Recently, on Twitter, there was some dismay that many pro-growth neoliberal ideas are not very popular. However, when it comes to policy discussion, good ideas are unpopular and popular ideas are bad!
This is a form of Berkson’s Paradox. If we consider all possible ideas, then good ideas will be more likely to be popular. However, we mostly consider ideas that are either good or popular, but not both.
Consider a standard 2x2 matrix of popular-vs-unpopular and good-vs-mad ideas:
Good and popular ideas are not debated (e.g., build and maintain the sewage system): they are implemented without much debate.
Bad and unpopular ideas are not debated (e.g., raise VAT to pay for a cow breeding programme to have cows with spots that resemble the flag): they are not done and nobody is arguing for them.
It’s only ideas that are good and unpopular or bad and popular that are part of the debate!1
Link of the Week
Day job link: A deep siamese neural network improves metagenome-assembled genomes in microbiome datasets across different environments, the manuscript describing and validating SemiBin
A surprisingly good video
/ht @dingding_peng
Tweets of the Week
Of course, we might not know exactly what ideas are good and which ones are bad, but, generally speaking, ideas that are part of the debate despite their unpopularity is because they are being pushed by policy experts and they will tend to be better than whatever is being done now. Carbon taxes are widely hated and yet are an idea that does not fully go away as policy experts love the idea. Obviously, past policy experts did recommend some things that we now recognize as bad (1960s urban planning).