Could single parent vs two parent households be a variable to look at? Most single-parent households are headed by women with no male parenting counterpart. If there is this big gendered shift, maybe it's stronger in single female parent households than in two-gender pair of parents households.
My theory is this has to do with the decline of marriage. The theory behind this being that men and women who commit to one another in marriage are more likely to have their political views converge over time. Apparently there are studies showing that women who have children become more conservative, so perhaps a decline in fertility is related in some way too. I'd be interested to see if there's anything to any of that in the data.
I would say that Burn-Murdoch's observation is correct, but that Paul Graham's explanation of it is not. Rather, there are cultural and memetic factors at work that are not chiefly about association.
Not sold on the hypothesis, but question is whether siblings "hang out." Personal experience says no. It's rare for siblings to hang out together socially, especially opposite sex.
Might help a little to throw age delta between siblings in, but probably not available?
Fairly straightforward explanation is that the left/right spectrum has become less about economic equality vs neoliberal economics, and more about "social justice" vs "traditional values". The "right" champions traditional gender roles (good for men), and the "left" champions egalitarianism (good for women/minorities).
If people follow their own interests selfishly, this is exactly what you would expect: more focus on gender roles between left and right leads to a harder gender split between left and right.
My vibes say: women used to assimilate into a relatively narrow mass culture controlled by men. Now there are so many more drivers of mass culture who are women, so the assimilation is less pronounced. This explains why the political attitude shift is so much stronger for women than for men.
No one cares what their siblings think. But peers that you're attracted to? You care what those people think, and might be willing to listen to their position / change your own in that scenario.
Something that’s not clear to me from that chart is whether it’s measuring the beliefs held by women/men, or just which group they say they belong to. I would be interested to know if the differences hold for particular policy positions. It seems plausible that men might be more likely to say they are conservative while holding many liberal beliefs, or for women to say they are liberal while holding many conservative beliefs.
Hi! First time commenting. The best explanation for your troubles replicating the graph might have to do with the original graph itself. Noah smith has a good investigation: https://www.noahpinion.blog/p/at-least-five-interesting-things-d7e (see number 2). One problem is that the graph seems to switch between two datasets at some point, and the divergence doesn't replicate in a lot of other datasets. So it might just not be a very large effect. Also, if we are sold on there being a lot more far left women than men, it is good to remember that in the context of normal distributions, small differences in means can lead to large differences in extremes.
But I think there's a ton of uncertainty around this question (how many people under thirty really know much about their own political ideology?), and this graph doesn't seem to shed a whole lot of valuable light.
As a point of reference, what is your definition for "gender"?
The problem is that pretty much every last man, woman, and otherkin has their own idiosyncratic and often contradictory definitions for the term. My efforts to put the concept on a more scientific footing:
This is great stuff, I'm mildly embarrassed I didn't think to bounce it off the ACX survey before now. Obv plenty of ways to make it more rigorous, but a good tool for a first gut-check.
Siblings would be a problematic way to judge this on its own terms. Much of the segregation-to-radicalization claims in Korea carry the implication that decreasing intersexual socialization *while the other is the object of each's sexual desire* makes the alien gender (and critiques of their gender-culture) a natural scapegoat for personal frustrations and discontent.
I think siblings are just too early to make a difference one way or the other. A lot of the """development""" of political beliefs happens in College, when kids are separated from family/siblings for the first extended period of time. I don't think this proves or disproves the hypothesis
Does anyone have the survey question used for lib/conservative in the cooperative election study?
I run across a decent amount of young self identified liberals who will call themselves right wing, but not conservative; and vote GOP only as a matter of contempt for the other side.
It seems reasonable that siblings might not be moderating. You might only moderate your views for people you view as a romantic prospect, or siblings may simply not “hang out” in such a way that moderates views.
I personally think Graham’s theory is debatable and probably wrong but not for the reasons given here
"It just suggests it doesn’t extend to siblings. Which would, admittedly, be weird."
But why? Relationships with opposite-sex siblings are different across the metric of "how likely I am to adopt their political views" than social ones with peers. I had a younger (3 years) sister, we were pretty antagonistic as kids. I often adopted positions (it is morally just that I get to watch the TV right now) purely in opposition to her. We have relatively similar politics now, as 39 and 42 year old adults, but 25 years ago if she believed X, there would've been at least a small movement for me to believe not-X.
- Does the survey data show the general trend in the chart that young women are more left than young men, and that this hasn't always been the case? From a quick analysis I did it seems like this is the case. Young men are more left-leaning than older men in the sample but young women are still much more left-leaning
- Your stats are for the whole survey group not just the young people. However, I redid the analysis (again quickly) on only the 30 and younger respondents and you still don't reject the null with much confidence
Again my analysis was done rather quickly so it's possible I made mistakes.
This gender gap seems like exactly the kind of clickbaity result that would get spread around by journalists and social media regardless of whether the paper it's from had any merit. Is there any reason to think this one in particular is worth taking seriously?
The disparity is deeply historic and physical, with women typically left holding the bag for male expansionism. Nursing the injured during war, caring for the vulnerable children, preparing food, experiencing childbirth, etc. Adventurism relies more on "male" detachment so that material, earthly, and bodily concerns are abstracted sufficiently to support bold action favoring the strong; whereas "females" (just like today's liberals) are left to clean up the mess and support anyone who the strong considers weak.
Yes, of course, this platform's statistics are far from significant. A bit like polling House GOP members, whose females somehow wound up like frat boys in Animal House (see MTG and LB).
Given that this is a community with a lot of contrarians and people on the spectrum, ACX readers' political views might be less influenced by their IRL peers than most people.
This isn't as simple to test as it may seem. I think you need to (at least) run a regression with some controls. One obvious control would be for the total number of siblings. Larger families are less likely to have all children of the same gender. Larger families are also different in income, education, race, culture, etc., which are all correlated with political ideology.
I don't think my sister and I shaped each other's political views at all, because by the time I was old enough to have political views she was out of the house. We have similar views, but I think that's our parents' influence.
On the other hand, her husband underwent a pretty major political change, largely because of her. And I've heard that Kids These Days don't date as much as we did, so maybe that has something to do with it?
If I recall correctly, there is evidence indicates that sex roles generally get more different as a country gets richer. Ie women in Norway are more likely to work in stereotypically female jobs than women in poorer countries. The usual explanation is that being richer allows women to indulge their preferences – in poorer countries women have to work at the highest paying job they can find, which are often stereotypically male jobs, whereas in richer countries women can afford to work in jobs they prefer. My theory is that divergence in political views is that the same phenomenon at the political level.
In the past 10 years, more young men in the US are having no sex, while the trend for women is much flatter. Growing conservatism and resentment among young men could be driven by increased female selectiveness in the era of dating apps and a resulting feeling among men of rejection from society. As women exceed men in college attendance and career success rates, they need men less and men tangibly feel this loss of status and access. Check out this chart--
I would think that in the US abortion would be a strong force for this. I don't have hard data to back that hypothesis up, but it does feel like a sort of obvious explanation
I have a hypothesis that this could be due to an increasing role of LGBT issues in political division. I suspect that men have a greater aversion to LGBT issues, whether innate or learnt. The timing of this is about right (corresponding to a rise sometime around the 2010s), and pew research center polls do record that women are more likely to be accepting than men of homosexuality. However, I don't think the magnitude of the effect can be accounted for by the pew polls - in the UK for example, ~80% of men and ~90% of women accept homosexuality. One possibility is that men may be accepting of homosexuality, but the political lines are drawn by other LGBT issues, for instance trans issues.
Siblings have zero effect for men, either young or old. But interestingly, young women with only sisters are *more conservative* than young women with only brothers.
