I can’t find evidence for ‘progressive cat ladies’
The ‘progressive cat lady’ is a staple of culture war discourse. Famously, J. D. Vance once dismissed Democrats as ‘a bunch of childless cat ladies with miserable lives’. After criticism, Vance apologised for the tone of his earlier comments but defended the content: ‘The substance of what I said … I’m sorry, it’s true’.
Despite the centrality of the cat, it distracts us from the meaning of these sentiments. In reality, such commentary is about female childlessness and its ideological implications. As the culture war has intensified, competing visions of the family (or lack of one) have become central to it. In contrast to the ‘childless cat lady’, figures such as Vance favour large and patriotic families with traditional gender roles. On social media, the ideal of the conservative ‘tradwife’ contrasts with the progressive cat lady.
Beyond social media caricatures, the ‘heterodox’ commentators who share certain sympathies with Vance have made interesting arguments about parenthood and ideology. Some assert that lack of children encourages emotional immaturity, underlining neurotic trends in progressive ideologies. Others stress the influence of motherhood in inculcating resistance to progressive ideologies; parenthood involves duty and obligation and such values are central to conservatism.
To be sure, hypotheses about parenthood and ideology are crucial and, in various forms, scholars have long debated these topics. Such arguments tend to focus on women, yet one should not forget men; fatherhood can be equally transformative. However, heterodox arguments have considerable problems. Journalists advance many of these and such commentators lack familiarity with academic methods. A community of heterodox scholars has emerged, but many of these figures indulge in polemics and/or fail to use data and conventional methods.
Admittedly, established academia has its own problems with this issue. Though some work exists on progressive ideology and parenthood – a 2009 qualitative account argues that the experience of not being a mother can involve a sense of social exclusion, potentially kindling radicalism – this scholarship tends not to engage with recent developments in progressive ideology. As heterodox commentators grasp, recent changes in such ideologies have been profound.
Our forthcoming book addresses such gaps. Within the broad tent of progressive ideologies, we distinguish between social justice ideology (others call this ‘wokeness’) – this emphasizes identity and anti-capitalist direct action and extends conceptions of harm – and liberalism; this stresses freedom, the common and equal basis of political participation and the legitimacy of rules.
Most of the book does not concern parenthood, yet some of our data speak to this debate. Generally, investigating the relationship between progressive ideologies and parenthood is challenging. Beyond the lack of established indices for measuring social justice ideology – we favour the recently developed Progressive Values Scale (PVS), yet this does not feature in established surveys – many surveys do not have suitable items on parenthood.
To address such problems, we fielded our own surveys in the US and UK which featured the PVS and items on parenthood. This investigation did not attempt to establish causality – the (quasi-)experimental manipulation of parenthood is next to impossible – but aimed to establish associations. Crucially, the PVS uses a sample of progressives (i.e. supporters of social justice ideology and/or liberalism); this means that the variable of age, which has a strong association with parenthood, has less influence on results than in samples of the general population in which there are major age differences between progressives and conservatives. Moreover, this approach enables comparison of tendencies within progressive ideologies, many associating the influence of childlessness with social justice ideology, rather than liberalism.
In both the US and UK, childlessness does not have a statistically significant relationship with social justice values. In other models, we created an interaction term between female gender and childlessness, this measuring whether the combination of childlessness and female gender increases the likelihood of respondents supporting social justice ideology. As with childlessness, the interaction term failed to achieve statistical significance.
We undertook similar analysis with the British Election Study (BES). This uses a sample of the wider British electorate, enabling comparison of progressives and conservatives, and, with its culture war index, measures social justice ideology differently. Though childlessness is associated with social justice values, the size of the effect is limited and, as noted above, also reflects the influence of age. Crucially, the interaction term between female gender and childlessness does not achieve statistical significance. Results for the BES liberalism index are similar. Overall, we find little evidence of a relationship between childlessness and progressive values.
The interpretation of such results requires caution. In an era of replication crises, single studies (though we analysed several datasets) are regarded as unreliable and this area has particular issues; for example, researchers disagree on appropriate measurements of social justice ideology and, to a lesser degree, liberalism. Quantitative methods have limits. Famously, such methods have difficulty with small numbers of cases. Potentially, the experience of childlessness may be important in certain cases and qualitative investigations, similar to those undertaken years ago, might ascertain this.
Whether such research will emerge is another question. About five years ago, when social justice ideology was growing quickly in influence, heterodox commentators made crucial points about emerging phenomena; the theory of luxury beliefs is another example. At that stage, hypotheses (rather than data) were welcome and acceptable; such phenomena were new, there had been no time to undertake empirical research and established academics were responding at an unsatisfactory pace.
As time has gone on, such hypotheses have gained much wider audiences, Vance’s comments reflecting this. But despite their popularity, evidence for such hypotheses remains unsatisfactory. One cannot entirely blame this dearth of evidence on the heterodox movement – the community has limited resources and, as I say above, this also reflects problems with established academia – yet this movement makes greatest use of such concepts.
And when longstanding theories have limited evidential bases, this becomes a problem.
If you enjoyed reading this, do think about subscribing! Subscription is free – all it means is that you’ll receive a weekly email. But every new subscriber makes me very happy 😊 😊 😊
I think there is a distinction to be made between what is academically 'provable' and ... let's call it 'lived experience'. There's a reason for these tropes and they don't come out of thin air. My theory is that there are archetypes of all political persuasions and that these are typically the people who are overrepresented especially online.
We had this during Brexit, with people who looked a certain way always being identifiable as Leave supporters just from their faces and misspelling borders as 'boarders'. They were Leave archetypes. Progressive, childless cat women may not be an empirical or epistemological fact identifying the leftishist cohort, but they certainly exist and we can all recognise them. I thought this too when you demonstrated the lack of empirical support for Luxury Beliefs.
In other words, I think you're dead right, but that there is also a there there.
> In both the US and UK, childlessness does not have a statistically significant relationship with social justice values. In other models, we created an interaction term between female gender and childlessness, this measuring whether the combination of childlessness and female gender increases the likelihood of respondents supporting social justice ideology. As with childlessness, the interaction term failed to achieve statistical significance.
Is the question whether "responders supporting social justice ideology are more likely to be childless" also covered?
Because the interaction can be assymetric (since women in the last decade at least are more likely to vote progressive from polls I've seen).