Sold! Painting Prices and Top Incomes From 1972-2013
MetadataShow full item record
This paper empirically investigates painting auction price trends from 1972-2013 and their relationship with income changes for the top 1% of earners in the United States and Japan. This researcher algorithmically collected 5.4 million art auction results between 1922 and November of 2014, representing a uniquely large dataset of individually compiled art auctions. The data was narrowed paintings that sold more than once between 1972 and 2013, giving 94,662 auction results. Subsequent sales of the same painting were matched and their prices differenced to conduct a repeat-sales regression (RSR) and construct a painting price index. This paper finds that paintings have an average real log return of 1.96% and a standard deviation of 15.22 percentage points. A dataset was constructed combining the painting price index with annual income data for the top 1% of earners in the United States and Japan to perform a time-series analysis of the relationship between changes in top incomes and the financial log return of paintings. This paper finds evidence that changes in top incomes in the United States and Japan were directly related to painting log returns between 1972 and 2012. The findings also suggest that changes in the top 0.5-1% of incomes are more strongly related to painting returns than the incomes of the top 0.01% of earners in both the United States and Japan.