And for older women, the effect is *exactly reversed*.
I was about to say: the decline in organized, traditional religion is a factor in this, particularly in Catholic countries and in countries that have large Catholic minorities. (Catholicism is an arguably male chauvinist church that - in many spcieties - appeals predominantly to women.) Maybe that's less of a factor in the US, where liberal Catholicism (however defined) is numerically relevant.
But that is just a factor that made very many women loyal to the hard right (well, again, less so in the US - and in Germany, it would be the center-right; well, I'd say CSU is hard right, but not far right). A different explanation is needed for men' rightwards shift.
(If liberalism/leftism were all about the loudest online woke folk, and conservative about Trad lunacy, then it would be clear: why would educated women be averse to the "women belong pregnant and barefoot in the kitchen" camp, and even mildly self-aware men averse to "stfu men, none you have opinions, only male opinions, get out of here and give up your jobs, also, male privilege" camp? )
One obvious difference between sibs and classmates is that most people aren't all that interested in impressing their siblings - in fact, they may just want to troll them instead. I think there are plenty of young men, for instance, who decide to moderate their views on abortion after hearing how zealous their dating partners' views on the issue are, but nobody cares that much about impressing their sister.
It's also possible that the shared effects of parental political beliefs outweigh the sibling effect somehow - the moderation effect may be more important when people differ widely in beliefs.
I have looked for evidence of these trends in data from the General Social Survey and the American National Election Studies. So far I have not found a growing gender gap in political alignment, party affiliation, or policy preferences.
The critical theory 'woke' ideological complex is hinged on Care-Harm moral values. This is really attractive to women. Some suggest almost necessary for some women who have kids later and later if at all and need a substitute
I think the increasing majority of women in college, increasingly in the social sciences and humanities, are adopting this belief and moral system. I am quite surprised that men have barely budged but it does fit that concept
Paul Graham's wider political takes are generally not very good outside of stuff he has direct experience with (bureaucracy, taxes). And even there he's not hugely insightful.
With the exception of South Korea, all three cases are men mostly varying within their normal range and women getting extremely liberal. The failure to replicate is also a non-sequitur: the first chart is about supporting conservative or liberal parties or self-identification. The debunking is about a battery of ideological questions. And it shows the same phenomenon: while men have gotten somewhat more liberal women have moved further left.
That's the phenomenon and so any explanation needs to focus on why women are getting more politically radical and becoming less likely to self-identify as conservatives. It is a change and one that's been ongoing for some time (before the internet age). MR at least explains the correct phenomenon with a plausible mechanism. But many people (like Paul Graham) are avoiding seeing this as something driven by women or a longer run trend. Others are essentializing it, saying women or men have always been X or Y, when there is a change that needs explaining.
South Korea meanwhile is caught up in East Asia's gender wars and also has its own, relatively intense versions. While it might be partly driven by some global trend there's also a specific regional trend that doesn't exist outside of the region.
Everybody seems to be just skipping past the "does not replicate" part...the Cooperative Election Study is a huge sampling collected carefully and consistently for nearly 20 years now. If a new divergence is real there needs to be some explanation for it not showing up there even a tiny bit.
Used to take PG's theories seriously until I learned of his horrible understanding of moral equivalence with regards to the Hamas-Israel war. A techno capitalist should not side with terrorists.
I very much dont agree with my siblings about politics, I'd bet family is an extremely bad predictor of politics for social media generations; maybe in an era where family was 2nd place(in thoery?) for information compared to tv news that all agreed; now people don't even share facts; my siblings least of all(they told me to stop sending them political videos)
Personally, as a Zoomer male, I think large segments of the left dislike us. You see constant complaints of the "cishet white man" and such from many of their circles. They're not exactly making it difficult for us to feel like they're the 'enemy' so to speak.
I know this isn't true of all leftists groups, but the fact that some parts of the edifice aren't completely rotten does not really make me any less keen on the whole thing coming down; not to mention that the seemingly-innocuous groups seem to have been the precursors to the rotten ones.
The above is a good video on the topic. So until the late 20th century men used to more left wing than women. Also when polled on individual issues the extent of polarisation is much lower than the initial data point would suggest. So the political positions of liberal and conservatives voters might be smaller than people expect.
Surely this is due to gender being a tool applied for the purposes of identity politics - gender is being politicized by social media algorithms and bad faith pundits and political actors.
The right have abortion bans, Trump (a rapist), anti-trans legislation, "traditional values" where women are housewives etc. The left have cancel culture, "believe victims", and vocally supporting every group except men (especially white ones).
I can't believe so many people are confused about this. It's clearly #MeToo-style feminism. (#MeToo, of course, started in 2017. But the variety of feminism that it epitomizes is much older, and was around in the aughts.)
Does no one remember the genderwars of the early 2010s? These issues aren't as discussed anymore, but that isn't because the underlying causes behind them went away.
“Boys and girls used to hang out together more” seems glaringly false to me. Things used to be WAY more gender-segregated, but have gotten more coed over time. In the past 10 years alone, my college dorm decreased its number of single-gender floors by 75%, as well as allowing opt-in ability to have opposite gender roommates in doubles. Coed bathrooms have also drastically increased in number (mostly due to the prevalence of nb people now), and while no one really hangs out in bathrooms, I think it’s a sign of just how many other things went coed first.
More anecdotally, everyone in my social group has opposite gender friends nowadays, while I know that my dad had basically never talked to women even by the time he’d graduated. Everyone I know also seems happy to live with opposite gender roommates in their apartments, while this would have been unpopular a few decades ago and outright scandalous a few decades before that.
I think a much simpler explanation for increasing political differences is that liberal politics have gotten more focused on social justice (which is often exclusive towards men) and conservative politics have become more focused on the pro-life side of the abortion debate (which is repellent to many women for obvious reasons).
As a sanity check, perhaps you could sample from liberal and conservative political groups and see if they’re male or female, rather than sampling men and women and asking if they’re liberal or conservative, just to see if the two approaches are roughly comparable to each other. Of course, you’d have to be able to account for age.
Part of this has to be a function of education, right? People with graduate degrees have very liberal views, generally speaking, and more and more such degrees are going to women (I think both in raw numerical terms and the proportion of such degrees that are awarded).
What do conservative and liberal even mean in this context? If we divide each category further with social and economic qualifiers would the resulting 4 groups relate to each other in the same way? What if the terms feminine and masculine were used in place of liberal and conservative, would the results be similar?
To me the most confounding part of anything like this is in a respondent's self-reporting what broad cohort they should be placed within. I think in truth this detail is situational and malleable, and is always a matter of relative perspective.
The problem with sampling ACX readers on this is not where they fall on the political spectrum, but WHY they fall where they did. Your readers seem generally more likely to be the sort of people who arrived at their political opinions through some kind of intentional deliberation (or at least formulated a defensible post hoc rationale for the ones they already have.) I suspect your readers are meaningfully less susceptible to changing positions merely to get along socially, and when they sometimes have to pretend to believe in X because of their peer group they're more likely to be able to compartmentalize that and maintain their true belief in not-X which is reported in private surveys.
If I was dating a gal in the Sierra Club and had to feign sympathy towards some useless endangered species, I could probably do so and still maintain my true belief in a basically libertarian property rights approach and accurately report that on a survey. But most people, if they say something enough times, start to inch towards believing it. I suggest that the kind of folks who read your blog and those like it are arrogant about our intellect and reasoning ability, think we're smarter than most people we meet on a regular basis, and are therefore less likely to be influenced by the herd effect being discussed in those tweets.
