This isn’t as dreadful a prediction as it seems, this idea that the risk, in the next 6 months, of a US recession is only 21%. This while the actual recession started a week or two back. Because Moody’s does point out that this takes no account at all of the coronavirus, a black swan event entirely outside the usual models:
The odds that the U.S. will be in recession in the next six months remained low in February, but the model incorporates data that don’t capture the impact of COVID-19 on the economy. The odds of a recession occurring in the next six months were 21% in February, but significantly underestimate the risks. Many recession probability models that are based on warning signs for classic causes of a U.S. recession will understate the risks because COVID-19 is a black swan event.
Well, Har, Har, eh?
And we could even go off and shout about how market economics doesn’t work we need planning, state action, right?
Except that is to take exactly the wrong information away from this little Boo Boo. For what is that model being used to predict recessions? Exactly the sort of model that an economic planner will use because it’s the best, state of the art, guess we’ve got about what the risks of recession are. That is, our planner will be insisting that nothing need be done because the basis of his plan says that the recession which is here doesn’t in fact exist and won’t.
It is the market out there which is telling us that this basis of planning is wrong. As it always is the market that sticks that inconvenient stick into the spokes of economic plans. As MacMillan put it, events dear boy, events.
Well, yes, we know there always are events. But how do we recognise when there are? When market prices diverge from the plan of course. This is no different from the manner that it is variances between corporate plans and outcomes that we monitor. It is when reality diverges from plan that we’ve got to pay attention.
Which gives us two things to think about. The first is that plans are always going to be late to hte party when there are events. And also that we’ve got to have markets and their price system to know when the plan is at divergence with reality. You know, when the planners have fucked up by not recognising the black swan.
Even this isn’t a new point. Kantirovich himself pointed out that a properly planned economy would still need a market one out there somewhere so they’d know what the prices should be set at in the plan.
And another computer model says reactions in the lab to impurities we detect in the air will result in “runaway” “global warming” solved only by UN control of everything. And another says our hospitals will be clogged by COVID patients. These “studies” are not peer-reviewed because peers rarely look at the code nor detect the ways the coder biased it – including, here, simply ignoring what we know to be happening!
Peer review isn’t really supposed to be much of a check. It’s more about obvious errors, inconsistencies, completeness and a sense-check. It is replication that matters, which is why the Replication Crisis is a crisis. Climate science seems to have not only corrupted peer review, but to have abandoned replication entirely. Every paper is simply accepted as true. As Ionnadis and others have shown, all science is riddled with errors, fake data, bias, and so on, such that at least 50% of published papers (peer reviewed) are wrong. Except in climate science.