In a financial crisis, the theory goes, people cut back on gym memberships, eating out, and holidays. But on a bad day (and during a financial crisis, those bad days are legion), a consumer will still reliably reach for retail therapy: picking up a cupcake, a magazine, or, famously, a tube of lipstick.

The “l(fā)ipstick index” theory is a shaky, but compelling economic view coined just after the Sept. 11 attacks by Leonard Lauder, chairman emeritus of the international cosmetics company Estée Lauder. Lauder’s theory—backed up, reportedly, by the company’s own lipstick sales—was that in times of economic uncertainty, affordable luxuries like a striking lipstick shade help prop up select retailers even through downturns.

This financial crisis might be different. After all, why wear lipstick when no one can see your smile anyway?

The theory itself is widely cited, but conveniently hard to prove: reliable lipstick sales are not typically published as a stand-alone metric, leaving serious economists unable to measure this effect in an economic downturn. And there are indications the lipstick business is not quite as robust as we might think: In 2009, at the height of the last economic crisis, sales declined nearly 10%, according to research group NPD.

There’s more to the argument that cosmetics, as a whole, are a robust sector. A McKinsey report from May, citing statistics from Euromonitor, showed that sales for the global beauty and cosmetics sector have risen every year since at least 2005, including through the global financial crisis.

COVID-19, however, could be a different crisis altogether. McKinsey estimates that global beauty sales could fall by 20% to 30% in 2020, and by as much as 35% in the U.S. if there is another wave of COVID-19 later in the year. (On Monday, Dr. Anthony Fauci, the country’s top epidemiologist, said the country is still “knee deep” in the first wave.)

While cosmetics are expected to rebound quickly, the market may also be fundamentally changed, the consultancy warned. The closure of the department stores, emporiums, and airport boutiques that sell the bulk of cosmetics hit sales this year, pushing brands to sell on digital platforms.

But it has also shifted our relationship with makeup, at least for the time being. In April, just as lockdowns were going into effect globally, demand shifted. With little need to wear makeup outside, demand for cosmetics dropped, while demand spiked for skin care products, creams, and DIY home beauty treatments.

Lipstick in particular took a hit. Amazon sales in the U.S. tracked by McKinsey in the four weeks up to April 11 showed “l(fā)ip care and color” saw the steepest decline in retail sales of any segment, with sales falling 15% and prices falling 28%.Meanwhile, hair coloring product sales, likely spurred by salon closures, spiked by 172%. Nail care products, meanwhile, rose the most of any product area, by 218%, a trend that was also reported at European e-retailers.

If the “l(fā)ipstick index” has any economic weight then, McKinsey proposed, there was now an alternative: the “nail-polish effect.”