...Yin in the ascendency, always brings chaos, thus the growing disequilibrium between men and women in the everyday, not just in the political. Pursuing kindness at the societal scale, is a path paved by a good intention, but as we are seeing across the West, compassion before justice, in the particular and at large, impoverishes all. The quiet part, but getting louder, is that new wave feminism/progressivism is the witchcraft of our age. (If you want some numbers to tell you something Scott, check out how the original witchcraft wave played out in the U.S. and Europe)
I wonder if young men are really conservative or if they are just opposed to wokism, which they feel is hurting them. I would expect them to be much more open to traditional leftist positions like tax the rich, distribute wealth, social security for everyone, limit corporate greed and similar.
My 2 cents: because politics has become less important, it's now fashion; and men and women have radically different fashion choices.
Politics used to be life or death for lots of people. Gay people's ability to marry people literally hung on which party was elected in their locality. And that's only the most recent example. Women's rights, policies that affected non-white people, all that stuff. These days, the gap between right and left is smaller. Example: even Donald Trump, who was incompetent and terrible in so many ways, failed to derail America. The Tories, who are pure expressions of self-serving abuse of power, have not managed to destroy the UK (I really thought they might). Right and left leaning parties bring only marginal levels of change.
Therefore, people realise that their lives and livelihoods are not at stake, and vote on other factors - essentially aesthetic or fashion factors. And right now, the fashion is swinging in one particular direction. It could easily change course, and make women more right wing.
This feels like one of those ideas that's just so salient that it's dangerously difficult to avoid immediately jumping on a pet hypothesis that obviously explains it.
Even after it doesn't replicate, and thus shouldn't need any explanation.
I think that increasing time online versus interacting in person are at the heart of this, like so many other things.
Think of it this way, a brother and sister may occupy the same house, but their minds may be a thousand miles apart thanks to the Internet. If one spends all evening on tumblr and the other on 4chan (1) occupying the same household may make less of a difference than it would have in 1980.
An interesting question that could start to test this would be to look at the both gender difference in specific site usage and how boys and girls on the same site differ in the politics. Are boys and girls and tumblr more similar in political leanings than boys and girls in total? Are girls on tumblr closer in affiliation to girls on Twitter than to boys on tumblr? Are intersite political differences greater or less than intrasite political differences? I hypothesize yes in some since to all of these, but I'd be curious about empirical data.
(1) Or the Zoomer equivalents. I admit, I'm almost to the age where I'm not sure where kids hang out online these days. Tiktok?
My view is that this is mostly untrue and the non-replications are correct, at least in the US; young women are more liberal than young women 10 years ago, but young men are also more liberal than young men 10 years ago (I'm willing to believe that there's a "divergence" in the sense that the change among young women has been larger, but the design of the graphs and the agenda of the discussion depends on today's young men being increasingly conservative).
I might be willing to buy young men being more willing to *call themselves* conservative conditional on having a given set of political opinions, but that is more about vibes. Maybe also the political identity spectrum increasingly including the "trust the credentialed" vs. "one weird trick" spectrum, which has probably always had a big gender gap.
One thing I tried here was looking at cisgender Americans who selected more moderate responses (not "1" or "5") on the immigration, abortion, minimum wage, and global warming issues. In all four cases (and in the smaller sample size for all four combined) men identified as a bit more conservative on the political spectrum question, though I did not look for statistical significance.
Your blog is where I learned that all politics is signalling, so let me return the favour: this is signalling, and it's pretty funny signalling at that.
If you're a woman, you want people to think that you're caring/educated/cosmopolitan. Liberalism (which is a consumer product marketed as an identity like everything else) sells itself as caring/educated/cosmopolitan,* and has reoriented itself in response to market forces to fill that niche. Your opinions don't matter for non-signalling purposes, so you loudly proclaim yourself to be liberal.
If you're a man, you want people to think that you're tough/self-sufficient/unresponsive-to-social-pressure. Conservatism (which is a consumer product marketed as an identity like everything else) sells itself as tough/valuing self-sufficiency/standing-up-to-social-pressure. Your opinions don't matter for non-signalling purposes, so you loudly proclaim yourself to be a conservative.
My prediction would be that legacy products with skewed branding will fit this pattern; e.g. tankie Communism will be more male than female.
If this has grown over time, it's because politics as brands to express your identity has been finally finishing off that weird 19th century thing where you think about what policies would be good for the country and vote for someone who says they'll enact them. Also, development/liberalism increase gender dimorphism in personality (presumably also for signalling-competition reasons), so that would lead some of the increase too.
*Also rural. You can squint and tilt your head to say environmentalism does this, but I think that's too much of a stretch and it's just the caring thing again. The party that combines liberalism with "voting for us proves you grew up on a farm" will be guaranteed 50% of the vote though.
For the ACX survey, you could check married people vs single. I think I was more influenced by my wife's politics vs any sibling. (I guess this is age dependent.)
I think this is obvious - the left has becoming increasingly misandrist and the excessive focus on DEI and growing dislike of men, especially white men, has started to alienate and exhaust males. Therefore they are less likely to be on board with liberalism. It is well known that feminism, originally a respectable ideology, has spiraled out of control and many of its proponents are now woke idiots that hate men.
Women who are not your sister are able to use additional incentives/rewards to shape male behavior (should they choose to accept the mission). It's unsurprising that they would do so, since unrelated males can be dangerous. It's also unsurprising that they tend to be more liberal as liberal ideology is more empowering to women, and at least tries to moderate male dominance.
I feel like it's less about gender integration (amount of time men and women spend together) and more about a divergence of where men and women get cultural information. It would be interesting to see how this holds up in say, single-sex private schools before social media, where boys and girls are mostly separated but are basically in the same cultural bubble.
There have always been gendered hobbies, skills, and occupations, but the past fifteen years have been historically weird with how much new, niche information can be accessed. Single women (who are doing the most political diverging) generally aren't coming up with these new ideologies themselves from hanging out with other women, but ideas are spreading in majority-female spaces that otherwise wouldn't have the chance to spread. The success rate of specific ideas is different in majority-male vs majority-female spaces, and with the vast magnitude of new ideas that get a chance to take root, this difference is accentuated.
Whereas for most of history, most people are coming up with new ideas from information they get from their family/spouse and their friends who are also in a small, local network. Coming across twenty hot takes a day from people that you've never met is pretty unprecedented.
Scott, I was thinking of re-doing the analysis using a non-parametric test, which I think is what's appropriate for data of this kind. But maybe you already did that? Did you do a t-test (parametric) or a Chi-square test (non-parametric)?
So what's interesting is that when you look at support for particularly policies there is much less discrepancy. My theory is that the difference here is largely one driven by a greater role of personal value signalling and different aesthetics about political talk in our political discussions (likely driven by social media). Basically, the theory is the way of talking about politics each side has puts the other side off -- and that these strongly correlate with male vs female differences in communication and interpersonal advertising.
I mean, I believe global warming is a serious concern and that we need to worry about how environmental degradation can harm people (eg haze in SE Asia) but when I hear the way many greens talk about how we are hurting mother earth with this vibe that seems to resent progress and our mastery over nature my emotional reaction is absolutely -- fuck that I'd rather let the earth burn than be on the same side as them.
This happens in a hundred different ways on a hundred different issues. It didn't happen before because politics was a broadcast message. The political classes talked to you (and they worked hard not to alienate anyone) but now we talk to each other and the differences between male and female communication styles and value advertising are driving us crazy.
The piece is great to show some tendencies, though it compares apples with pears - still fruits, but not really the same. The US data shows the support for ideology, while the rest is the support of a certain party. Since there are no Gallups in other countries, this is the best we have, though it does not account to how true to their ideology those parties are. In Romania, the social-democratic party cares little about social democracy, yet it will be classified as a left-wing party there.
