Re-thinking how we measure benevolent sexism

by Anna O’Sullivan (class of ‘25)

“Alright, class! I need a couple of strong boys to carry some heavy stuff. Don’t make the girls do it!”

“It’s my job as the man to protect my girlfriend.”

“You make the best meals; women just have that special touch in the kitchen.”

“Let me open that door for you, miss.”

Nice, right? Not really. While these conversational snippets seem like social pleasantries, they’re examples of benevolent sexism.

The term comes from a 1996 paper by Peter Glick and Susan Fiske, where they argued that two different types of sexism need to be measured and accounted for separately. Hostile sexism is the more obvious form of sexism. It encompasses overtly negative attitudes that reinforce women’s subordinate status — ideas like “women are weaker than men” or “men should hold higher positions than women.”

On the other hand, benevolent sexism often professes seemingly positive character traits about women, things like “a man is not complete without a woman” or “women should be rescued before men in a crisis.” Because of this, benevolent sexism can be harder to identify. It often comes from people who may consider themselves anti-sexist or even feminist. Glick and Fiske explain that benevolently sexist ideas promote “prosocial behavior” like “helping” and “intimacy” and might seem like a social good — even beneficial to women. However, benevolent sexism is an insidious series of attitudes that trap women in stereotypical, subordinate roles. These attitudes reinforce a gendered power structure and place undue burdens on women.

Glick and Fiske explain that benevolent sexism can be broken down into three categories: complementary gender differentiation: the belief that men and women have specific and separate complementary traits that further justify the traditional division of social roles where women work inside the home and men work outside the home, protective paternalism: attitudes and behaviors that justify male dominance, saying that women aren’t as capable of certain things and need a man’s help or guidance, and heterosexual intimacy: men’s desire for intimacy and sexual reproduction with women, coupled with expectations of the fulfillment of these desires.

Some of this scale-defining definition still seems to capture benevolently sexist attitudes accurately. However, other parts of it seem weak. If we follow the logic of the tri-category definition, we come to terms that either seem to be measuring hostile sexism or measuring something other than sexism altogether. The traditional definition of protective paternalism means the statement “I would walk a woman home at night” could be considered a term measuring benevolent sexism. After all, this captures the benevolent attitude of seeing women as needing help or guidance. However, a respondent may know that sexual offenses occur at a higher rate to women, especially women who are alone. The staunchest feminist could, for good reason, offer to walk a woman home. Perhaps this measure could gain more weight if it said, “Even if a woman said no, I would still walk her home” or “Even in broad daylight, I would still walk a woman home.” Still, these seem to drift from the benevolently sexist attitude they’re trying to measure. The first seems to veer into consent rejection, appearing like a more hostile attitude. Even the second seems to be more complicated than a matter of sexism — perhaps a test of how seriously you take the potential danger to women or how willing you are to give up an afternoon to protect somebody.

These are the problems I’ve repeatedly encountered as I’ve been working to create an amended Benevolent Sexism Scale. How each item on the scale is phrased can significantly change its meaning. And just because there is a benevolently sexist attitude, like thinking women need more help than men, doesn’t mean that any example that could be related to this attitude (like walking a woman home) is sexist.

Some phrasing issues can be solved through adjustments in language. The idea that “women are better caretakers than men” may be true, as society has shouldered far more caretaking labor onto women, giving them practice from early childhood when many are gifted their first baby doll, their male counterparts receiving race cars. So, an amended phrasing to better capture truly benevolently sexist beliefs, and not just informed reactions to lived realities under the patriarchy, might say, “Women are naturally better caretakers than men.”

As another example of the traditional definition’s failings, this sassy Psychology Today article points out that benevolent sexism scale items about how ‘a man should cherish his partner’ seem to misclassify affection between partners as something sinister. Sure, that phrase might be rated ‘highly agree’ by the man who treats his wife as a trophy he’s won — something he’s acquired that he can hold on a pedestal and pin emotional expectations on (or by that wife). However, it might also be rated ‘highly agree’ by somebody who genuinely loves and appreciates their partner. Perhaps deep down, if you really examine the root of the phrase, it has strange connotations for anyone, but it doesn’t seem like a clear marker of benevolently sexist attitudes. However, if this gendered cherishing was placed in contrast to its reverse, that could change things. If we measured people’s responses not just to whether men should cherish their partner but if women should, perhaps that could get at the dynamics we’re missing.

