I saw two statements written about economics over the weekend that could easily have been written about climate as well. These are both complex systems where researchers try to link one output variable (e.g. global average surface temperatures or economic growth) to one input variable (e.g. CO2 or government spending).
Via Cafe Hayek, here is Bob Gelfond discussing Keynesian multiples
When it comes to the “evidence” demonstrating the magic of the Keynesian Multiplier, what we see, in fact, is merely careful curation of statistical flukes on a grand scale over decades. Economist Ryan Murphy, who runs a project called govtmultiplier.com that attempts to catalog scholarly measurements of the Keynesian Multiplier, has categorized and analyzed 128 papers on the subject. Only four papers even attempt to include this kind of statistical test, and none of these validate the original results, meaning simply that none of them prove the Keynesian Multiplier actually leads to more dollar-for-dollar economic growth. And this is after these models are ginned up to make their theory look as good as possible. If attempts to employ macroeconomics purport to be science, they must boldly make predictions about the future, not rummage around for convenient data from the past. But no peddler of the Keynesian Multiplier has been able to make demonstrable predictions borne out by the test of time.
Morgan Housel on economic data, but applies to climate without changing a word.
Ideally we’d have 500 years of unimpeachably perfect data. In reality we have about 50 years of so-so data. If we had the former, we’d learn that so much of what we’ve learned from the latter is wrong and incomplete.
Update: Here is a third bit from Arnold Kling in the same vein:
Sometimes, I think that there are macroeconomists (Krugman is not the only one) for whom there is no path of economic variables that could ever contradict their point of view. They remind me of the climate scientists who tell us that Buffalo’s Snowvember came from global warming.
Macroeconomics is infinitely confirmable because of its high causal density and lack of controlled experiments. The macroeconomist has enough interpretative degrees of freedom to twist any pattern of economic activity to fit his or her priors.