How the “prestige penalty” keeps us from crossing gender lines in the labor market


by Lauren Valentino

It’s an oft-repeated line: “women earn 80 cents for every dollar a man earns.”

Typically, social scientists explain this gap by pointing out the fact that men and women end up in different job sectors. In some versions of the explanation, it is assumed that men and women pick different jobs on the basis of their masculine or feminine preferences. Since feminine-type work pays less, women, on average, earn less.

But why would people choose to go into gender-typical sectors (men into construction, women into teaching), even in today’s society, where substantial – though incomplete – progress has been made toward gender equality? Many social scientists have shown that discrimination is an important part of the answer.

My research suggests that there may also be a “prestige penalty” at play. It is not necessarily that women freely choose “women’s work” and men freely choose “men’s work,” but instead that there are social repercussions to being a man in a “woman’s job,” or a woman in a “man’s job”.

If a woman or man chooses a job that is the opposite for their gender in terms of societal expectations, she or he is seen as having a lower social standing.

As a cultural sociologist who studies the labor market, I am interested in more than just how much money people earn in their jobs. That’s because work affords us more than material rewards: it also gives us a sense of purpose and meaning, a role in society, and a certain level of social status more generally.

Sociologists see work-related status as conferring “occupational prestige.” In my work, I try to understand what it is about a particular job that makes us think of it as a “good” or “bad” job.

I analyzed data from the General Social Survey (GSS), which asks survey-takers to rate over 850 occupations in terms of prestige. These occupations span the gamut from “potato chip sack machine operator” to “member of a board of directors of a large corporation.” The survey-takers were asked to rate each occupation on a scale of 1 to 9, where 1 is an extremely un-prestigious occupation, and 9 is an extremely prestigious one.

In the GSS a handful of occupations were rated twice – in their male and female moniker. In three out of four cases, the male version of the occupation is rated as more prestigious than the female version.

Actors receive a mean prestige rating of 5.740, whereas actresses receive a mean prestige rating of 5.463. Automobile repairmen are on average rated as 4.638, repairwomen as 4.583. Waiters received a score of 3.579, waitresses 3.556.

Occupations are seen as more prestigious when a man, rather than a woman, does them.

The one exception to this pattern is masseuses, who are rated slightly higher than masseurs in terms of prestige (3.700 versus 3.632). This led me to wonder whether the masculinized or feminized nature of the job was an important factor that influences how people perceive a job’s prestige.

One of the most striking features of the labor market from a social science perspective is just how much gender segregation exists there. So it should not surprise us that the gendered reality of the labor market seeps into our perceptions of that occupation’s social standing as well.

Indeed, acting and automobile repairwork are male-dominated professions. Census data shows that women make up 25% of the acting profession, and a mere 11% of the automobile repairwork profession. And women make up 76% of the massage industry.

These figures suggest that we see a worker’s job as prestigious when it matches our gendered expectation of what that job is. Acting and automobile repairwork are perceived as ‘man’s work,’ and when a woman does that job, it is seen as less prestigious.

On the other hand, massage therapy – a form of intimate labor and “care work” – is feminized, and men who go into that career experience a “prestige penalty” as a result.

Waiters and waitresses present an interesting counter case. Although waiters are rated as slightly more prestigious than waitresses, there are more waitresses than there are waiters in the US: 243,097 waiters versus 437,680 waitresses.

This suggests to me that our perceptions of how masculinized or feminized a job is may not necessarily reflect the true gender breakdown of that job. Waiting tables has a complex history as both a serving job and as a part of the masculine “restaurant culture.” These two dynamics may provide countervailing cultural messages about just how “masculine” and “feminine” the job of waiting tables is.

It is important to note that these prestige differences between the male and female occupations are not statistically significant by conventional social science standards. This is the case because GSS survey-takers were each given a subset of 90 occupations to rate, since asking each person to rate all 850 jobs would have been prohibitive. Each occupation is thus only rated by a handful of survey-takers (between 73 and 84).

Nevertheless, most of the differences I discuss here are large enough that they would likely hold up in larger samples more typical of social science datasets. Moreover, the differences I do observe all point to the same story: when men do the job, we see it as more prestigious, unless those men are doing “women’s work.”

These prestige differences in gendered occupational titles reveal how we evaluate men and women differently, even when they are in the same job.

Our perceptions are influenced by the labor market’s gender segregation – how many men or women do a job. We subconsciously use this information to form expectations about the context of that job – how masculine or feminine it is. Then we penalize men who do women’s work, and women who do men’s work in terms of their social standing.

These differences are likely both a cause and an effect of the gendered process of occupational sorting that social scientists have long observed. We seem to be trapped in a cycle in which people avoid the “prestige penalty” they would suffer by crossing occupational gender barriers, further perpetuating the entrenched gender lines in the labor market.

Lauren Valentino is a PhD candidate in sociology at Duke University.

Image: Fleet & Family Support Centers via Wikimedia Commons

  1. Hi Lauren!

    Is there a paper available somewhere that discusses this research in more detail? I don’t see one in the links. Thanks!

    • Lauren said:

      Hi Dan! There’s not a paper on this “finding” since the data are so sparse. (I’m hoping to use the intriguing preliminary findings presented here as justification for a funding request to collect more data on the topic!) I do have a paper testing some of these ideas in a slightly different way that I will be happy to post a SocArXiv link to soon.

      • Thanks for the reply! I admit to being quite skeptical, and was hoping the underlying paper would have more detail. We’re theorizing here with differences in the third sig fig. And we’re seeing that 4 of 5 go in one direction (some just barely) while 1 in 5 go the other way. Both the magnitude and distribution seem very consistent with a – quite surprising! – null finding. What if we switch our framing from “is this difference statistically significant” (or “would this difference be statistically significant with enough data”) to a question of “how big do we think this difference is?” Our best estimate would be quite close to zero, which is in some sense fascinating given that “waitress” and “waiter” are so clearly gendered, and “repairwoman” is downright clunky. What this would suggest, I think, is the opposite: that men and women gain (or suffer) equally in the prestige of gendered jobs.

      • Lauren said:

        I take away from this a slightly different point than a mere ‘women’s prestige penalty’: that being gender-atypical in a job (a misalignment between perceived masculinity/femininity associated with a job and the job incumbent’s actual gender) may carry a penalty. In that case, if one does buy that waiting tables is perceived as a masculinized occupation (a big if), then the directions of differences do, in all 4 cases, point to that pattern. Certainly the differences are small, but I think they suggest that it’s worth looking into these dynamics with a lot more data!

  2. matt vidal said:

    It’s hard to be precise about the practical significance of this measure. We’d need to see the means and SDs across the results for the range of occupations in order to begin to judge this. What we see right now is a 2.04 difference in means across occupations (actor to waiter). If that were close to the max variation around means, then perhaps a .28 difference (between men and women actors) has some practical significance.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: