one in which they are made to study it 15 minutes a week. The claim that men are naturally more interested in mathematics than women will only be true if that is the case in every possible world that belongs to this region. If that is not the case, as in the example I imagined above, then it’s not true that men are naturally more interested in mathematics than women, though in that example it’s also not true that women are naturally more interested than men in mathematics.

As far as I know, nobody has ever proposed this account of the nature/nurture distinction, but I think it’s the right way of analyzing it. This account of the distinction admits that it’s not value-neutral, but this need not be a problem and it doesn’t mean that we should throw it away. To be sure, it would be a problem if, because of that, feminists and their opponents were talking past each other, but I think it’s usually not the case. Indeed, it seems to me that, in the vast majority of cases when feminists and their opponents disagree about whether a difference between men and women is natural, their disagreement is not purely verbal. It’s not that, when feminists say that a different is cultural and their opponents say it’s natural, the former are talking about what would happen if some environmental factors were absent/present while the latter are talking about what would happen if different environmental factors were absent/present. They really disagree about what would be the case if roughly the same environmental factors were absent/present.

For instance, in recent years, feminists have managed to convince a lot of people that the underrepresentation of women in STEM was a problem. They claim that, in large part, it’s the result of discrimination against women. Now, I don’t think the evidence really supports this claim, but I don’t want to discuss this here. (For the record, I think the evidence is mixed, though ultimately my view is that, to the extent bias has any effect on the representation of women in STEM, it probably favors them overall. But this is a story for another time.) Indeed, whatever effect bias might have, what’s clear is that women would still be underrepresented in STEM even if there were no bias because abilities and interests are not distributed identically in men and women. As a lot of research in recent years has demonstrated, this alone explains most if not all of the underrepresentation of women in STEM.

Indeed, the fact that both abilities and interests are not identically distributed in men and women plays a role, though not in the same way. Let’s start with abilities before we turn to interests. There are differences in cognitive abilities between men and women, especially at the extremes, but the reason why men vastly outnumber women in STEM does not seem to be that women are less capable of studying these subjects on average than men. In fact, in many countries, the opposite seems to be true. (Incidentally, as the same paper also shows, the difference in academic achievement between men and women does not seem to be related to gender equality.) But it doesn’t follow that differences in the distribution of abilities between men and women do not explain the underrepresentation of women in STEM.

Indeed, although girls do not perform worse than boys in science and mathematics on average, they perform significantly better in reading.This means that although boys are not stronger than girls in science and mathematics, they are more likely than girls to be stronger in mathematics and science than in reading. It’s plausible that people are more likely to pursue a career in a domain that is a relative strength of them and, as I point out immediately below, this is supported by the data, so it explains in part why women are less likely to pursue a career in STEM. In other words, except perhaps among very high ability individuals, it’s not that women are less capable than men, they just have more options.

Differences in interests also seem to play a major role in the underrepresentation of women in STEM. As the same paper where I found the graph above shows, women are just less interested in STEM on average, which also goes a long way toward explaining why they are underrepresented in STEM. As you can see on this graph, the fact that women are less interested in STEM and that men are more likely to have a relative academic strength in science and mathematics explains most, though not all, of the underrepresentation of women in STEM.As in the case of the gender pay gap, the part of the gap that remains after you have taken into account these factors could be the result of discrimination, but it could also be a million other things.

The fact that women are less interested in STEM than men on average seems to have a lot to do with the fact that STEM occupations are more things-oriented than other jobs. Indeed, it’s well-established that women are more people-oriented, whereas men are more things-oriented. To be clear, we are talking about a very large difference, since it’s approximately equal to one standard deviation. This difference has been shown to explain a large part of the underrepresentation of women in STEM. Moreover, not only does the fact that women are more people-oriented and men are more things-oriented predicts in part the overall underrepresentation of women in STEM, but it also predicts relatively well the differences in the proportion of women in the various disciplines that are classified as STEM.

Thus, even if feminists were right that discrimination in hiring and a hostile environment explain in part the underrepresentation of women in STEM, they would still be underrepresented even if there were no such thing, because abilities and interests are not distributed identically among men and women. As in the case of the effect of occupational preferences on the gender pay gap, feminists usually acknowledge that once presented with the facts, but they insist that differences in the distribution of abilities and interests among men and women are the product of socialization. The problem is that it’s not enough to say that, you also have to show it, but the evidence doesn’t support this claim.

