Goodhart's Law, or duck-shoving

While I have talked briefly about Goodhart's Law before (here), Charles Goodhart was a chief economic advisor to the Bank of England. He looked at central banking institutions' inflation-control methods, through restricting and easing of circulating cash in the system. Goodhart's Law was an aside said in a 1975 paper that he wrote, saying "Ignoring Goodhart's law, [i.e.] that any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes", then he went on to explore the actual impacts of risk management around monetary policy (Chrystal & Mizen, 2003, p. 222).

Goodhart was apparently noting that once banks focused on a particular inflation target, they changed their behaviour - for example, creating and using new instruments not counted in the measure - making the original 'money supply' statistic useless for policy-making. Experts say that what Goodhart (1975, as cited by Chrystal & Mizen, 2003) meant by this self-proclaimed law is that we shouldn't put our blind trust in statistics once they become our goal, because all we did was observe a normal pattern of behaviour then turned our observation into a target. It is A target; but who knows if it is THE target? Or the RIGHT target?

It is a good point. Because as soon as we have a goal, we don't want to 'spoil' our ability to meet the target. So we change our behaviour... and the observation no longer holds.

There are two other very similar phrases:

• Anthropologist Marilyn Strathern simplified Goodhart's Law to "When a measure becomes a target, it ceases to be a good measure" (Mattson et al., 2021, p. 2), which I don't feel has quite enough sting; and
• Social scientist researcher Donald Campbell said that "The more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor" (Mattson et al., 2021, p. 2). Ouch. That has sting. We could reframe this as the more we use a stat as a marker, the more we are likely to try to game it. 

How about this for a current example. Let's think about hospitals who say surgery cases are only allowed to be on a waiting list for 6 months; surgery is expected to have taken place, the case resolved. So to meet that statistic, at the end of 6 months, the surgical department sends the patients back to the specialist for reassessment. The clock resets, and the countdown to six months begins again. The statistic has been met: regardless of the surgery not having been done, the person's quality of life is still poor, but the statistics look great.

I would call this duck shoving. Rearranging deckchairs on the Titanic. 

Or Goodhart's Law in action ;-)

Sam

References:

Chrystal, K. A., & Mizen, P. D. (2003). Chapter 8: Goodhart's Law: Its Origins, Meaning and Implications for Monetary Policy. In P. D. Mizen (Ed.), Central banking, monetary theory and practice: Essays in honour of Charles Goodhart (pp. 221-243). Edward Elgar.

Goodhart, C. A. (1975). Monetary relationships: A view from threadneedle street in papers in monetary economics. In Reserve Bank of Australia Conference Proceedings, Papers in Monetary Economics (Volume 1, pp 1-20). Author.

Mattson, C., Bushardt, R. L., & Artino, A. R., Jr (2021). "When a Measure Becomes a Target, It Ceases to be a Good Measure". Journal of Graduate Medical Education, 13(1), 2–5. https://doi.org/10.4300/JGME-D-20-01492.1