« Who Are Macro Experts? | Main | What Do Schools Sell? »

February 23, 2009


There is a vast space of possible things that can go wrong, so each plan will have to cover a pretty wide range of scenarios. Even to include a scenario as one with a plan will signal to viewers that you consider it more likely and/or important.

I doubt we'd find much comfort in the government's plans for an A-bomb in San Fran, and probably much less in their execution.

Robin, there are many possible causes but the convergent results of systemic risk, asset price declines, deleveraging, slows in money velocity, etc. etc. might not be that different across causes. Would Bear Stearns have gone much differently, if an atom bomb had gone off in San Francisco? Staring into one abyss would increase your ability to deal with many others without panicking.

Burger, comfort isn't the point. You can't expect to suddenly find a trick method in the actual event, so whatever uncomfortable thing it is you're going to end up doing, why not know it in advance?

Isn't there a bias something like: "If something actually happens, then people believe that it was foreseeable before it happened."?

I think you're correct (and also) governments should be judged more by how well they protect against extreme circumstances than how well they do in the no-disaster, short term.

However, an opponent of your ideas might make an argument that if they believed the probabilities of collapse implied by the market before the collapse, it would have been a waste of time and effort to make plans for such worst-case scenarios.

Here's a sick idea.

Call for volunteers. Ask people to volunteer for positions in that list, first-come, first-serve. The first to volunteer are the last to be fired if money becomes short.

When money becomes short, flip a coin, and if it comes up heads - reverse the list. Now the first to volunteer will be the first to be fired.

This reconciles the fundamental issue that those who rush to volunteer if they will be the LAST fired are probably those with the least confidence in the project. The coin toss roughly equalises the chance that anyone on the team will be fired, based on their position in the list, leaving them all roughly equivalent in terms of their motivation to keep the project running.

I don't know what good that is in the grand scheme of things, I just thought it made things interesting.

In "David's Sling", Marc Stiegler suggested that even just visualizing in advance what could go wrong would be helpful, in that if the worst happened you at least wouldn't be fighting disbelief as well as trying to make things work.

And does anyone think that the gov't would actually keep to a plan with special interests screaming for help. It's the same problem of trusting gov't; Congress can change any law any time it wants, you cannot rely on what they have said, or even what they have enacted into law.

"Why didn't various governments create and publish a plan for what they would do in the event of various forms of financial collapse, before it actually happened?"

Same reason the money that was supposed to go to flood control in New Orleans got spent on more-visible projects. What politicians do is get elected, not solve people's problems. How would devoting energy to this sort of plan win votes? The sort of person who would even consider this sort of thing wouldn't be running for office and wouldn't get campaign contributions if they did.

This would be like asking an evangelical Christian to prepare a plan for the eventuality that scripture is incorrect.
"Why would we want to prepare one of those, Sir? The free market works, it is proven"

Hardly any potential catastrophies actually occur. If you only plan for the ones that actually occur (say, by waiting until they happen, or by flawlessly predicting the future), then you save a lot of mental effort.

Also, consider the difference between potential and actual catastrophe regarding how willing you will be to make a desperate effort to find the best solution.

What if, realistically, your plan turns out to be "Do nothing of consequence, and let my successor deal with the mess?" You probably don't want to actually tell people that, even if that's the way the incentives turn out. (For example, my father works for a company that recently fired its CEO, after some decisions he made turned out to be disastrous. Unfortunately, according to the CEO's contract, they had to give him a huge severance package if they fired him. Ergo: the company is screwed, but he isn't.)

...the American government knows what it will do in terms of immediate rescue operations if an atomic bomb goes off in San Francisco.
Call on the experts who handled Katrina?

There are number reasons:

1) Economics, being the dismal science, has been very bad at prediction. Sometimes, some schools get stuff right, other times they get them dead wrong.

2) Due to the huge number of variables, it's hard to even make a plan

3) With Economics, what worked last time may not in fact work this time. Economics is not like Physics. Sometimes it /seems/ deterministic, other times it doesn't. It's also very difficult to test in a "lab-like" setting. There's no such thing as a clean-room country that you can test your theories on without outside influence.