Indeed, sex differences in academic achievement are not related to gender equality across countries, which is not what you would expect if feminists were right. The same is true for sex differences in occupational preferences, which do not seem to be related to gender equality, as you can see below.Again this is not what you would expect to find if differences between the occupational preferences of men and women were mostly the result of the fact that boys and girls are socialized to conform to gender roles. In fact, what is remarkable is how stable those differences are cross-culturally, despite large variation in gender equality.

Another paper about the US shows that, as women entered high-status occupations in greater number, the amount of variance in the proportion of women in various occupations accounted for by job status decreased, while the amount of variance accounted for by people-things orientation increased.This paper also shows that women are more likely to work in people-oriented occupations and that this relationship did not change over time. Since gender equality significantly increased in the US during that period, this suggests that women’s preference for people-related occupations and men’s preference for things-related occupation is not related to gender equality, which again is at odds with the feminist explanation of why women are underrepresented in STEM and other male-typical occupations.

Even worse for the feminist narrative, the paper I already cited above on how men were more likely than women to be stronger in mathematics and science than in reading also showed that not only more gender equality was not related to a larger proportion of women in STEM, but that in fact the opposite was true.The authors of this paper hypothesize that, as women don’t have to worry as much about their economic future, they are more free to act on their preferences and, since they are more likely than men to be interested in non-STEM occupations, their representation in STEM decreases.

The evidence I have briefly reviewed supports the claim that differences in the distribution of abilities and interests among men and women, which result in different occupational choices and explain in part the underrepresentation of women in STEM and the gender pay gap, are to a large extent natural. As I explained above, this doesn’t mean that, in some possible environments, the distribution of abilities and interests among men and women could not be vastly different. Again, if you systematically beat up boys whenever they manifest interest for STEM-related pursuits, you would almost certainly succeed in reversing the sex differences we currently observe. But feminists obviously do not advocate such brutality and, while they disagree with their opponents about precisely what kind of interventions would be acceptable to increase the interest of women for STEM occupations, the kind of societies they imagine when they make claims about what would be the case if boys and girls were not socialized to conform to gender roles are pretty close if not totally similar.

It’s possible that, unlike their opponents, feminists envision a society where policies are in place to actively encourage women but not men to pursue a career in STEM, although even that is not clear since they usually say that such policies are only necessary because women are currently socialized not to be interested in STEM, but they are still pretty similar. Thus, if feminists and their opponents disagree about whether differences in the distribution of abilities and interests among men and women are natural, it is not for the most part because they disagree about what a desirable society where gender roles have been eliminated would be like. It’s because they disagree about what sex differences in abilities and interests would persist in such a society and, as we have seen, the evidence strongly suggests that feminists are wrong.

If I’m right about that, then putting aside the effect it may have on the gender pay gap, it’s doubtful that the fact that women are underrepresented in STEM really constitutes a disadvantage. It’s just that women don’t have the same interests as men and tend not to have the same conception of what a good life is.  As Susan Pinker argued, there is a lot of evidence that, beyond the fact that women are less interested in STEM than men, this is indeed the case. She suggests that, by assuming male-typical measures of success, feminists might actually contribute to make women feel bad about their choices or push them in directions that will prevent them from flourishing. Even if she is wrong, feminists should at least take that possibility seriously, but they don’t seem interested in doing so, as the way in which they reacted to her book shows.

Coming back to the gender pay gap, feminists could accept that, to the extent that the gender pay gap is the result of the fact that women don’t have the same occupational preferences as men, it’s not because girls and boys are socialized to conform to gender roles, yet still insist that it’s unfair. For instance, they could say that, while it’s true that women are more interested in nursing than in software engineering, it’s unfair that nurses are paid less than software engineers, because given the differences between the occupational preferences of men and women, even if they are natural, it means that men will end up being paid better on average than women. Indeed, feminists sometimes claim that it’s unfair that people in male-dominated professions are typically paid more than people in female-dominated professions, but it’s really not clear to me on what theory of justice it would be fair to pay nurses the same as software engineers. It just happens to be the case that software engineers create more wealth than nurses, so it seems fair that they should be paid more.

Now, how productive the workers in various occupations are depends in part on the relative prices of the goods and services they produce and the demand for them, which in turn depend on people’s preferences, so I guess one could try to argue that the fact that workers in male-dominated professions are more productive than workers in female-dominated professions is partly the result of the fact that people’s preferences are unduly influenced by gender stereotypes. But I honestly don’t see how one could go to show that and, in any case, one would have to actually show it and not just assert it. Short of that, one has to explain why, in the interest of closing the gender pay gap, it would be fair to pay the same amount of money people who don’t create the same amount of wealth. But again I don’t see on what conception of justice it would be fair. This would be economically very inefficient and therefore it’s likely that most people, including women, would be worse off. I don’t think the case for this view can be made without appealing to implausibly strong egalitarian assumptions. Perhaps I’m wrong, but in any case, the argument still has to be made and, to my knowledge, nobody has made it yet.

