Do authors who retract papers end up cited less often? Depends how eminent you are

St. Matthew, by Frans Hals, via Wikimedia

A picture of the downstream effects of retractions is starting to emerge.

In a new working paper, economists at the University of Maryland, the University of Rochester, and Northwestern University focused on work by teams of scientists. Their main findings:

After retraction, ordinary authors experience large citation losses to their prior work, especially when working with an eminent coauthor. Eminent authors, by contrast, show little citation losses to their prior work, regardless of the status of their coauthors. Furthermore, the presence of coauthors with no prior publications predicts that established authors experience smaller citation losses.

The researchers looked at several explanations for the findings, including:

…one may argue that eminent authors typically lead in the conceptual design of the research rather than in the technical analysis, where problems are more likely to emerge. In this view, eminent authors may receive less blame when retraction occurs because they are seen as unlikely to be responsible for the relevant tasks. To test this idea, one can examine citation effects based not on author status at the time of the retraction but at the time the research was conducted, when task allocation would be determined.

When they ran that analysis, however, the researchers found that “task allocation does not appear to be a key explanation for our main finding.”

The authors conclude that there is a “Matthew Effect” at work in scientific credit:

Not only do the rich get richer, when riches are to be had, but the poor get poorer when catastrophe strikes.

We learned about the paper from Freakonomics, which kindly referred its readers to us to read more about retractions. The new work joins other research finding that retractions are linked with a 65% decrease in citations to the papers themselves, and to a 5-10% decline in citations to related papers along with a decline in funding.

3 thoughts on “Do authors who retract papers end up cited less often? Depends how eminent you are”

  1. Does the Matthew effect reflects hierarchical, non-democratic levels of organisation in institutions that lack transparency? It is the antithesis of the naval custom, where the captain has clear responsibilities and is the last off the ship. A corollary is that it is the antithesis of “leadership”, so suggesting that many of these senior people fail to deliver significant leadership (that is, they were first into the lifeboat). To become senior, they were promoted and this in part because their institutions perceived that they had leadership? To me this all chimes with the over promotion of men and the under promotion of women in science.

  2. Um, I hate to say it, but that ‘working paper’ is pretty abominable. I’ve heard bad things said about the economics field. And I’ve seen it written that where science uses complexity to make things clear, economics uses complexity to obscure the facts. This paper seems to bear that out.

    There’s no chance I’ll get through the 48 pages. And if they have something worthwhile to say about retractions and scientific careers, I seriously doubt it can be extracted from their mathematical models.

    If this blog post is correct, and the ultimate point is that senior scientists are more buffered from the effects of a retraction, then this paper is working hard to make that as unclear and difficult to understand as possible.

  3. A cynical view.
    But what this also says is that a retracted paper continues to amass citations – even if only at 35% of its unretracted rate.
    Better therefore to publish quickly – even with mistakes – and then retract later if and when the mistake is found.
    “Publish quickly and be prepared to retract (if bad)” could then then be the strategy which leads to a maximisation of citations!!!!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.