The lipstick index, as reported by NYTimesAfter Estée Lauder chairman Leonard Lauder revived the truism that the cosmetics industry is recession-proof and actually suggested that the “lipstick index” could indicate economic fluctuations in 2001, writers in the business section and the magazine, including pop economist Adam Davidson, simply repeated it and attributed it vaguely to economists. (So did Maureen Dowd, but in pure Dowd fashion.) Only in the Styles section did reporters ask economists about it and look at some numbers, much less correctly attribute the concept’s revival to Lauder. Also, they noted that it had “been largely discredited.”
Critics have pointed out that Dr. Keys violated several basic scientific norms in his study. For one, he didn't choose countries randomly but instead selected only those likely to prove his beliefs, including Yugoslavia, Finland and Italy. Excluded were France, land of the famously healthy omelet eater, as well as other countries where people consumed a lot of fat yet didn't suffer from high rates of heart disease, such as Switzerland, Sweden and West Germany. The study's star subjects—upon whom much of our current understanding of the Mediterranean diet is based—were peasants from Crete, islanders who tilled their fields well into old age and who appeared to eat very little meat or cheese. As it turns out, Dr. Keys visited Crete during an unrepresentative period of extreme hardship after World War II. Furthermore, he made the mistake of measuring the islanders' diet partly during Lent, when they were forgoing meat and cheese. Dr. Keys therefore undercounted their consumption of saturated fat. Also, due to problems with the surveys, he ended up relying on data from just a few dozen men—far from the representative sample of 655 that he had initially selected. These flaws weren't revealed until much later, in a 2002 paper by scientists investigating the work on Crete—but by then, the misimpression left by his erroneous data had become international dogma.
We now demand a higher standard of proof for predictions. We now need to optimize our operations and products based not on popular opinion, but on probability. We now demand data.
Despite the ample statistical analysis, the study hits a few stumbling blocks. Firstly it only takes into account analytical intelligence, disregarding creative and emotional intelligence. Moreover, it could be argued the study is not representative, as over 87 per cent of the participants involved in the various studies were from the US, the UK and Canada. Also, the predominant religion is the study is Protestantism, while other beliefs are not investigated.