“Follow the science,” we are told, especially the junk science that climate alarmists invent. I recently debunked a piece of junk climate science whose alarmism was featured in the Wall Street Journal, Bloomberg, CNN, CBS, and elsewhere. The junk science was produced by the Federal Reserve. Fed officials claimed that warming will cut economic growth by a third, but my simple statistical analysis showed their results were within the margin of error and that minor improvements and new data flip their result!
Mainstream economics shows that warming will have minor economic effects compared to the economic growth we expect over the next century. That is a problem for the climate lobby, which unfortunately includes the Fed. Since economic growth will swamp the economic effects of global warming, the Fed set out, it seems, to prove that warming will reduce growth. The fact that Florida, on average 26 degrees warmer than Michigan, has grown faster didn’t faze them.
The Fed isn’t the only institution to fall into chicanery. The study was published in a peer-reviewed academic economics journal, given wide media coverage, and cited in a congressional report to justify the Green New Deal. Bogus research makes big splashes. But when it is debunked, there isn’t a ripple.
The best economic model, validated by a Nobel Prize for William Nordhaus, shows that if nothing is done to reduce emissions, warming will reduce world GDP by about three percent by the year 2100. If global GDP continues to grow at the rate it has been growing, then the world in 2100 will be five times richer than it is today. A three-percent reduction in GDP would make us 4.8 times richer instead of 5.0 times. Not exactly catastrophic! Mainstream economics doesn’t deny climate change and accepts that some policies to mitigate it might pass a cost-benefit test. But it does not predict a climate apocalypse, even if we do nothing.
To support apocalyptic predictions, the Fed, it seems, went after growth. Their study looked at the relationship between seasonal temperatures and growth, state by state and year by year, from 1957-2012. Higher summer temperatures were statistically associated with lower growth, while higher fall temperatures were associated with higher growth, but the fall effect was smaller than the summer effect. There was no statistically significant association between growth and temperatures in the winter and spring, so they subtracted the summer effect from the fall effect and concluded that the overall effect of higher temperatures was to lower economic growth.
The problem is that even if two estimates are statistically significant individually, their sum is not necessarily significant. For example, if moving to Florida would increase one person’s income by $1,000, plus or minus $100, but would lower his spouse’s income by $1,100 plus or minus $100, it is reasonable to say that the move will almost certainly raise the husband’s income and lower his wife’s income, but it is not nearly as clear that their total income will be lower. Even though the wife’s expected loss is higher than her husband’s gain, the odds that their overall income will be higher are substantial.
In addition, I found that removing California from the sample switched the result to an increase in overall growth from a temperature increase, though without statistical significance. Using different data that measured the same things, the sign of the effect also flipped, though again without statistical significance. A statistically insignificant result that changes sign when estimated with different samples is exactly what we should expect if no true relationship exists. My work was published in Econ Journal Watch, another peer-reviewed academic journal that specializes in critiques of articles published in other journals.
That the Fed engages in politically biased research should not surprise us. A recent study of Fed personnel shows that the institution is very lopsided, politically.
Following the habits of good science is a good idea, but following the dictates of people who call themselves scientists is not the same thing. Climate change might be real, but there are good reasons to reject calls for draconian policies that fail cost-benefit tests.