4) Finally, we have the issue of the dominant economic ideology, Neoliberalism, close cousin of perennial favorite Economic Libertarianism that insists that *process legitimizes outcome*. Why mess with anything if it was /supposed/ to happen as a result of market forces? It may even be *fair* (but it makes a lot of people miserable). That's OK, though, since it's proper adjustment.

Correct me if I'm wrong, but wasn't the problem that the government's economic "experts" seriously believed that the worst possible thing that could happen was that the system needed a "liquidity" injection?

Well, now the major institutions are clamoring for "liquidity", but that has by now become a euphemism for "free money in the form of a loan that no sane person expects to be paid back."

I'm with Doug S. on this one. None of the "leaders" who be in a position to create such plans are in any way motivated to develop or publish such things, or to follow them after the fact.

For those that ARE interested and motivated to make such things, a better term would be "prediction" than "plan". And we're very bad at such predictions, as we always seem to underestimate the shortsightness of every participant, from government to private industry.

hello, I am a econ/math student at Binghamton university and I am looking for some introductory books on rationality. Ideally I would like to get my paws on the most cogent apologies and criticisms. If any of you guys have any suggestions you may get one more to your camp as this blog has whet my appetite for rationality. Thanks in advance to any with some thoughts.

Does the federal budget have a line item saying "Money for things that go wrong?" Because every year, something expensive goes wrong.

Kellen: "I am looking for some introductory books on rationality. [...] If any of you [...] have any suggestions [...]"

Cf. "Recommended Rationalist Reading."


I don't know about the rest of you, but I don't look to the military as a role-model often enough.

They can't publish such a plan because it would depend almost entirely on who was in power. In the US specifically, also because no executive has any incentive to plan for anything at all beyond their 8th year.

Hmm. It seems to me that Eliezer had a plan for the economic crisis, namely to keep his mouth shut about it, and when the crisis actually happened the pressure to talk about it apparently exceeded his expectations and he didn't follow through. That with the level of agency and unitary identity of a practiced human rationalist. Imagine how much less able to follow through he would be if he fell entirely under the sway of an arbitrary coalition of sub-motivations every 4 years (hmm... he does deny identity with the authors of his earlier papers).
Oh, and I have to ditto "Katrina".

Off topic but... did not anyone notice overcomingbias.com being named as one of the best 100 blogs by Times Online? Congratulations... though the write-up is a bit odd:


Category: SCIENCE

A strange, very much out-there, two-man science blog. “We want to avoid, or at least minimise, the startling systematic mistakes that science is discovering.” In part bonkers, it is nevertheless a window on another world.

End quote.

D. Alex


Original draft of this post contained the specific reasons I was trying to keep my mouth shut (everyone already talking, and seems like the sort of thing a non-professional would screw up) to which this post is a probable exception (haven't seen it pointed out, and the interaction with domain-specific expertise that makes everything I said about the psychology obviously wrong is at least not visible at a glance). Deleted because it wasn't really all that relevant.


Isn't there a bias something like: "If something actually happens, then people believe that it was foreseeable before it happened."?

Hindsight Bias and, to an extent, Taleb's Narrative Fallacy. This whole topic is quite Talebian. How do we plan for disasters we can't foresee? As Robin says,

There is a vast space of possible things that can go wrong, so each plan will have to cover a pretty wide range of scenarios.

While there might be a very wide range of causes for disasters, the possible effects are likely to be fewer. A government can plan for a crisis by making a shortlist of bad things and planning how to limit their effect, and can deal with the underlying cause on-the-fly. By analogy to emergency medicine, deal with easy-to-see life-threatening symptoms first and figure out the underlying cause later when the patient is stablised. In financial disaster planning, decide how to deal with known knowns like high unemployment, high interest rates, inflation and so on beforehand then figure out how to deal with the failing banks as you go along.

My impolite assumption is that you want to discuss having plans for when things go abyssal, but I don't think the current economical turbulence is a good example. What's happening is natural, a reaction to mindless tampering. What would that plan look like? "If we tamper too much, lets tamper some more in THIS way instead". The thought of stop tampering will not occur to them.

The underlying problem is that the ones in charge don't understand human nature. They they don't understand people, they don't understand themselves, and think economics is something unrelated to what makes people people tick. In a couple hundred years, I believe current economic theories will be regarded they same way as alchemy. If things doesn't go abyssal before then.(Is it not depressing to go around thinking like this all the time?)