Similarly, although I will not defend this claim here, I think the evidence supports the hypothesis that, if men work longer hours on average than women, it’s not just because of sexism which means that women have to take care of a disproportionate amount of domestic tasks and child rearing, but also because women are naturally more likely to prefer spending time with their family and their friends. (If you want to get a sense of the kind of evidence I have in mind, you can read the essay by Susan Pinker I cited above.) Feminists typically disagree, but even if they agreed, they could insist that to the extent that this difference explains the gender pay gap, it’s unfair because the economy should accommodate women’s preferences better.

In a fascinating and influential paper about the gender pay gap, Claudia Goldin made a point very close to this. She showed that, if women makes less than men, it’s partly because in many occupations wage is a non-linear function of the number of hours worked. In other words, in those occupations, a person who works 60 hours a week is paid more than twice what a person who works 30 hours a week gets paid. Moreover, even if someone works the same number of hours as someone else, they are often paid more if their work schedule is more flexible and their hours are less continuous. Since women are less able and/or willing to work longer hours and have a flexible work schedule, this makes even their hourly wage inferior to that of men.

Goldin proposes that, in order to close the gender pay gap, the pay structure of the occupations where that is not yet the case be made more linear and less sensitive to the continuity of the hours worked. The problem with the proposal is that, as even Goldin notes, this is probably not possible in many professions and/or would be a source of inefficiency. Indeed, since the marginal cost of labor increases with the number of hours worked, workers will be less willing to work longer hours if the pay structure becomes more linear. Insofar as workers can’t easily be substituted to each other in some occupations, this will decrease productivity. Another possibility is that, where employers are in a position of strength, workers will be forced to work more than they would like.

Now, to be clear, it doesn’t mean that it shouldn’t be done. Indeed, as Goldin also points out, even male workers might enjoy this change. Still, given that men do not seem to mind as much as women the current pay structure, they will presumably benefit less and even women might be harmed by such a change if the negative effect it has on productivity is sufficiently large. I discussed this proposal not because I think it’s obviously unreasonable, which I don’t think it is (Goldin is a very subtle thinker and I’m sure she is aware of the issues I raised), but on the contrary because it illustrates how non-obvious some of the claims that feminists make about structural injustice are. In fact, I don’t know whether Goldin identifies as feminist, but I think her proposal is exactly the kind of things feminists should emulate. I’m not sure that we should embrace it, but unlike what feminists often do when they talk about structural injustice, she makes very precise claims grounded in original empirical research.

I discussed gender bias in politics, the gender pay gap and the underrepresentation of women in STEM at some length because I wanted to illustrate my claim that, while it’s true that, even in contemporary Western society women are systematically disadvantaged in a variety of ways, feminists often make claims about this which are not justified. Sometimes, as when they claim that female politicians are disadvantaged, the disadvantage they allege simply does not seem to exist. Other times, as when they claim that women are paid less on average than men, women really are disadvantaged in the way feminists allege,  but it’s really not clear that it has anything to do with discrimination or at least it’s not clear that it has to do with discrimination to the extent that feminists claim. There are also cases, as when they talk about the underrepresentation of women in STEM, where feminists are pointing out a real phenomenon, but whether it really constitutes a disadvantage for women is contingent on whether their account of what causes that phenomenon is correct and the evidence suggests it is not. Indeed, if women are underrepresented in STEM because they have more options which they are more interested in pursuing, then putting aside the effect it might have on the gender pay gap, it’s not clear why this fact constitutes a disadvantage at all.

Thus, to the extent that, on Frye’s account of oppression, women are supposed to be oppressed because of the ways in which feminists allege they are systematically and unfairly disadvantaged, their case for the claim that women are oppressed is much weaker than they think. Moreover, even if you nevertheless think that women are sufficiently disadvantaged to count as oppressed on Frye’s definition, there is a very good case to be made that not just women but also men are oppressed. Of course, Frye explicitly denies that, but that’s because she ignores the many ways in which men are disadvantaged. To be fair, she is not alone in this, for the claim that men face disadvantages which are different but comparable in seriousness to those faced by women is usually met with disbelief or even ridicule. Yet it’s a fact that men are systematically disadvantaged in a variety of ways and, while some of it can be explained by differences between men and women (just as we have seen that some of the disadvantages faced by women could be explained in that way), the evidence is overwhelming that wrongful discrimination also plays a role. Finally, in many cases, one can also show that women are the beneficiaries. If you go back to Frye’s definition of oppression, you will see that all the ingredients are present.