The cultural aspect is important in understanding a rise in conservatism. Coming from Romania, it is impossible not to hear Andrew Tate's name on a weekly basis. What has he done for society? Absolutely nothing, turned a bunch of young boys into self-obsessed mean that speak fast, drink sparkling water and have an issue with fat, liberal and "woke" people. Tracing all the problems of the world to those groups is easy and requires little effort to investigate the world, thus a growth in those conservative thoughts.
My guess would be siblings don't moderate each others' political views much because they start out too similar, on account of having the same parents. Look for people who share a house, but didn't grow up together, and didn't have much chance to pre-screen prospective housemates based on political views.
It's not that it is or isn't a problem, but as with anything trends are interesting and it's a worthwhile exercise to try and understand why things are playing out the way they are.
Another thing worth noting is the established trend that male and female professions tend to diverge more as gender roles become weaker in society (this can be observed in international comparisons). This could be a reflection of the same underlying dynamics, men and women increasingly self-sorting into different tribes as they have the freedom to do so. This would imply that sexual differences are strong and persistent and naturally carry over into various areas of life if given the freedom and opportunity to do so.
It's also somewhat paradoxical in the case of politics though, since historically women have tended to be more right wing and men more left wing. But this just underscores the fact that the tribalism is itself the driver of the differences and not the specific orthodoxies of the tribes that they've self-sorted into.
Everybody (left and right) wants to make it about their personal gender war hobby horse, but it seems more likely it's due to demographic shifts, such as those mentioned by Marginal Revolution. Women historically tended to be more right wing than men, since they were less likely to be members of unions and were more religious; as women became near-equal contributors to the work force and secularized, these influences disappeared and they became more left-wing than men. If in the past such trends have been primarily demographic, why would we expect now they'd be based on gender war stuff?
If this doesn't replicate and may not exist... why elevate the topic to a discussion at all? It seems like the very stereotype of "angels dancing on the head of a pin" style discussion that generates no knowledge or information if you're merely speculating about what could cause an effect that doesn't exist.
Can anyone find the Washington Post-ABC News poll that found people with daughters were more likely to vote for Hillary in 2016?
As I recall the poll actually showed that people with children of mixed sexes were about as likely to vote Trump as people with only boys, whereas people with only daughters were more likely to vote Trump, so the actual finding was that having at least one son made you more likely to vote Trump. The poll appears to be paywalled, but Huffpo of all places has the finding:
My interpretation was the presence of at least one son means you would have been worried about the effects of feminism on your son if Hillary won, but I am obviously not a neutral observer here...
Why omit people with both brothers and sisters? Seems like the important part is having at least one opposite-sex sibling; not having a same-sex sibling doesn't seem important.
(I say this as a male with both a brother and sister.)
Every failed hypothesis highlights the incredible ability the human brain has to concoct a plausible-sounding hypothesis to literally ANY set of data, and that we shouldn't accept any hypothesis until we have some supporting data (and NOT what was available before the hypothesis was generated).
My personal experience is that I (M) have two sisters and a wife who are all significantly farther to the right than I am. Not that my anecdote is statistically meaningful, so much as that my personal experience leads me to be skeptical of the hypothesis sans compelling evidence.
Meanwhile, is there a more broad assessment of country trends? I'm skeptical this is just something happening in these four countries, and if we did a true international survey we'd find no general trend, just a lot of noise that only looks like 'signal' when you cherry pick the data. The failure to replicate suggests this isn't a robust observation on which to base any hypotheses. Once again, the prudence of starting any discussion of anomalous data with, "But what about the null hypothesis?" suggests itself.
I am much less concerned about my sisters approval than I am of a potential partner. Eventhough I start getting into cooky political stuff my sister will still be my sister. However my long time partner might no longer be my partner
Presumably siblings are comfortable with each other, and there is no need to impress the opposite sex by exhibiting their approved behaviours. In other words, stranger girls and boys might rub off on one another much more than siblings.
I hate to say it, but I think an ACX sample is not very useful for assessing the USA or the rich world as a whole.
You can see from his sample it's got almost 10 times as many men as women (a big problem if you're studying gender divergence), and is composed heavily of younger men with technical interests. Given the history of the blog involves a doxxing attack on its author by a major arm of the liberal media, it tends to lean center-right.
I think it's a great sample for looking at attitudes among techies (who are a culturally important segment given the importance of computers and the internet these days), but not much more than that.
I'm a young adult and still live with my siblings, and have fairly different political views from them. I think being in the same family and interacting with each other moderates us a *bit*, but our internet siloes are still very different and it generally doesn't feel worth it to work through our different worldviews built out of watching hundreds of hours of video essays and podcasts leading in different directions. The number of different assumptions and logical steps built on other steps just seems overwhelming.
to me the obvious explanation is that news feed algorithms and balkanized media amplify initially small differences, ala James Flynn's basketball argument. Media used to be more ideologically integrated when people were reading local newspapers and watching local TV channels that tried to serve everyone in a particular geographic area. Now media is all fine tuned to pander to the biases of particular ideological niche.
Politics in general has polarised, which seems to be something not solely correlated to gender. If politics polarises, and there's already a gap between men and women on average, maybe the gap widens as a result of the landscape moving rather than a change in gender relationships?
I wonder how the gap would look when tested for individual ideas, rather than as a bundle. E.g., separating out women's rights from issues like tax burden, would be interesting. It's possible that answers around reproductive rights skews things, though I don't have a convincing argument for why the gap around those in particular would have changed.
Nope, not Chinese. Not American, just British colonial ancestry.
I think the crackdown in China is more about cultural progressivism than consumerism/capitalism. In fact in economic management terms the Chinese are actively trying to redirect from exports, to internal consumption. But the Chinese are savers, ( because there is no social welfare) so it's not working very well. The same Chinese prof I mentioned before, was also happy to explain the chaos of 1960's 'cultural revolution' as very much a shift in the Yin. ( It was Mao's wife, Jiang Qing who led the 'Gang of four').
As for other episodes of mass mania, check out Richard Landes' 2010 book, "Heaven on Earth-The varieties of the Millennial Experience". The chapters on the tragedy of the Xhosa people in the mid-19th century is especially relevant to the Greta Thunberg phenomenon.
Could single parent vs two parent households be a variable to look at? Most single-parent households are headed by women with no male parenting counterpart. If there is this big gendered shift, maybe it's stronger in single female parent households than in two-gender pair of parents households.
I wonder if a lack of dating and increasing singlehood is contributing to these numbers.
My theory is this has to do with the decline of marriage. The theory behind this being that men and women who commit to one another in marriage are more likely to have their political views converge over time. Apparently there are studies showing that women who have children become more conservative, so perhaps a decline in fertility is related in some way too. I'd be interested to see if there's anything to any of that in the data.
I would say that Burn-Murdoch's observation is correct, but that Paul Graham's explanation of it is not. Rather, there are cultural and memetic factors at work that are not chiefly about association.
The left gives preferential treatment to one of these groups over the other. Of course support diverges
Not sold on the hypothesis, but question is whether siblings "hang out." Personal experience says no. It's rare for siblings to hang out together socially, especially opposite sex.
Might help a little to throw age delta between siblings in, but probably not available?
Fairly straightforward explanation is that the left/right spectrum has become less about economic equality vs neoliberal economics, and more about "social justice" vs "traditional values". The "right" champions traditional gender roles (good for men), and the "left" champions egalitarianism (good for women/minorities).
If people follow their own interests selfishly, this is exactly what you would expect: more focus on gender roles between left and right leads to a harder gender split between left and right.
