The hemline index

Many years ago I remember reading a sci fi book where the author talked about the intersection between the price of gold, hemlines and men's beards (Heinlein, 1982), specifically:

The main character has been asked about "the relationships between men's beards, women's skirts, and the price of gold. I had ceased to wonder at silly questions; around Boss anything can happen. But this one seemed super-silly. Why should there be any relationship? [...] But I had learned not to ignore questions merely because they were obvious nonsense; I tackled this one by calling up all the data I could, including punching out some most unlikely association chains. I then told the machine to tabulate all retrieved data by categories.

"Durned if I didn't begin to find connections!

"As more data accumulated I found that the only way I could see all of it was to tell the computer to plot and display a three-dimensional graph. And that looked so promising that I told it to convert to holographic in color. Beautiful! I did not know why these three variables fitted together but they did" (Heinlein, 1982, p. 224).

I remember at the time of reading that this sounded suspiciously like a spurious correlation to me, but I was reminded about this recently via a YouTube video thumbnail (while still thinking it sounded like crap). So I turned to GoogleScholar, searching for a connection between skirt length, gold and facial hair. I found nothing.

However, I did find a 1926 "hemline" index theory, or HIT, proposed by a US economist, George Taylor (Kim & Kim, 2012; Škrinjarić, 2020) where fashionistas apparently suggest skirts shorten during economic contractions, due to "designers and retailers [needing to] create and display unique styles of clothing in order to attract consumers’ attention" (Kim & Kim, 2012, p. 1).

While that sounds plausible, there is apparently no evidence to support it: "there is no statistical evidence that economic state can predict the future changes in skirt length" (Škrinjarić, 2020, p. 6) and "After completing several regression analyses which revealed no statistical significance, and a thorough observation of graphed movements between hem lengths, overall stock returns, and fashion companies’ stock returns, it is clear that there is no causal relationship indicated between these variables" (Humes, 2010, p. 59).

Why? Well, what we wear tends to be influenced by many "economic, social, cultural, social and environmental effects. Today, fashion is not just a statement of someone’s economic status" (Škrinjarić, 2020, p. 7). A good point. And even more likely: in a downturn at present, we buy nothing as we already have so much.

All claims need to pass the sniff test.

Sam

References:

Kim, J. E., & Kim, J. (2012). Skirt length theory: The impact of perceived financial status on skirt length preference. Asia-Pacific Advances in Consumer Research, 10, 23-25. http://www.acrwebsite.org/volumes/ap10/apacr_vol10_1011042.pdf

Škrinjarić, T. (2020). Hemline Index Theory: empirical analysis with Google data. International Journal of Fashion Design, Technology and Education, 13(3), 325-333. https://doi.org/10.1080/17543266.2020.1799079

Heinlein, R. A. H. (1982). Friday. Ballantine Books.

Humes, G. M. (2021). Hemline Theory: Testing the Relationship Between Fashion Trends and the Stock Market [Honours thesis, Kent State University]. https://etd.ohiolink.edu/acprod/odb_etd/ws/send_file/send?accession=ksuhonors1620591215357509&disposition=inline