4 ms·
The biggest problem is figuring out the truth from lies and mistakes and competing narratives. Maybe the pundit predicted that unemployment would go up, and som
by tsimionescu 6y ago
The biggest problem is figuring out the truth from lies and mistakes and competing narratives. Maybe the pundit predicted that unemployment would go up, and some say it did, but others explain that it's not actually up because people are leaving the workforce voluntarily, others claim that it was miscalculated in the past, yet others claim that it's actually a bad measure that shouldn't be trusted at all.
There are very few directly testable predictions. Even something as basic as whether the Earth is flat is hard to conclusively prove without enough background - how could you hope to understand the economy?
- cousin_it 6y agoI imagine predictions about unemployment would be based on numbers published by specific agencies. For example, see this Metaculus question: https://www.metaculus.com/questions/6296/us-unemployment-above-5-through-nov-2021/ https://www.metaculus.com/questions/6296/us-unemployment-abo...
- tsimionescu 6y agoThe question is, why should I believe those numebrs? "Unemployment" is a pretty technical measurement (the number of people who have not been employed for some amount of time X < T < Y but are still looking for a job) that is estimated through a pretty complex process (surveys and all of the statistical analysis that has to be done on those), so I can easily question: 1. The competency (are they measuring correctly?) or truth-fullness of the institutions (are they cooking the numbers?) 2. The relevance of the number (why not just look at people older than X who don't have a job?) 3. The relative importance - maybe unemployment has gone down, but underemployment has gone up, which some would claim is more relevant These are not even hypothetical concerns, these are all real complaints reasonable people have about the unemployment numbers in various places.