My vibes say: women used to assimilate into a relatively narrow mass culture controlled by men. Now there are so many more drivers of mass culture who are women, so the assimilation is less pronounced. This explains why the political attitude shift is so much stronger for women than for men.
No one cares what their siblings think. But peers that you're attracted to? You care what those people think, and might be willing to listen to their position / change your own in that scenario.
Wasn't that supposed widening ideological divide disproven? I can't remember where I saw that, it might have been on Zvi's substack.
Something that’s not clear to me from that chart is whether it’s measuring the beliefs held by women/men, or just which group they say they belong to. I would be interested to know if the differences hold for particular policy positions. It seems plausible that men might be more likely to say they are conservative while holding many liberal beliefs, or for women to say they are liberal while holding many conservative beliefs.
Hi! First time commenting. The best explanation for your troubles replicating the graph might have to do with the original graph itself. Noah smith has a good investigation: https://www.noahpinion.blog/p/at-least-five-interesting-things-d7e (see number 2). One problem is that the graph seems to switch between two datasets at some point, and the divergence doesn't replicate in a lot of other datasets. So it might just not be a very large effect. Also, if we are sold on there being a lot more far left women than men, it is good to remember that in the context of normal distributions, small differences in means can lead to large differences in extremes.
But I think there's a ton of uncertainty around this question (how many people under thirty really know much about their own political ideology?), and this graph doesn't seem to shed a whole lot of valuable light.
As a point of reference, what is your definition for "gender"?
The problem is that pretty much every last man, woman, and otherkin has their own idiosyncratic and often contradictory definitions for the term. My efforts to put the concept on a more scientific footing:
"A Multi-Dimensional Gender Spectrum"
https://humanuseofhumanbeings.substack.com/p/a-multi-dimensional-gender-spectrum
Political scientist John Sides was not impressed with the viral graphs; there might not be much to explain:
https://goodauthority.org/news/maybe-young-men-and-women-arent-so-ideologically-different/
This is great stuff, I'm mildly embarrassed I didn't think to bounce it off the ACX survey before now. Obv plenty of ways to make it more rigorous, but a good tool for a first gut-check.
Siblings would be a problematic way to judge this on its own terms. Much of the segregation-to-radicalization claims in Korea carry the implication that decreasing intersexual socialization *while the other is the object of each's sexual desire* makes the alien gender (and critiques of their gender-culture) a natural scapegoat for personal frustrations and discontent.
I think siblings are just too early to make a difference one way or the other. A lot of the """development""" of political beliefs happens in College, when kids are separated from family/siblings for the first extended period of time. I don't think this proves or disproves the hypothesis
Some other artifact that doesn’t show in polls:
Does anyone have the survey question used for lib/conservative in the cooperative election study?
I run across a decent amount of young self identified liberals who will call themselves right wing, but not conservative; and vote GOP only as a matter of contempt for the other side.
The observation is that women and men have different politics and the proposed mechanism is that women and men consume different media.
I think using ACX survey data won't work, not because it selects for certain politics, but because it selects for certain forms of media consumption.
Maybe you should ask about the use of Instagram and First Person Shooters in the next survey, just to make sure.
It seems reasonable that siblings might not be moderating. You might only moderate your views for people you view as a romantic prospect, or siblings may simply not “hang out” in such a way that moderates views.
I personally think Graham’s theory is debatable and probably wrong but not for the reasons given here
"It just suggests it doesn’t extend to siblings. Which would, admittedly, be weird."
But why? Relationships with opposite-sex siblings are different across the metric of "how likely I am to adopt their political views" than social ones with peers. I had a younger (3 years) sister, we were pretty antagonistic as kids. I often adopted positions (it is morally just that I get to watch the TV right now) purely in opposition to her. We have relatively similar politics now, as 39 and 42 year old adults, but 25 years ago if she believed X, there would've been at least a small movement for me to believe not-X.
Some things I thought of while reading
- Does the survey data show the general trend in the chart that young women are more left than young men, and that this hasn't always been the case? From a quick analysis I did it seems like this is the case. Young men are more left-leaning than older men in the sample but young women are still much more left-leaning
- Your stats are for the whole survey group not just the young people. However, I redid the analysis (again quickly) on only the 30 and younger respondents and you still don't reject the null with much confidence
Again my analysis was done rather quickly so it's possible I made mistakes.
This gender gap seems like exactly the kind of clickbaity result that would get spread around by journalists and social media regardless of whether the paper it's from had any merit. Is there any reason to think this one in particular is worth taking seriously?
The disparity is deeply historic and physical, with women typically left holding the bag for male expansionism. Nursing the injured during war, caring for the vulnerable children, preparing food, experiencing childbirth, etc. Adventurism relies more on "male" detachment so that material, earthly, and bodily concerns are abstracted sufficiently to support bold action favoring the strong; whereas "females" (just like today's liberals) are left to clean up the mess and support anyone who the strong considers weak.
Yes, of course, this platform's statistics are far from significant. A bit like polling House GOP members, whose females somehow wound up like frat boys in Animal House (see MTG and LB).
Given that this is a community with a lot of contrarians and people on the spectrum, ACX readers' political views might be less influenced by their IRL peers than most people.
This isn't as simple to test as it may seem. I think you need to (at least) run a regression with some controls. One obvious control would be for the total number of siblings. Larger families are less likely to have all children of the same gender. Larger families are also different in income, education, race, culture, etc., which are all correlated with political ideology.
Also see: https://www.pewresearch.org/short-reads/2013/06/27/why-boys-with-sisters-are-more-likely-to-be-republicans/
I don't think my sister and I shaped each other's political views at all, because by the time I was old enough to have political views she was out of the house. We have similar views, but I think that's our parents' influence.
On the other hand, her husband underwent a pretty major political change, largely because of her. And I've heard that Kids These Days don't date as much as we did, so maybe that has something to do with it?
There's tremendous social pressure for young women to dislike Trump
What about men/women with no siblings?
If I recall correctly, there is evidence indicates that sex roles generally get more different as a country gets richer. Ie women in Norway are more likely to work in stereotypically female jobs than women in poorer countries. The usual explanation is that being richer allows women to indulge their preferences – in poorer countries women have to work at the highest paying job they can find, which are often stereotypically male jobs, whereas in richer countries women can afford to work in jobs they prefer. My theory is that divergence in political views is that the same phenomenon at the political level.
In the past 10 years, more young men in the US are having no sex, while the trend for women is much flatter. Growing conservatism and resentment among young men could be driven by increased female selectiveness in the era of dating apps and a resulting feeling among men of rejection from society. As women exceed men in college attendance and career success rates, they need men less and men tangibly feel this loss of status and access. Check out this chart--
https://www.knowledgeformen.com/wp-content/uploads/2023/10/young-men-driving-the-decline-in-sex.png
I would think that in the US abortion would be a strong force for this. I don't have hard data to back that hypothesis up, but it does feel like a sort of obvious explanation
I have a hypothesis that this could be due to an increasing role of LGBT issues in political division. I suspect that men have a greater aversion to LGBT issues, whether innate or learnt. The timing of this is about right (corresponding to a rise sometime around the 2010s), and pew research center polls do record that women are more likely to be accepting than men of homosexuality. However, I don't think the magnitude of the effect can be accounted for by the pew polls - in the UK for example, ~80% of men and ~90% of women accept homosexuality. One possibility is that men may be accepting of homosexuality, but the political lines are drawn by other LGBT issues, for instance trans issues.
I ran it broken down by age -- which is crucial to the initial hypothesis, fwiw.
https://kmunger.github.io/im/acx.png
Siblings have zero effect for men, either young or old. But interestingly, young women with only sisters are *more conservative* than young women with only brothers.
And for older women, the effect is *exactly reversed*.