I guess people prefer to think about how to make the worst case not happen rather than plan what to do when/if they get there. What are the possible bad outcomes, what lies outside our control and how can we minimize the risk. Emergency plans are good for things we have little control over, like natural disasters(disregarding that they can avoid building huge cities at stupid locations, and subsidize rebuilding them when they are wiped out). People like to think they understand economics and therefor can control it. Problem is that reality doesn't fit their model.

Oh, and I don't claim to have anything close to a complete understanding of people. But you don't need to understand everything to see that someone else doesn't get it either.

I guess people prefer to think about how to make the worst case not happen rather than plan what to do when/if they get there.


This may provide some insight on the topic:

Politico: Yap trap: Pols talk, markets dive

It was a perfectly reasonable question, and on the surface it seemed like a perfectly reasonable answer. But when Senate Banking Committee Chairman Chris Dodd went on Bloomberg TV Friday and mused about the possibility of bank nationalization, panicked investors sent the Dow plummeting a hundred points in the next hour.
Dodd’s casual remark and the not-so-casual consequences it caused were among the most vivid examples of a new Washington phenomenon. The city’s sudden status as the de facto world financial capital means that briefings and interviews that once would have passed with a yawn can create instant terror on Wall Street and Main Street alike.
“Sen. Dodd’s ‘nationalization’ comment wins the Oscar for most irresponsible remark by a U.S. official,” said Tony Fratto, a financial spokesman at the White House and the Treasury Department during the Bush years. “The last thing markets need is to have officials speculating about hypothetical policy choices.”

"The underlying problem is that the ones in charge don't understand human nature. They they don't understand people, they don't understand themselves, and think economics is something unrelated to what makes people people tick."

I suspect that if you could go back in time and confront the Federal Reserve on their lack of abyssal planning, the knee-jerk justification for it would be very similar to the current justification for Obama's vocal and continual support of the Stimulus package - namely, that the economy works on unquestioning Faith. The Stimulus will work as long as people believe it will work, the math be damned. We can't make an Abyssal Plan because doing so will cause it to manifest.

They seem to think that the economy is a form of Consensus Reality, and nothing more. While you can point to economic bubbles (such as Tech in the 90s, or Tulip Mania in the 1600s) as an example of this, all bubbles share a common feature - eventually cold, hard, deterministic reality came crashing in to lay the smack down.

Refusing to plan for economic abysses seems on par with the driver of a car not asking his passengers to buckle up, because he doesn't want to jinx himself and get into an accident.

While it is true that people hate to think about the abyssal scenario (lord knows, I never do in my own life) there is one branch of government which does a great job of it - the military. I'd wager they even have contingency plans for an intra-NATO war. I see no reason you couldn't find a bunch of Economists who would take a morbid glee in planning for an apocalypse.

When I worked remodeling and landscaping I used to try to visualize what could go wrong and head potential problems off while we were still planning the work. I finally got tired of the man I was working for calling me a pessimist and complaining about my negativity and quit saying anything.

Stay well away from anyone who uses the word "negativity". Every time I have heard it used it was an attack on someone attempting to show some foresight.

Great post.

Here are some possible reasons for not wanting to explicitly plan for future failure, more at the level of the institute than at the level of the economy:

-The board wants to build consensus and forge a diverse group of people into a coherent whole. Telling some people that they are marginal would damage this focus.

-If a funding agency finds out that you can withstand an X% cut with vital operations intact, your odds of getting your grant cut by X% skyrocket.

-The board and officers are largely tasked with selling your concept to funding agencies. Given that funding agencies are usually risk-averse in practice (despite being risk-seeking in rhetoric), any suggestion that you might fail makes it harder to get funding.

You are making a conscious effort not to fool yourself. This habit of mind is excellent in a scientist or philosopher, but does not work well in contexts involving leadership, salesmanship, and negotiation.

Years ago at the Singularity Institute, the Board was entertaining a proposal to expand somewhat. I wasn't sure our funding was able to support the expansion, so I insisted that - if we started running out of money - we decide in advance who got fired and what got shut down, in what order.
Even over the electronic aether, you could hear the uncomfortable silence.

