Recalculating The Gender War

Recently, Science Advances published a study entitled How many cents on the dollar? Women and men in product markets. The paper uses a study of eBay sellers and transactions to claim that female sellers products are unjustly valued less than male seller’s products. Conclusions are converted into cents on the dollar estimates reminiscent of false gender wage gap statistics:

“on average, when selling a product, women received 94.7 cents for every dollar received by men…when selling a new product, women received 80 cents for every dollar received by men.” (4)

However, I am suspicious of both the studies methods and conclusions. The study’s authors claim to have accurate gender data on seemingly anonymous eBay sellers. Furthermore, they claim that eBay buyers accurately guess the gender of these largely anonymous eBay sellers (even though their own data suggests otherwise) and base their decisions on it. The study is an obvious attempt to invent a “wage gap”-style societal grievance feminists can rally around.

Both Time Magazine and Fortune wrote articles about the new study and uncritically accepted its findings. Both Time and Fortune magazine also used the opportunity to push even more sensationalist misinformation about the “gender wage gap” and women in the workplace.

In its article entitled “Even on eBay, Men Make More Money Than Women”, Time magazine writes:

“The wage gap between men and women has been shown to be real time and again: men tend to make significantly more than their female counterparts for the same work.”

No, it doesn’t. It doesn’t show that men make more that women for the “same work”. It shows that men tend to choose career fields that are more demanding, more dangerous and more lucrative. Even when they are doing the “same work”, it’s often not really the same work. For example, male doctors tend toward more demanding medical specializations than women. Men also tend to work longer hours and take less time off work.

This is why gender wage gap propaganda is so dastardly - there technically is a gender wage gap, but it is not likely due to be gender discrimination, but due to the different career choices men and women tend to make. The claim that the wage gap is caused by gender discrimination has been repeatedly debunked. However, the gender discrimination line is still spread by feminist ideologues and politicians who think they can use it to garner voters (even when they have a gender wage gap in their own ranks).

The Study: Is eBay representative of product markets?

The study is entitled “How many cents on the dollar? Women and men in product markets”. However, eBay is not a very good general representation of all product markets. For one thing…

Problem #1 - How do you reliably determine an eBay seller’s gender?

The study claims:

“Data were extracted in 2014 at the eBay Research Lab. The 420 most popular products from each of the meta-categories in the eBay catalog were selected for analysis. The gender of sellers and buyers was identified on the basis of the gender reported by users at the time of registration” (7) [emphasis added]

However, seller profiles do not appear to list gender. Clicking through a bunch of eBay listings I couldn’t find a single seller profile that listed gender. Sure, sellers could list their respective genders in their general seller information if they really wanted to (most seller’s don’t because there isn’t any reason to), just as a sellers could just as easily put his/her race, shoe size or favorite pizza topping. However, there is not a separate field that specifically lists a seller’s gender.

I even made an eBay account to double-check. I was never asked my gender during account setup and no where in the account settings could I find an option to identify my gender.

Furthermore, the Time magazine article claims:

“eBay released a statement noting that it was not involved and the research, and nearly half of its top consumer sellers in the U.S. are female. “We are passionate about harnessing our platform to empower millions of people by leveling the playing field for them,” the company said. “We do not reveal the gender of our sellers, although they can choose to do that themselves. When our sellers succeed, eBay succeeds."” [emphasis added]

I haven’t found a primary copy of this statement yet, but its telling how Time magazine expressed little skepticism of the study despite having this statement from eBay.

The study could have guessed gender from the first names of eBay members. However, it is risky and imprecise to guess gender based on names. Furthermore, while I could believe that eBay might share gender data of its sellers, I find it harder to believe that eBay would share the first names of its sellers with the study’s authors - two outside researchers that don’t seem very noteworthy. Frankly, I would be pretty angry if eBay shared any of my personal information with these people without my explicit consent.

However, the study supposedly used eBay transaction data from 2009-2012 (which makes all the data in this new study at least 4 years old). Maybe between 2009-2012 eBay asked sellers for/gave sellers the option to provide gender identity and changed their practices later? In addition, maybe eBay decided to keep this data years later despite changing their practices. Then (strangely), eBay decided to share it with researchers with seemingly no connection to eBay, despite eBay supposedly claiming "we do not reveal the gender of our sellers”.

There are a lot of questions about how exactly the study got this data.

