Good art by women is cheaper

A rose painted by another name would cost more. In a new paper*, four academics show that art made by women sells for lower prices at auction than men’s, and suggest that this discount has nothing to do with talent or thematic choices. It is solely because the artists are female.

The authors used a sample of 1.9m transactions in art auctions across 49 countries in the period from 1970 to 2016. They found that art made by women sold at an average discount of 42% compared with works by men. However, auction prices can be distorted by a few famous artists whose output is perceived as extremely valuable. If transactions above $1m are excluded, then the discount falls to 19%.

…the researchers used a computer programme to generate paintings and randomly assign the results to artists with male or female names. They then asked participants to rate the paintings and ascribe a value. The experiment found that affluent individuals (those most likely to bid at auctions) attributed a lower value to works which the programme assigned to a woman. Clearly, this gap was unrelated to the artistic merit of the picture.

I don’t quite think that shows (non-statistical) discrimination, perhaps more convincing is this:

The average discount applied to the work of a given female artist was lowest in countries where women were more equal. (There are some exceptions to the rule, such as Brazil, where women’s art was highly rated.)

The good news is that the female discount has fallen over time. For transactions under $1m, the study calculated, the discount has dropped from 33% in the 1970s to 8% after 2010.

If you not wealthy and wish to collect art, buy textiles, they are much cheaper and very often the creators are women rather than men.  They are not as a whole less aesthetically valuable than paintings, except perhaps for paintings at the very very highest levels.  But within painting, prices for Gwen John vs. Augustus John have been in parity for some while now, same with Frida Kahlo and her male contemporaries, or Natalia Goncharova vs. her peers, the latter I check on a regular basis in fact I was just perusing them today.

Here is The Economist source piece.  Here is the original research., “Is gender in the eye of the beholder? Identifying cultural attitudes with art auction prices”, by Renée Adams, Roman Kräussl, Marco Navone and Patrick Verwijmeren.


One could take that to mean that there really isn’t any good art by women

The history of art makes it abundantly clear that there isn't any consistent and sensible criteria for "good art" (a salient example being the mostly unremarkable history of the Mona Lisa), which is why art was a good choice for the authors -- because valuation is arbitrary, your interpretation is ruled out.

Hi, mouse!

Hey! That's not me!

Besides Will looks like the anti- mouse!

I mean, they literally used random assignment of the same paintings... ?

That would require he have read even the synopsis ...

I wonder if the study distinguishes between dead artists and living ones? Seems to me that art auctions are a very small and very specialized niche when it comes to the economics of art. More of a East Coast, 1%, thing. But looking at the top Pop Singers for 2018 or the top box office stars, the discount is even greater.

"the researchers used a computer programme to generate paintings and randomly assign the results to artists with male or female names. They then asked participants to rate the paintings and ascribe a value"

The actual auctions was just a lead-in thought prompt; they did an actual test with randomized inputs.

Clever of them.

James Damore was right. Women have certain biological traits confirmed by science that prevent them from doing great things with their life. As a result we pay a fair price for their fair work. To pay an exceptional price, we need to see exceptional work.

"the discount has dropped from 33% in the 1970s to 8% after 2010".

Biological traits have changed over 40 years, definitively =)

So, you ... didn't read his white paper, I take it?

Because he said no such thing, of course.

Definitely. Several men were shown boobies and their ability to do math declined considerably. Imagine what effect actually having boobies must have on one's ability to do math.

We're all gender fluid now, so how can the authors possibly assign binary gender identifications to the fake artists?

In my country you can be arrested for not keeping your gender fluid to yourself.

I get the joke, but this is indeed an interesting point: what fraction of the male and female painters self-identified as something else?

Andy Warhol should be on the data skewing the men income up. I doubt the Damore fan above would take him as the epitome of heterosexual men biological superiority ;)

There's a conundrum with this research. If aspiring women artists are discouraged from painting then there's less quality female produced artwork and the gap is justified. Is this what the research wants to tell us?

Or does it want to tell us that young women have equal opportunities to produce painting and that buyers simply, after encouraging female artists wholeheartedly, don't want to pay them as much?

Read it again: it randomly assigns (gendered) names to paintings.

That was only one part of it. The other part examines real-world prices.

"The experiment found that *affluent individuals* (those most likely to bid at auctions) attributed a lower value to works which the programme assigned to a woman." (emphasis added)

I thought the "affluent" part was a strange way to state a result. What about the non-affluent? Did the authors start out limiting their sample to the affluent because the affluent are most likely to bid at auction or was this data mining something the authors decided to do after the fact? Surprise, surprise --- it appears to be the latter.

If I'm reading Table 11 correctly, the authors actually find that "Female Name" *adds* (insignificantly) to the value of a painting in their overall original regression (Table 11, Panel A, column 1). However, they run 13 different regressions, 7 over the whole sample and 6 over people that visit museums, with different "interaction terms". The quoted result is from Table 11B, column 6, which the authors explain corresponds to a 6% gender discount. What they don't tell you is that in that same regression males, mature people (age 45+), and the college educated add a *premium* to female art, with the mature premium being statistically significant. I guess we should commend the authors' persistence: they seemed to have done a lot of work to find the result they were looking for. If at first you don't succeed, try another regression, try another regression, again.


My impression from reading art history is that there were more than a few women who were famous painters in their own lifetimes going back to the 17th Century. But they tended not to stay famous after their deaths, unless their memory is specifically revived for being a woman.

I have a model in mind about why it's harder for a woman painter to stay famous:

- Enduring fame for a painter is largely due to influence on subsequent painters (e.g., Cezanne remains a huge name because of his influence on subsequent huge names like Picasso and Matisse)

- Most famous painters are men, and men tend to be more influenced by other men painters than by women painters

- So, very good women painters tend to have fewer famous followers than equally good men painters.