People are really, really reluctant to plan in advance for the abyss. But what good reason is there not to? How can you be worse off from knowing in advance what you'll do in the worse cases?

I don’t suppose you can. But the process of deciding in advance can cause a lot of trouble. It would be necessary for people to argue in favour of e.g. firing Eliezer Yudkowsky first, rather than anybody else. Then you might have to work with the person who made that argument. Perhaps after arguing that he should be fired first, instead.

Perhaps it would have been easier to decide in advance that everyone should take a pay cut...

Short version: People try to avoid hard choices because they are hard. If the choice will not have to be implemented for a long time, if ever, there is therefore a lot of pressure to defer making the choice. After all, if you defer it long enough, you might never have to make it at all.

I think the answer is, beforehand there are too many contingencies and it seems like a waste in mental effort to consider them all.

Consider the stock example, which is quite simple. I buy a share at $100. I don't think it's so simple that I decide in advance I'm going to sell if it drops below $80. What if, in one day, it drops to $83, and I don't have access to a broker except to sell it at the end of the next day? What if the stock is doing well, and then there's a rumor that quickly sends it down to $80? What if I happen to have good reason to think the rumor is false? What if it's doing very well, but then a rumor sends it down to $85? And this time, I think the rumor is correct. What if I'm away from my broker while bad news develops and it drops to $50, but then the bad news appears increasingly not so bad and the stock is rising when I get in contact with my broker again? Etc.

And all this has to do with when to sell a share of stock, which is seemingly characterized by a single number. There is much more that can develop in connection to the scenario you describe.

Does the Singularity Institute have plans for what to do if an unfriendly AI appears from nowhere? (Not that you should make such plans public.)

For the same reason that when you're buying a stock you think will go up, you decide how far it has to decline before it means you were wrong

Do any investors actually do that? I don't mean to be rude - but why haven't they got better things to do with their time?

But what good reason is there not to? How can you be worse off from knowing in advance what you'll do in the worse cases?

The answer seems trivial: you may have wasted a bunch of time and energy performing calculations relating to what to do in a hypothetical situation that you might never face.

If the calculations can be performed later, then that will often be better - since then more information will be available - and possibly the calculations may not have to be performed at all.

Calculating in advance can be good - if you fear that you may not have time to calculate later - or (obviously) if the calculations affect the choices to be taken now. However, the act of performing calculations has associated time and energy costs - so it is best to use your "calculating" time wisely.

Another reason why not...

Aurini: there is one branch of government which does a great job of [the abyssal scenario] - the military.

Millenium Challenge. (A defence of it is here for sake of balance).

I'd expect lots of Advance Abyssal plans by government, especially war gaming as Eliezer suggested, to end up something like that. As a hypothetical future it falls more prone to signalling and various biases.

A step in the right direction: http://money.cnn.com/2009/02/25/news/companies/banks_test/index.htm?postversion=2009022514

It seems reasonable to me to not come up with fixed solutions to crisis, ahead of time.

I'd propose that it is very difficult to, in general, ahead of time, properly assess the viability of a solution. Planning who to fire ahead of time doesn't make sense for a business, since it is difficult to figure out ahead of time, who will be most likely to be useful or harmful during a crisis. For a world wide crisis, it may be more reasonable to have an arsenal of well thought out, potential solutions, and an ordering for considering them in any situation - assuming it can be done more efficiently than coming up with a solution on the spot.

In the specific case of a political body, this sort of planning faces even more problems. If the government had planned ahead on how to deal with a financial meltdown, the plans would have been heavily influenced by those with political clout, increasing the risk of profiteers getting into the fray, to attempt and abuse any recovery plans. The government has to be competent enough at dealing with planned solutions in general, for the plans to be worth making.

The pain of a crisis is in having to deal with it in the first place. What you are proposing is to deal with a crisis before it occurs, in order to prepare ourselves better for it. This means incurring the pain upfront even though there is a low probability that the crisis will ever occur. Now multiply this by the number of different crises - New Orleans, after all, is different from Bear Sterns - and we might spend all of our spare time preparing for low-likelihood catastrophes.

Perhaps you'll argue that not everyone needs to prepare, only a few people need to do it on everyone's behalf, but I don't think that's true. Preparedness has to be pervasive, or else we're not prepared.

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30