Also, even if the seller is registered as a man or women (which currently there does not appear to be any way for a seller to do) that doesn’t necessarily mean that seller is actually that gender. First, sellers could lie about their gender (again - sellers don’t actually appear to be asked about their gender anyway). Second, while the study supposedly eliminated eBay “stores” from its statistics, some private sellers may still be multiple person operations. You can’t really say the seller is a woman or a man if a man takes the product pictures and a woman writes the product descriptions.

Buyers’ perception of gender

In addition, the study has an Achilles heel. While the authors claim to know the gender of eBay sellers, they admit that most sellers don’t publicly identify their genders:

“Potential buyers do not receive direct information from eBay about the sellers’ gender. Yet, the gender of a private seller can be gleaned from the range of items a merchant is offering for sale (for example, selling female clothing suggests that the seller is likely a woman) and, at times, from the seller’s username.”

The authors want to attribute their findings to gender discrimination, but they can’t unless they can prove that buyers can figure out the gender of anonymous eBay sellers. So the authors are trying to claim that buyers routinely accurately guess the gender of a seller (again - assuming the selling actually has one gender).

“Four hundred people participated in the experiment, each evaluating five user profiles. Of the 2000 evaluations, the gender was correctly identified in 1127 cases and mistakenly identified in 170. In 701 cases, participants reported that they could not discern whether the sellers were male or female”

So 170 (mistaken gender) + 701 (could not discern) = 871 (couldn’t correctly guess gender). 871 (couldn’t correctly guess gender)/1127 (could correctly guess gender) = 56.35% chance can correctly guess gender

1127 (could correctly guess gender) + 170 (incorrectly guessed gender) = 1297 (could make any guess about gender)/2000 = 64.85% chance confident enough to make make any guess (true/false) about gender.

A 56.35% chance respondents could guess the correct gender and a 64.85% respondents felt confident enough to make any guess regarding gender. This doesn’t seem like a very convincing argument that buyers consistently guess a seller’s gender, especially when guessing gender is already only a 50% coin toss. Indeed, it shows that over one-third (100%-64.85%=35.15%) of the time a buyer doesn’t even feel confident enough to guess a seller’s gender. Also, none of this actually proves that actual buyers try to guess the gender of the seller or that it affects their bidding decisions.

Gender is least likely of a sea of factors

The study claims:

“Auctions are ideal for testing for gender differences in outcomes because, after an item has been listed, the final price is not affected by the seller’s behavior, only by the bidding of potential buyers.”

This is not true. There are numerous other variables that could affect a buyer’s bid: the sellers rating, the sellers return policy, the seller’s page layout, the product description, shipping fees, shipping location, warranties, pictures of the item, whether the buyer is currently bidding on competing items, time left in the auction, number of other bids, number of competing auctions currently available, etc.

These seem much more obvious and likely decisive factors than what gender a buyer thinks a seller might be. The study claims it controls for many (but not all) of these of possible factors, but I don’t see how they could possible effectively control for qualitative subjective variables such as the quality of a product photo or product description. They also don’t appear to include major factors like return policies or warranties.

Furthermore, the study finds that men and women tend to sell very different items. Women also tend to sell cheaper items and set lower starting bids. In addition, female sellers only make up only 23.07% of the sellers in the study’s dataset. So even more statistical voodoo is used to try to make data on male and female sellers comparable.

Personally, I have never considered someone’s gender when buying from eBay. It just isn’t a factor. I’m sure the authors would argue I still have a unconscious bias (supposed unconscious biases being conveniently almost impossible to disprove). It seems like the study is trying to attribute its findings to one of the least likely factors that may affect a buyer’s decisions.

Brad and Alison

I’m pretty skeptical about this study. The authors also are claiming to have data on the gender sellers that it doesn’t seem like they could have and claiming that eBay buyers reliably sniff out gender when it appears they do not. However, there is one part of the study that seems somewhat okay:

“To test whether people evaluate products that women sell as less valuable than the same products when sold by men, we conducted another experiment on Amazon Mechanical Turk, asking participants to report the monetary value they assign to an Amazon $100 money-value gift card (“How much are you willing to pay?”) when sold by either Alison or Brad…One hundred sixteen people participated in the experiment; 59 were asked to report their evaluation of a $100 gift card sold by Alison, and 57 were asked to report their evaluation of a $100 gift card sold by Brad. The average value assigned to the gift card sold by a woman was $83.34, whereas the average value assigned to the same card sold by a man was $87.42” (6)

Okay, so that’s a difference. Doesn’t strike me as a major difference, especially when you consider how small the sample size for the survey is. I would also be interested to know how they recruited people through Mechanical Turk. The wording of the recruitment ad could have biased responses. However, while it seems kinda interesting, I’m not prepared to jump to the same conclusions as the study’s authors.

