How Often Do Economists Commit Misconduct?

schwit1 (797399) writes A survey of professional academic economists finds that a large percentage are quite willing to cheat or fake data to get the results they want. From the paper's abstract: "This study reports the results of a survey of professional, mostly academic economists about their research norms and scientific misbehavior. Behavior such as data fabrication or plagiarism are (almost) unanimously rejected and admitted by less than 4% of participants. Research practices that are often considered 'questionable,' e.g., strategic behavior while analyzing results or in the publication process, are rejected by at least 60%. Despite their low justifiability, these behaviors are widespread. Ninety-four percent report having engaged in at least one unaccepted research practice."

That less than 4% engage in "data fabrication or plagiarism" might seem low, but it is a terrible statistic . ... 40% admit to doing what they agree are "questionable" research practices, while 94% admit to committing "at least one unaccepted research practice." In other words, almost none of these academic economists can be trusted in the slightest. As the paper notes, "these behaviors are widespread.""

Comments

Re:What does it matter?

[lgw]

It's worth noting a pattern of "adjusting" the data that predominately favors the leading theories. That's some seriously questionable stuff in an field. Fortunately for the climate change guys, you can just ignore the ground station data entirely and still have a reasonable conversation about this stuff.

I'm far, far more concerned with constant tuning of the models to meet the data then vice versa. That may sound backwards, but ask anyone who's made a model to predict the stock market based on fitting his model to all historical data how that worked out. Descriptive power is not a significant reason to expect predictive power from a hypothesis (necessary, but very far from sufficient). And every time you tinker, you reset the clock on knowing if you have any predictive power.

It takes many years to test the predictive power of a climate model. 15? 20? Depends on who you ask, but the better part of a career. The models from 15 years ago failed pretty hard, prediction-wise. We're a long way from anyone having the right to be arrogant about this stuff, and every time someone adjusts their model to make it match observations, that's one more model reset, one less chance to move from hand-wavy descriptivism to a tested theory.