(US only, cisgender only)
I was about to say: the decline in organized, traditional religion is a factor in this, particularly in Catholic countries and in countries that have large Catholic minorities. (Catholicism is an arguably male chauvinist church that - in many spcieties - appeals predominantly to women.) Maybe that's less of a factor in the US, where liberal Catholicism (however defined) is numerically relevant.
But that is just a factor that made very many women loyal to the hard right (well, again, less so in the US - and in Germany, it would be the center-right; well, I'd say CSU is hard right, but not far right). A different explanation is needed for men' rightwards shift.
(If liberalism/leftism were all about the loudest online woke folk, and conservative about Trad lunacy, then it would be clear: why would educated women be averse to the "women belong pregnant and barefoot in the kitchen" camp, and even mildly self-aware men averse to "stfu men, none you have opinions, only male opinions, get out of here and give up your jobs, also, male privilege" camp? )
One obvious difference between sibs and classmates is that most people aren't all that interested in impressing their siblings - in fact, they may just want to troll them instead. I think there are plenty of young men, for instance, who decide to moderate their views on abortion after hearing how zealous their dating partners' views on the issue are, but nobody cares that much about impressing their sister.
It's also possible that the shared effects of parental political beliefs outweigh the sibling effect somehow - the moderation effect may be more important when people differ widely in beliefs.
I have looked for evidence of these trends in data from the General Social Survey and the American National Election Studies. So far I have not found a growing gender gap in political alignment, party affiliation, or policy preferences.
https://allendowney.substack.com/p/the-political-gender-gap-is-not-growing
The critical theory 'woke' ideological complex is hinged on Care-Harm moral values. This is really attractive to women. Some suggest almost necessary for some women who have kids later and later if at all and need a substitute
I think the increasing majority of women in college, increasingly in the social sciences and humanities, are adopting this belief and moral system. I am quite surprised that men have barely budged but it does fit that concept
Paul Graham's wider political takes are generally not very good outside of stuff he has direct experience with (bureaucracy, taxes). And even there he's not hugely insightful.
With the exception of South Korea, all three cases are men mostly varying within their normal range and women getting extremely liberal. The failure to replicate is also a non-sequitur: the first chart is about supporting conservative or liberal parties or self-identification. The debunking is about a battery of ideological questions. And it shows the same phenomenon: while men have gotten somewhat more liberal women have moved further left.
That's the phenomenon and so any explanation needs to focus on why women are getting more politically radical and becoming less likely to self-identify as conservatives. It is a change and one that's been ongoing for some time (before the internet age). MR at least explains the correct phenomenon with a plausible mechanism. But many people (like Paul Graham) are avoiding seeing this as something driven by women or a longer run trend. Others are essentializing it, saying women or men have always been X or Y, when there is a change that needs explaining.
South Korea meanwhile is caught up in East Asia's gender wars and also has its own, relatively intense versions. While it might be partly driven by some global trend there's also a specific regional trend that doesn't exist outside of the region.
I would guess that for a man, having a sister, specially a younger sister, might make you more conservative.
Everybody seems to be just skipping past the "does not replicate" part...the Cooperative Election Study is a huge sampling collected carefully and consistently for nearly 20 years now. If a new divergence is real there needs to be some explanation for it not showing up there even a tiny bit.
Used to take PG's theories seriously until I learned of his horrible understanding of moral equivalence with regards to the Hamas-Israel war. A techno capitalist should not side with terrorists.
I very much dont agree with my siblings about politics, I'd bet family is an extremely bad predictor of politics for social media generations; maybe in an era where family was 2nd place(in thoery?) for information compared to tv news that all agreed; now people don't even share facts; my siblings least of all(they told me to stop sending them political videos)
Personally, as a Zoomer male, I think large segments of the left dislike us. You see constant complaints of the "cishet white man" and such from many of their circles. They're not exactly making it difficult for us to feel like they're the 'enemy' so to speak.
I know this isn't true of all leftists groups, but the fact that some parts of the edifice aren't completely rotten does not really make me any less keen on the whole thing coming down; not to mention that the seemingly-innocuous groups seem to have been the precursors to the rotten ones.
Allen Downey asserts that it is untrue: https://allendowney.substack.com/p/the-political-gender-gap-is-not-growing
https://youtu.be/ywtKokFjYb4
The above is a good video on the topic. So until the late 20th century men used to more left wing than women. Also when polled on individual issues the extent of polarisation is much lower than the initial data point would suggest. So the political positions of liberal and conservatives voters might be smaller than people expect.
Surely this is due to gender being a tool applied for the purposes of identity politics - gender is being politicized by social media algorithms and bad faith pundits and political actors.
The right have abortion bans, Trump (a rapist), anti-trans legislation, "traditional values" where women are housewives etc. The left have cancel culture, "believe victims", and vocally supporting every group except men (especially white ones).
I can't believe so many people are confused about this. It's clearly #MeToo-style feminism. (#MeToo, of course, started in 2017. But the variety of feminism that it epitomizes is much older, and was around in the aughts.)
Does no one remember the genderwars of the early 2010s? These issues aren't as discussed anymore, but that isn't because the underlying causes behind them went away.
“Boys and girls used to hang out together more” seems glaringly false to me. Things used to be WAY more gender-segregated, but have gotten more coed over time. In the past 10 years alone, my college dorm decreased its number of single-gender floors by 75%, as well as allowing opt-in ability to have opposite gender roommates in doubles. Coed bathrooms have also drastically increased in number (mostly due to the prevalence of nb people now), and while no one really hangs out in bathrooms, I think it’s a sign of just how many other things went coed first.
More anecdotally, everyone in my social group has opposite gender friends nowadays, while I know that my dad had basically never talked to women even by the time he’d graduated. Everyone I know also seems happy to live with opposite gender roommates in their apartments, while this would have been unpopular a few decades ago and outright scandalous a few decades before that.
I think a much simpler explanation for increasing political differences is that liberal politics have gotten more focused on social justice (which is often exclusive towards men) and conservative politics have become more focused on the pro-life side of the abortion debate (which is repellent to many women for obvious reasons).
As a sanity check, perhaps you could sample from liberal and conservative political groups and see if they’re male or female, rather than sampling men and women and asking if they’re liberal or conservative, just to see if the two approaches are roughly comparable to each other. Of course, you’d have to be able to account for age.
Does anyone have numbers for this?
> the finding doesn’t replicate and might not be real at all.
That link still shows the gap increasing recently, even if it also shows men getting more liberal relative to the past.
Part of this has to be a function of education, right? People with graduate degrees have very liberal views, generally speaking, and more and more such degrees are going to women (I think both in raw numerical terms and the proportion of such degrees that are awarded).
What do conservative and liberal even mean in this context? If we divide each category further with social and economic qualifiers would the resulting 4 groups relate to each other in the same way? What if the terms feminine and masculine were used in place of liberal and conservative, would the results be similar?
To me the most confounding part of anything like this is in a respondent's self-reporting what broad cohort they should be placed within. I think in truth this detail is situational and malleable, and is always a matter of relative perspective.
The problem with sampling ACX readers on this is not where they fall on the political spectrum, but WHY they fall where they did. Your readers seem generally more likely to be the sort of people who arrived at their political opinions through some kind of intentional deliberation (or at least formulated a defensible post hoc rationale for the ones they already have.) I suspect your readers are meaningfully less susceptible to changing positions merely to get along socially, and when they sometimes have to pretend to believe in X because of their peer group they're more likely to be able to compartmentalize that and maintain their true belief in not-X which is reported in private surveys.