Authors push most feminist interpretation of data

I often talk about feminist bias in research. I have explained that feminism is not simply (or even always) the belief in the equality of women. Its an ideology with various ideological beliefs. The main feminist ideological belief is the existence of a “patriarchy” - a vaguely defined societal system that seeks to oppress women as a class (normally for the benefit of men as a class). When I say that the study’s authors are pushing the most “feminist” interpretation of their findings, I mean that they are pushing a interpretation of their data most in-line with this belief

Notice the study’s authors introduce their study with the claim that:

“Gender inequality persists in U.S. society to this day. In the labor market, progress toward equality in employment…” (1)

Then they drop a slew of stats on the gender wage gap and employment differences. However, as I pointed out with the gender wage gap, none of these stats may be the result of oppressive gender discrimination against women, but may reflect different work and lifestyle choices that men and women tend to make. However, the authors only seem to see gender discrimination as the main possible factor.

Look at what they authors write about their own study:

“To test whether people evaluate products that women sell as less valuable than the same products when sold by men…” (7)

“One explanation for this disparity is that people tend to assign a lower value to products when sold by women rather than by men.” (7)

with new identical products, the gender price gap cannot be attributed to the quality of the product but rather to beliefs about gender.“ [emphasis added] (6)

"One mechanism that could generate disparities between women and men in product markets is the attribution of a lower value to a product sold by a woman than to a product sold by a man.” (6) [emphasis added]

“Women’s willingness to pay more than men for identical products may be related to the cultural beliefs and stereotypes about women being less instrumental and agentic than men.”(4) [emphasis added]

Note the authors strangely also found it didn’t really matter much for sellers for buyers were men or women:

“The gender of the participants in the experiments did not affect the final price, nor did it affect the differences between the prices of gift cards sold by a woman and the prices of gift cards sold by a man” (6)

So women are willing to pay more, but ultimately devalue female sellers just as much as men? This is probably the sort this thing feminists would write off to supposed “internalized misogyny”.

I don’t think the author’s base findings are trustworthy. However, for sake of argument, even if we assume that both men and women give lower bids on products from female sellers, that doesn’t necessarily imply anything negative toward female eBay sellers, much less women as a group.

Instead of claiming this shows that people devalue women, we could just as easily claim it shows that people view women as more generous and less greedy. This interpretation could be bolstered by the study’s finding that female seller’s tend to get more positive feedback even though they tend to have slightly less experience as sellers.

We could also consider that male sellers might tend to do certain things that female seller don’t, which make their products more attractive. Male sellers might get higher bids, not because people devalue female sellers’ products, but because male sellers tend to write more complete product descriptions that buyers respond to even when they don’t know the sellers’ gender.

Also note that authors largely brush off their finding that:

“Note that although women tended to buy more from women than from men, the effect of the interaction between being a woman seller and being a woman buyer was insignificant.” (4)

If the authors were of a different mindset, they could talk about the anti-male bias of female eBay buyers (more likely female sellers just sell more products female buyers want to buy; even more likely the study is all garbage anyway).

I don’t think any of these alternative interpretations are very accurate either, since I doubt the basic methodology of the study. My point is there are alternative interpretations. However, the feminist bias of the study’s authors causes them to push readers toward the very dim oppressive view of women’s place in the society espoused by feminism.

This is a feminist grievance study

I am very skeptical of how this study attained and manipulated it’s data. However, the Achilles heel of the study is that it depends on the assumption that eBay buyers accurately and consistently guess the gender of eBay sellers, which the study data itself suggests isn’t true.

It is very obvious the the study’s authors wanted to create a new politically useful “wage-gap” style statistic. The authors outright state they seek to “improve on the existing literature on the gender wage gap in the labor force.” The questionable narrative the study wants to push is that women are valued less than men by society, so people value the products women sell less.

This study isn’t really about eBay. There are no suggested solutions for how the already gender-blind eBay could fix this supposed problem. It’s just another ridiculous strained attempt to prove that women as a class uniquely live under a system of oppression. Another dubious statistical arrow for the feminist quiver. I wouldn’t be surprised if feminists show up on eBay’s doorstep soon and start making demands.