[andrewla]

There's a name for the phenomenon, Goodhart's Law [1]. Generally, once a metric is used for decision making, it no longer reflects the phenomenon that it is attempting to measure.

Making it much more complicated is the fact that methodologies vary from year to year, and are often created by averaging data from a number of different sources. I've become fairly convinced that almost all macroeconomic measures are so flawed that using them in almost any dimension (across time, between countries) is useless. It may be that everyone involved in creating the overall measures is acting in good faith, rather than some politically motivated conspiracy, but still in aggregate create data that does not reflect reality in a useful way.

The exceptions tend to be metrics that require skin in the game to shift -- the yield curve inversion or S&P500 index, for example, are fairly real, since shifting them would cost a ton of money.

[1] https://en.wikipedia.org/wiki/Goodhart%27s_law

[jedberg]

> The exceptions tend to be metrics that require skin in the game to shift -- the yield curve inversion or S&P500 index

I agree with you on the S&P500, but isn't the yield curve fairly easily manipulated by the Federal Reserve and their rate setting activities?

[rossdavidh]

I think the idea is that would impact more than just the politics, it would actually impact the rate of interest banks and others pay, so there is incentive to disregard the politics at least to some degree.