If I was dating a gal in the Sierra Club and had to feign sympathy towards some useless endangered species, I could probably do so and still maintain my true belief in a basically libertarian property rights approach and accurately report that on a survey. But most people, if they say something enough times, start to inch towards believing it. I suggest that the kind of folks who read your blog and those like it are arrogant about our intellect and reasoning ability, think we're smarter than most people we meet on a regular basis, and are therefore less likely to be influenced by the herd effect being discussed in those tweets.
If the gender split in politics is pretty recent, it might make more sense to just look at the younger respondents to the survey.
...Yin in the ascendency, always brings chaos, thus the growing disequilibrium between men and women in the everyday, not just in the political. Pursuing kindness at the societal scale, is a path paved by a good intention, but as we are seeing across the West, compassion before justice, in the particular and at large, impoverishes all. The quiet part, but getting louder, is that new wave feminism/progressivism is the witchcraft of our age. (If you want some numbers to tell you something Scott, check out how the original witchcraft wave played out in the U.S. and Europe)
I wonder if young men are really conservative or if they are just opposed to wokism, which they feel is hurting them. I would expect them to be much more open to traditional leftist positions like tax the rich, distribute wealth, social security for everyone, limit corporate greed and similar.
My 2 cents: because politics has become less important, it's now fashion; and men and women have radically different fashion choices.
Politics used to be life or death for lots of people. Gay people's ability to marry people literally hung on which party was elected in their locality. And that's only the most recent example. Women's rights, policies that affected non-white people, all that stuff. These days, the gap between right and left is smaller. Example: even Donald Trump, who was incompetent and terrible in so many ways, failed to derail America. The Tories, who are pure expressions of self-serving abuse of power, have not managed to destroy the UK (I really thought they might). Right and left leaning parties bring only marginal levels of change.
Therefore, people realise that their lives and livelihoods are not at stake, and vote on other factors - essentially aesthetic or fashion factors. And right now, the fashion is swinging in one particular direction. It could easily change course, and make women more right wing.
https://manifold.markets/oh/will-the-findings-about-an-politica?r=b2g
This feels like one of those ideas that's just so salient that it's dangerously difficult to avoid immediately jumping on a pet hypothesis that obviously explains it.
Even after it doesn't replicate, and thus shouldn't need any explanation.
I think that increasing time online versus interacting in person are at the heart of this, like so many other things.
Think of it this way, a brother and sister may occupy the same house, but their minds may be a thousand miles apart thanks to the Internet. If one spends all evening on tumblr and the other on 4chan (1) occupying the same household may make less of a difference than it would have in 1980.
An interesting question that could start to test this would be to look at the both gender difference in specific site usage and how boys and girls on the same site differ in the politics. Are boys and girls and tumblr more similar in political leanings than boys and girls in total? Are girls on tumblr closer in affiliation to girls on Twitter than to boys on tumblr? Are intersite political differences greater or less than intrasite political differences? I hypothesize yes in some since to all of these, but I'd be curious about empirical data.
(1) Or the Zoomer equivalents. I admit, I'm almost to the age where I'm not sure where kids hang out online these days. Tiktok?
My view is that this is mostly untrue and the non-replications are correct, at least in the US; young women are more liberal than young women 10 years ago, but young men are also more liberal than young men 10 years ago (I'm willing to believe that there's a "divergence" in the sense that the change among young women has been larger, but the design of the graphs and the agenda of the discussion depends on today's young men being increasingly conservative).
I might be willing to buy young men being more willing to *call themselves* conservative conditional on having a given set of political opinions, but that is more about vibes. Maybe also the political identity spectrum increasingly including the "trust the credentialed" vs. "one weird trick" spectrum, which has probably always had a big gender gap.
One thing I tried here was looking at cisgender Americans who selected more moderate responses (not "1" or "5") on the immigration, abortion, minimum wage, and global warming issues. In all four cases (and in the smaller sample size for all four combined) men identified as a bit more conservative on the political spectrum question, though I did not look for statistical significance.
Your blog is where I learned that all politics is signalling, so let me return the favour: this is signalling, and it's pretty funny signalling at that.
If you're a woman, you want people to think that you're caring/educated/cosmopolitan. Liberalism (which is a consumer product marketed as an identity like everything else) sells itself as caring/educated/cosmopolitan,* and has reoriented itself in response to market forces to fill that niche. Your opinions don't matter for non-signalling purposes, so you loudly proclaim yourself to be liberal.
If you're a man, you want people to think that you're tough/self-sufficient/unresponsive-to-social-pressure. Conservatism (which is a consumer product marketed as an identity like everything else) sells itself as tough/valuing self-sufficiency/standing-up-to-social-pressure. Your opinions don't matter for non-signalling purposes, so you loudly proclaim yourself to be a conservative.
My prediction would be that legacy products with skewed branding will fit this pattern; e.g. tankie Communism will be more male than female.
If this has grown over time, it's because politics as brands to express your identity has been finally finishing off that weird 19th century thing where you think about what policies would be good for the country and vote for someone who says they'll enact them. Also, development/liberalism increase gender dimorphism in personality (presumably also for signalling-competition reasons), so that would lead some of the increase too.
*Also rural. You can squint and tilt your head to say environmentalism does this, but I think that's too much of a stretch and it's just the caring thing again. The party that combines liberalism with "voting for us proves you grew up on a farm" will be guaranteed 50% of the vote though.
For the ACX survey, you could check married people vs single. I think I was more influenced by my wife's politics vs any sibling. (I guess this is age dependent.)
I think this is obvious - the left has becoming increasingly misandrist and the excessive focus on DEI and growing dislike of men, especially white men, has started to alienate and exhaust males. Therefore they are less likely to be on board with liberalism. It is well known that feminism, originally a respectable ideology, has spiraled out of control and many of its proponents are now woke idiots that hate men.
Women who are not your sister are able to use additional incentives/rewards to shape male behavior (should they choose to accept the mission). It's unsurprising that they would do so, since unrelated males can be dangerous. It's also unsurprising that they tend to be more liberal as liberal ideology is more empowering to women, and at least tries to moderate male dominance.
I feel like it's less about gender integration (amount of time men and women spend together) and more about a divergence of where men and women get cultural information. It would be interesting to see how this holds up in say, single-sex private schools before social media, where boys and girls are mostly separated but are basically in the same cultural bubble.
There have always been gendered hobbies, skills, and occupations, but the past fifteen years have been historically weird with how much new, niche information can be accessed. Single women (who are doing the most political diverging) generally aren't coming up with these new ideologies themselves from hanging out with other women, but ideas are spreading in majority-female spaces that otherwise wouldn't have the chance to spread. The success rate of specific ideas is different in majority-male vs majority-female spaces, and with the vast magnitude of new ideas that get a chance to take root, this difference is accentuated.
Whereas for most of history, most people are coming up with new ideas from information they get from their family/spouse and their friends who are also in a small, local network. Coming across twenty hot takes a day from people that you've never met is pretty unprecedented.
Scott, I was thinking of re-doing the analysis using a non-parametric test, which I think is what's appropriate for data of this kind. But maybe you already did that? Did you do a t-test (parametric) or a Chi-square test (non-parametric)?
I don't understand how merely *having* an opposite-sex sibling is a good proxy for *hanging out* with the opposite sex.
So what's interesting is that when you look at support for particularly policies there is much less discrepancy. My theory is that the difference here is largely one driven by a greater role of personal value signalling and different aesthetics about political talk in our political discussions (likely driven by social media). Basically, the theory is the way of talking about politics each side has puts the other side off -- and that these strongly correlate with male vs female differences in communication and interpersonal advertising.
