Readers mirror the Caldecott judges

Noodling around in the back of my head has been the problem of proving bias in award selection. I have posted elsewhere the perennial controversy arising from roughly 65% of the Caldecott (children's illustration awards) going to male illustrators. Interestingly, the fact that roughly 65% of the Newbery awards (children's story books) go to female authors does not excite comparable concern. I am agnostic about whether this is an important issue or not. Someone at some point will have to gather the necessary data to indicate whether this is even a real problem. If the awards are going in proportion to the genders of the accomplished authors and illustrators, then there is no real evidence of bias and discrimination in the way the awards are being allocated. In other words, if 65% of the full-time illustrators are male, then the fact that 65% of awards goes to males is not particularly alarming.

But how do you show what the proportion of full-time illustrators might be by gender? That data doesn't seem readily available.

I did come across Elizabeth Bird's periodic listing of the top 100 illustrated children's books. Her column is in the School Library Journal and some 1,000 readers send in their nominations. This data set has the drawback that it is across time and not the most recent year so it describes readers assessment of the full portfolio of illustrated books and not what is new to the market. The readership of SLJ likely follows the demographic of the profession, i.e. 84% of librarians are female, so there is that skew to take into account. Of course the participants are self-selected so that is another mystery variable. The benefit is that it is likely a knowledgeable base of participants, which helps.

Turning a blind eye to all those issues, my question was: Would 1,000 self-selected, probably mostly female, probably well informed, participants comes up with a list of books with an illustrator gender ratio materially different than do the Caldecott judges? The answer is, No. The Caldecott's go male 69% of the time. This 100 Best list is 68% male illustrator.

Doesn't really resolve anything because there are so many potentially confounding issues. None the less, it does moderately suggest that there is likely something else going on than straightforward bias.