That seems question-begging to me. Very few artists enjoy enduring influence, and very few artists were women. I don't think there's any inherent reason for women to have fewer followers. And I don't think women's reputations have been less lasting. There really have been only a few major women artists, but some have retained their fame (Rachel Ruysch, Vigee Lebrun). Anecdotally I just don't see evidence for anti-female discrimination in the art market, at least for old masters.

Something to remember: Prior to the 1900s daily life required tremendous effort. Getting water, for example, had a fairly high (by modern standards) body count--you were often going to a river and scooping water out with a bucket, and if you fell in and hit your head you were gone. Other tasks that we take for granted or even have eliminated entirely were likewise much more laborious than we generally consider.

Also, in the past many tasks were gendered. The dairy was a woman's domain; men weren't allowed in, or were only allowed in if invited. Laundry and house cleaning were also women's work. This involved hauling water, cooking, mending cloths, etc.

Note that while nobles had different tasks, they were similarly gendered. Noble women wove and spun and sewed, though generally for higher-status folks (alter cloths, clerical garb, etc).

What this means is that a woman had much less time to do art than a man. A man could reasonably expect that many of these tasks would be taken care of by women in his life; a woman had to hire someone, adding expense to her artistic lifestyle, or had to do it herself, decreasing the amount of time available for artistic activities.

So this report is yet another piece of evidence that social “science” is agenda driven propaganda. I thinks that’s become the best default lens with which to view such results in future.

Thanks for reading the paper and tables. Bad on Tyler... good thing I read the comments.

"The average discount applied to the work of a given female artist was lowest in countries where women were more equal."

So, does that mean more gender-equal countries have actually gone past the point of equal? Presumably, the global market price of art reflects the purported female discount across all countries. If the female discount is lower in more gender-equal countries, then doesn't that mean people in those countries are willing to pay higher than market price for art simply because it's produced by a woman?

The article is oblivious to male pseudonyms used by female artists. The classification of gender looks far from perfect because there's no hard data in first place: "Overall, we are able to classify gender for just over 87% of the BASI painting data set."

They report they were not able to estimate the author's gender for 13% of the paintings. What they don't report is the uncertainty of establishing the gender of the 87% of the sample. It was possible, but how accurate?

The use of a male pseudonym is common in literature. I'd expect something similar in paintings. Some women chose income over personal fame and taking down the patriarchy.

As above, this finding doesn't seem to hold firmly. If men and women are not equally skilled or famous in the field of painting, it is entirely reasonable for gender to be used as a proxy for valuation. We use names as a proxy for all sorts of things in the absence of better information.

So maybe it just proves "affluent" art consumers rely overly on weak proxies to guide their evaluation. No surprises there.

What if the same experiment had been repeated with chefs? Or chess puzzles "this was composed by Kasparov / Polgar"? I suspect a similar result.

The closer result in greater gender equal countries also fits this narrative; those countries have generated more good / famous female artists than others. So the male/female proxy for value is weaker there.

This paper seems to join a growing list of "gotcha!" papers seeking to prove sex discrimination, which, under strict scrutiny, fall apart. I now scrutinise all such claims closely.

Is this what Cowen means by textile art:

I suppose art is quite different from music, where everyone listens predominantly to the output of female composers, and they are rewarded accordingly.

The small number of female composers is quite striking, especially compared to the huge number of female musical performers.

Most popular female musicians that I’m familiar with from pop culture perform, and generally appear in public and the media, underdressed. Underdressing is less prevalent amongst female composers.

"There are some exceptions to the rule, such as Brazil, where women’s art was highly rated."

I thank President Captain Bolsonaro's leadership for making Brazil a fairer, freer country.

The highest levels of contemporary art, in terms of auction prices, are dominated by homosexual male artists. In fact, a case could be made that "modern" art is predominantly a homosexual field. How does that figure into the study?

The work of a few female artists is quite valuable. Ultimately, why would anyone care? It's art, the most subjective subject on earth.

"Ultimately, why would anyone care? It's art, the most subjective subject on earth."
Where, therefore, the true values of a society, free from the compromises survival and the struggle for economical or military progress entail, make themselves known. You must have to tolerate Mendel or Einstein because you need good harvests or nuclear weapons, but who will make you like a watercolor sketches?

This suggests the possibility of an arbitrage trade from high-inequality countries to low-inequality countries, no?

Art prices are set largely by the extremely rich, such as Jho Low, who often became extremely rich by dubious means, such as by embezzling billions from the Malaysian sovereign wealth fund.

That people would pay more for art by men than by women, quality being equal, is only puzzling if you think that people buy expensive art for its quality. If, on the other hand, they are buying it to signal their high status, it might be more enlightening to ask whether and why male artists are considered to be higher status. One might also reasonably ask who are the buyers of expensive art. Mostly high status men?

I was going to make a similar comment. The auction of Jeff Koons' "Rabbit" that recently went for $91M suggests that much of the art market has little to do with aesthetics.

Besides, art is supposed to be about universality rather than the money, right?

I bet art prices follow an exponential/Pareto distribution. How much of the difference (measured as averages, for chrissake) could be explained by there simply being fewer women painters around - and especially early on?

"in fact I was just perusing them today." IP think 'perusing' is not the word you're looking for.

This is so most important article written by Tyler, Women's always been the BEST at everything. As a fashion designer, I truly agree with your point but that only limited to some industry. Nowadays the women's are on top as a fashion artist. Some industry dominates by men or some by women. It's always been balanced between gender. But study might have some different factor to show up. Great article.

Comments for this post are closed