I’d also like to push on the traditional benevolent sexism scale’s broad category of ‘heterosexual intimacy.’ Many items that fall under a man’s expectation of the fulfillment of his desires — particularly sexual desires — do not seem like benevolent sexism. For example, the idea that women owe men something sexually if they let the man pay for a date. Indeed, the idea that men should pay for a date is benevolent sexism. This is a classic dynamic that the scale seeks to measure in which the man is the provider, and the woman is dependent. However, the sexual pressure that follows from this, while related to this benevolent sexism item, doesn’t seem benevolent itself. The idea that women should fulfill their partner’s sexual needs and desires does not seem to fall under the definition at all. Perhaps it is hostile sexism; perhaps it exists in its own category that I’ll need to explore. But benevolent sexism does not seem like the right label.

I am not refuting the entire heterosexual intimacy category — certainly, many of the items describing expectations about women’s emotional labor do seem to be benevolently sexist. The idea that women are “natural caretakers,” for example. However, even in the realm of emotional expectations, there seems to be a line. As I continue to develop and explore this scale, I’ll also be thinking about where this line falls and if we need to be thinking about a gradient between hostile and benevolent sexism, the introduction of a third category, or something else entirely.

This fall, I plan to test my scale to amend the phrasing further. These tests will include randomly changing the wording of items to help gauge how responses change as a result. For example, to test my intuition that asking respondents to respond to statements in how they relate to both men and women will capture attitudes the original scale is missing, certain respondents would be asked their agreement/disagreement with the statement “Women are incomplete without men.” Another respondent would be given “Women are incomplete without men” and “men are incomplete without women.” This will help me refine the language used in the scale, which is incredibly important and I predict will be hugely influential to respondent’s benevolent sexism measures.

The items will likely change as I conduct these tests and further develop this scale. However, what I hope to offer with this initial blog post, rather than the 33 items as a definitive scale, is an understanding of the closer scrutiny we need to have when measuring benevolent sexism. The scale was developed to measure a nuanced, insidious form of sexism. As such, it’s especially important to be wary of the terms we use when measuring it.

Here are the items I am considering for the new benevolent sexism scale I hope to construct:

1. In heterosexual dynamics, men should pay for dates.
2. Women are incomplete without men.
3. Men are incomplete without women.
4. Women are more naturally sensitive than men.
5. Part of the benefit of being in a relationship with a woman is their ability to care for emotional needs.
6. Part of the benefit of being in a relationship with a man is their ability to care for emotional needs.
7. Women have a quality of purity few men possess.
8. Women are naturally better teachers than men.
9. Women are natural caretakers.
10. Men are natural caretakers.
11. Women should interrupt their career or education plans to care for loved ones.
12. Men should interrupt their career or education plans to care for loved ones.
13. Women should be willing to move, stop working, or take on caretaking responsibilities for their partner’s career.
14. Men should be willing to move, stop working, or take on caretaking responsibilities for their partner’s career.
15. A woman can’t truly be fulfilled without being a mother.
16. A man can’t truly be fulfilled without being a father.
17. It’s pitiable when a woman doesn’t get married.
18. It’s pitiable when a man doesn’t get married.
19. Women are natural leaders.
20. Men are natural leaders.
21. A woman’s vulnerability is a positive trait.
22. A man’s vulnerability is a positive trait.
23. A woman’s outspokenness is a positive trait.
24. A man’s outspokenness is a positive trait.
25. A woman cursing is unnatural.
26. A man cursing is unnatural.
27. Women should seek a financially well-off partner.
28. Men should seek a financially well-off partner.
29. A man should open the door for a woman, even if she doesn’t believe it is necessary.
30. A man should walk a woman home at night, even if she doesn’t believe it is necessary.
31. Women should be saved before men in a crisis.