David Benatar wrote a whole book where he carefully defend the claims I just made, so I won’t try to make the case here, but I will just mention some of the ways in which men are disadvantaged. First, anti-male sexism is openly expressed in a way that would be unimaginable, at least in polite company, in the case of anti-female sexism. Indeed, for a large segment of the educated class, it has become socially acceptable to use expressions such as “old white men” and “straight white men” as terms of abuse. But this is not very important compared to the other disadvantages that men face. They are much more likely to be victim of violence, significantly less likely to be granted custody of their children in the event of a divorce, overwhelmingly more likely to have a work-related accident, much more likely to be imprisoned, commit suicide at several times the rate of women, etc. Moreover, when they express concern about these issues, they are often mocked, which adds insult to injury. Of course, neither I nor Benatar claim that if men are disadvantaged in those ways, it’s only because of wrongful discrimination against them. But wrongful discrimination nevertheless seems to play a significant role. At any rate, the case that wrongful discrimination plays a significant role is at least as strong as in the case of the disadvantages faced by women and other minorities, so feminists are not really in a position to deny it.

For instance, a recent meta-analysis found that even when controlling for a variety of legally relevant variables, men still received harsher sentences than women. Another study about gender bias in federal criminal cases in the US even found that, after controlling for a variety of legally relevant variables, men received sentences that were 63% longer than women. Now, I don’t claim that the whole effect is the result of unfair discrimination against men, for as should be clear by now it can be quite difficult to demonstrate the existence of that kind of discrimination. But this disparity is still more than 6 times as large as the anti-black bias the same author found in another paper using the same methodology. As far as I can tell, even other papers on gender bias in the criminal justice system, which typically find smaller effects, tend to find effect sizes that are similar to those found in the literature on racial bias in the criminal justice system. Given that feminists are usually committed to the view that racial bias exists in the criminal justice system, it seems difficult for them to deny that anti-male discrimination also plays a role in criminal cases.

The conclusion I want to draw from everything I have said is not that both women and men oppressed, but that Frye’s conception of oppression should be rejected, because it’s not helpful. Moreover, while I focused on Frye’s account of oppression because it’s arguably the most influential among feminists, I believe that I could say very similar things against any account of oppression on which women but not men are oppressed. Of course, it’s always possible to define “oppression” so as to directly or indirectly stipulate that women but not men are oppressed, but I don’t think any such definition would serve a useful purpose. Even if you insisted on defining “oppression” so that only women can be oppressed, it would not make the ways in which men are systematically disadvantaged any less serious than what women have to deal with. But this is precisely why we should not concede this linguistic point to feminists, because as I noted above, if “oppression” is only used to describe what women have to deal with this is exactly what people will think. (On the other hand, I’m happy to leave the word “feminism” to feminists, because as we have seen it’s rapidly becoming toxic anyway.) That being said, I repeat that I don’t say that because I think men are oppressed, which I don’t.

Men are not oppressed, but neither are women anymore, at least not in contemporary Western societies. Women are systematically and unfairly disadvantaged in some ways, but so are men and I don’t think it even makes sense to say that overall women have it worse. I don’t even know how one would go to show that without begging the question . To the extent that it’s possible, we should try to eliminate wrongful discrimination against both men and women, but we should do so in a responsible manner. We should only make claims that are supported by the evidence. Before we propose to address the injustices that we have identified, we should make sure that the policies we propose will actually help, instead of making things worse. We should refrain from stretching the ordinary meaning of words for political gain. On all these counts, I think feminists usually fall short.

NOTE: This post is a draft of a paper I was asked to write for a volume published by Oxford University Press and edited by Bob Fischer from Texas State University. It’s called Ethics, Left and Right: The Moral Issues That Divide Us. It’s a collection of essays on social/ethical issues organized so as to encourage debate between left and right. On each issue, one essay is written from a left-wing perspective and the other from a right-wing perspective, then each contributor briefly responds to the person opposite them on the issue they wrote about. If you read this blog regularly, you won’t be surprised to know that I think it’s a great idea, which is why I was glad that Bob asked me to contribute. The version of the paper I end up publishing in the volume will probably be very different from this post, if only because in order to respect the space limitations, I will have to cut it roughly in half. Moreover, I wrote this in just a few days on my spare time and it’s still very much a draft, so I’m sure the final product will be very different.