Repeat offenders and “publication inflation” have both contributed to the rising number of retractions over the past 10 years.

The 4-year sums of WoS records represent the time frame in which most (78.6%) of the articles that are retracted in any given year were published. Thus, changes in these sums represent changes in the size of the literature involved in most of that year’s retractions.1Ratio is the ratio of the 2010 value to the 2001 value for each column.