I mean, I believe global warming is a serious concern and that we need to worry about how environmental degradation can harm people (eg haze in SE Asia) but when I hear the way many greens talk about how we are hurting mother earth with this vibe that seems to resent progress and our mastery over nature my emotional reaction is absolutely -- fuck that I'd rather let the earth burn than be on the same side as them.
This happens in a hundred different ways on a hundred different issues. It didn't happen before because politics was a broadcast message. The political classes talked to you (and they worked hard not to alienate anyone) but now we talk to each other and the differences between male and female communication styles and value advertising are driving us crazy.
The test might be incorrect: having a sibling doesn't mean you spend a lot of time together.
The piece is great to show some tendencies, though it compares apples with pears - still fruits, but not really the same. The US data shows the support for ideology, while the rest is the support of a certain party. Since there are no Gallups in other countries, this is the best we have, though it does not account to how true to their ideology those parties are. In Romania, the social-democratic party cares little about social democracy, yet it will be classified as a left-wing party there.
The cultural aspect is important in understanding a rise in conservatism. Coming from Romania, it is impossible not to hear Andrew Tate's name on a weekly basis. What has he done for society? Absolutely nothing, turned a bunch of young boys into self-obsessed mean that speak fast, drink sparkling water and have an issue with fat, liberal and "woke" people. Tracing all the problems of the world to those groups is easy and requires little effort to investigate the world, thus a growth in those conservative thoughts.
My guess would be siblings don't moderate each others' political views much because they start out too similar, on account of having the same parents. Look for people who share a house, but didn't grow up together, and didn't have much chance to pre-screen prospective housemates based on political views.
It's not that it is or isn't a problem, but as with anything trends are interesting and it's a worthwhile exercise to try and understand why things are playing out the way they are.
Another thing worth noting is the established trend that male and female professions tend to diverge more as gender roles become weaker in society (this can be observed in international comparisons). This could be a reflection of the same underlying dynamics, men and women increasingly self-sorting into different tribes as they have the freedom to do so. This would imply that sexual differences are strong and persistent and naturally carry over into various areas of life if given the freedom and opportunity to do so.
It's also somewhat paradoxical in the case of politics though, since historically women have tended to be more right wing and men more left wing. But this just underscores the fact that the tribalism is itself the driver of the differences and not the specific orthodoxies of the tribes that they've self-sorted into.
Everybody (left and right) wants to make it about their personal gender war hobby horse, but it seems more likely it's due to demographic shifts, such as those mentioned by Marginal Revolution. Women historically tended to be more right wing than men, since they were less likely to be members of unions and were more religious; as women became near-equal contributors to the work force and secularized, these influences disappeared and they became more left-wing than men. If in the past such trends have been primarily demographic, why would we expect now they'd be based on gender war stuff?
If the effect is real, of course.
If this doesn't replicate and may not exist... why elevate the topic to a discussion at all? It seems like the very stereotype of "angels dancing on the head of a pin" style discussion that generates no knowledge or information if you're merely speculating about what could cause an effect that doesn't exist.
Can anyone find the Washington Post-ABC News poll that found people with daughters were more likely to vote for Hillary in 2016?
As I recall the poll actually showed that people with children of mixed sexes were about as likely to vote Trump as people with only boys, whereas people with only daughters were more likely to vote Trump, so the actual finding was that having at least one son made you more likely to vote Trump. The poll appears to be paywalled, but Huffpo of all places has the finding:
https://www.huffpost.com/entry/people-with-daughters-are-more-likely-to-vote-for-hillary-clinton_n_58053b6ce4b0b994d4c0e164
My interpretation was the presence of at least one son means you would have been worried about the effects of feminism on your son if Hillary won, but I am obviously not a neutral observer here...
Allen Downey has tried to replicate these results with publicly available data and found no effect
https://www.allendowney.com/blog/2024/02/11/the-political-gender-gap-is-not-growing/
Why omit people with both brothers and sisters? Seems like the important part is having at least one opposite-sex sibling; not having a same-sex sibling doesn't seem important.
(I say this as a male with both a brother and sister.)
Every failed hypothesis highlights the incredible ability the human brain has to concoct a plausible-sounding hypothesis to literally ANY set of data, and that we shouldn't accept any hypothesis until we have some supporting data (and NOT what was available before the hypothesis was generated).
My personal experience is that I (M) have two sisters and a wife who are all significantly farther to the right than I am. Not that my anecdote is statistically meaningful, so much as that my personal experience leads me to be skeptical of the hypothesis sans compelling evidence.
Meanwhile, is there a more broad assessment of country trends? I'm skeptical this is just something happening in these four countries, and if we did a true international survey we'd find no general trend, just a lot of noise that only looks like 'signal' when you cherry pick the data. The failure to replicate suggests this isn't a robust observation on which to base any hypotheses. Once again, the prudence of starting any discussion of anomalous data with, "But what about the null hypothesis?" suggests itself.
I am much less concerned about my sisters approval than I am of a potential partner. Eventhough I start getting into cooky political stuff my sister will still be my sister. However my long time partner might no longer be my partner
Presumably siblings are comfortable with each other, and there is no need to impress the opposite sex by exhibiting their approved behaviours. In other words, stranger girls and boys might rub off on one another much more than siblings.
I hate to say it, but I think an ACX sample is not very useful for assessing the USA or the rich world as a whole.
You can see from his sample it's got almost 10 times as many men as women (a big problem if you're studying gender divergence), and is composed heavily of younger men with technical interests. Given the history of the blog involves a doxxing attack on its author by a major arm of the liberal media, it tends to lean center-right.
I think it's a great sample for looking at attitudes among techies (who are a culturally important segment given the importance of computers and the internet these days), but not much more than that.
Александр здравствуй, позвони мне или почини этот сайт
I'm a young adult and still live with my siblings, and have fairly different political views from them. I think being in the same family and interacting with each other moderates us a *bit*, but our internet siloes are still very different and it generally doesn't feel worth it to work through our different worldviews built out of watching hundreds of hours of video essays and podcasts leading in different directions. The number of different assumptions and logical steps built on other steps just seems overwhelming.
to me the obvious explanation is that news feed algorithms and balkanized media amplify initially small differences, ala James Flynn's basketball argument. Media used to be more ideologically integrated when people were reading local newspapers and watching local TV channels that tried to serve everyone in a particular geographic area. Now media is all fine tuned to pander to the biases of particular ideological niche.
Politics in general has polarised, which seems to be something not solely correlated to gender. If politics polarises, and there's already a gap between men and women on average, maybe the gap widens as a result of the landscape moving rather than a change in gender relationships?
I wonder how the gap would look when tested for individual ideas, rather than as a bundle. E.g., separating out women's rights from issues like tax burden, would be interesting. It's possible that answers around reproductive rights skews things, though I don't have a convincing argument for why the gap around those in particular would have changed.
Nope, not Chinese. Not American, just British colonial ancestry.
I think the crackdown in China is more about cultural progressivism than consumerism/capitalism. In fact in economic management terms the Chinese are actively trying to redirect from exports, to internal consumption. But the Chinese are savers, ( because there is no social welfare) so it's not working very well. The same Chinese prof I mentioned before, was also happy to explain the chaos of 1960's 'cultural revolution' as very much a shift in the Yin. ( It was Mao's wife, Jiang Qing who led the 'Gang of four').
As for other episodes of mass mania, check out Richard Landes' 2010 book, "Heaven on Earth-The varieties of the Millennial Experience". The chapters on the tragedy of the Xhosa people in the mid-19th century is especially relevant to the Greta Thunberg phenomenon.
I just realized that in every case more popular within girls ideology is "left" and within boys "right"
And not only "left" and "right" have very different meaning in different countries, but they also changed during 90s-20s
I am too lazy to argument for a specific explanation