At the end of an article by Kevin Roose in New York about the effects of a debt-ceiling breach:
The bottom line: A debt-ceiling breach would be very, very, very bad.
Keep in mind that these are all hypothetical scenarios. Reality could be better, or much worse. The truth is that while we sort of know what a government shutdown would look like (since it’s happened in the past), we have no idea what chaos a debt-ceiling breach could bring. If, in a month, we reach the X Date, run out of money, and are stuck in political stalemate, we’ll be entering truly uncharted waters. And we’ll be dealing our already-fragile economy what could amount to a knockout blow.
This is an example of something common: Someone who has never correctly predicted anything (in this case, Roose) telling the rest of us what will happen with certainty. If Roose is repeating what experts told him, he should have said who, and their track record. Roose is far from the only person making scary predictions without any evidence he can do better than chance. Here is another example by Derek Thompson in The Atlantic.
The same thing happens with climate change, except that it is models, not people, making predictions. Models that have never predicted climate correctly — for example, none predicted the current pause in warming — are assumed to predict climate correctly. We are supposed to be really alarmed by their predictions. This makes no sense, but there it is. Hal Pashler and I wrote about this problem in psychology.
A third example is the 2008 financial crisis. People who failed to predict the crisis were put in charge of fixing it. By failing to predict the crisis, they showed they didn’t understand what caused it. It is transparently unwise to have your car fixed by someone who doesn’t understand how cars work, but that’s what happened. Only Nassim Taleb seems to have emphasized this. We expect scary predictions based on nothing from religious leaders — that’s where the word apocalypse comes from. From journalists and the experts they rely on, not so attractive.
I don’t know what will happen if there is a debt-ceiling breach. But at least I don’t claim to (“very very very bad”). And at least I am aware of a possibility that Roose (and presumably the experts he consulted) don’t seem to have thought of. A system is badly designed if a relatively-likely event (debt-ceiling breach) can cause disaster — as Roose claims. The apocalyptic possibilities give those in control of whether that event happens (e.g., Republican leaders in Congress) too much power — the power to scare credulous people. If there is a breach, we will find out what happens. If a poorly-built system falls down, it will be much easier to build a better one. Roose and other doom-sayers fail to see there are plausible arguments on both sides.
The climate models are written by people, so it’s still basically people making the predictions. It just sounds more official to say “the model says”. The model says what the people who built the model want it to say. And with many models, it’s GIGO.
If a poorly-built system falls down, it will be much easier to build a better one.
While I agree with some of what you say, not all.
(1) We can know from experience and history how to treat diseases or financial crises even if we don’t fully understand fundamentals.
Science is never-ending. In that sense, there is always more to learn, which means we are all, always, ignorant about everything.
(2) I can fail to predict earthquakes, but still can know how to send emergency relief to cities hurt by them. I can fail to predict a bout of the flu, but still know how to treat symptoms. I can fail to predict financial crises, but still can know how to respond to one.
(3) History has plenty of examples of default on sovereign debt. Books of case histories have been written. We can see how Greece is doing right now. We can look at recent decades in Argentina. We can look at Russia in 1998. But there has not been a default on sovereign debt by a reserve currency ever that I know of, perhaps with the exception of the collapse of the Roman Empire. It is like playing Russian roulette: better not to test the hypothesis outside of a laboratory.
(4) If someone with a bad record of making predictions warns you not to step in a bear trap, it might be wise to listen.
“I can fail to predict earthquakes, but still can know how to send emergency relief to cities hurt by them. I can fail to predict a bout of the flu, but still know how to treat symptoms. I can fail to predict financial crises, but still can know how to respond to one.”
You are arguing because you know how to send emergency relief and treat flu symptoms you know how to respond to a financial crisis? I hope not. 1. Earthquakes. We don’t know what causes them, don’t pretend to. 2. Flu. Do you know how to reduce how often you get the flu in the future? Probably not. Yet that would be a good idea. In addition to treating symptoms. 3. Financial crisis. Surely it would be a good idea to reduce the chances of later crises. People who failed to predict the 2008 crisis are not good choices for doing this.
I was talking about reducing risk, not treating symptoms.
bingo!