Beauty or money? Statistical analysis

There are many studies investigating relationship between paintings’ meta information (author, age, etc.) and their prices, however, there is limited research on how people’s aesthetic perception of paintings is related to their prices. To bridge this gap, we designed a website (pollart1000.com) collecting survey responses on people’s opinions regarding aesthetic values of paintings; survey and data were complemented with price information. Comprehensive statistical analysis of that data is presented, including scoring of the best and worst works, correlations between variables, typology of respondents, and more. The results quantify relationships between aesthetic values and art prices. Particularly, we showed that aesthetic preference is inclined to figurative and innovational rather than to abstract art and that relations between price and aesthetic values are rather weak except of special cases.

American Art, “Experts,” and Auction Institutions

Estimates made by “experts” regarding the value of a painting to be offered at auction are evaluated as to whether or not they provide a “fair” or “unbiased” range. Using a sample of eighty American artists over the period 1987–2013, evidence is found that in general art prices are underestimated, but that very high priced art may be overestimated in the interests of profit-maximizing auction houses; we also find that the higher the low estimate, the less likely a painting is to sell, and the wider the range between high and low estimated, the more likely it is to sell; that if a painting does not sell on first offer, its hammer price will be about 60 percent lower if it sells on the next offering (it is “burned”); and that the rapid rise in buyer’s premiums charged by auction houses has had, in combination with other factors, a reduction in auction house revenues.