Nuclear Mangos

This blog is intended to provide reliable technical analysis of nuclear issues with non-state actors and nuclear beginner states. Some technical issues have important policy implications that citizens in a democracy should be able to make informed decisions about. The motivation for the blog has been the incredible amount of lies & hyperbole on the Iran situation of early 2006. The blog title is to remind you constantly of the quality of minds in charge of our nuclear security today.

Name:
Location: MA

Until recently I was a physics professor at Harvard, where I taught the nuclear and particle physics course, among others. I've recently left that position to work as an R&D physicist in security applications. I have never done classified weapons work.

Thursday, August 14, 2008

The True and Lawful End of Aspiring

This administration has a long record of using a vapid and naive cover story of democratic evangelism to justify misadventures. For the most part, these misadventures seem to have been fueled by a mix of hubris and corruption. I occasionally suspect that the policies are sold to a President easily gulled by this sort of naive moralism, and that sales pitch then doubles easily for an MSM too uninvested in the truth to care and a populace too busy trying to cover the mortgage to spend time caring.

So for the purposes of this post I wanted to lay out what I think ought to be a non-vapid role of moral considerations in various sorts of foreign policies. To do so, I'm going to make a number of assumptions: that a "right" side exists, that it can be unambiguously identified, that the highest foreign policy goal is to serve the forces of Sweetness And Light (henceforth SAL), and that the amount of SAL at stake in any one conflict can be accurately quantified in units commensurate with resources. These assumptions are not only debatable but probably mostly counterfactual. But taking them we can make some progress on how to think about what one ought to do. Countering assumptions can then be folded into that framework afterwards.

The first observation is that the moral obligation, in such a framework, is not to fix all the world's wrongs and struggle for SAL in every instance, but instead, to maximize the impact of your resources upon the extent of SAL.

The second observation is that it's OK--even necesssary--to take the long view. If you can expand the reach of SAL by 1.5% a year, then after half a century, you will have doubled the world's SAL. That would be a real contribution to humanity, and that it took half a century would not be a substantive criticism. A President might well make his policy goal to end his term with the world 6% better off than when he started it, and be satisfied he was contributing to real progress. Or, as one president put it, heed the

call to bear the burden of a long twilight struggle, year in and year out, "rejoicing in hope, patient in tribulation"—a struggle against the common enemies of man: tyranny, poverty, disease, and war itself.


Put something--even a small thing--in the win column, every year, year after year, and you'll look back after fifty years and be amazed at what's happened. You don't have to win it all in a blaze of glory. In fact, you just can't win that way.

Some years, of course, you won't make the 1.5%. You might have to average 1.7 for a while to make that up. You might make a big mistake, and have to average 2.5 for a decade or so to make up for it.

Taking that long view, one wants to approach every SAL opportunity asking: will I increase SAL by an amount commensurate with my long-term resources, or will I deplete those resources to an extent that it will limit my ability to increase SAL in the future, by more than I could possibly get out of this opportunity?

You do not do the world a favor by fecklessly chasing every single tiny thing and expending vast resources for tiny gains. That comes at the expense of much greater gains that become sacrificed.

So, one can view the twilight struggle as a series of opportunities: at each one, resources are staked, with some probability SAL increases; with some probability, there is a loss and SAL decreases by the amount of your lost resources. How do you maximize the growth of SAL, given a finite resource base, a set of given likelihoods and outcomes, and a desire to make sure that you never face ruin?

This is, as it turns out, just a math problem. The result is the Kelly formula. When the expected value of the stake is positive, you stake a fraction of your resources according to:

f*=(bp-q)/b


where b is relative amount of SAL to be increased in good versus decreased in bad outcomes, p is the likelihood of a good SAL outcome, and q is the likelihood of a bad outcome (i.e. 1-p).

However, when the expected payout is negative (i.e. bp-q is negative), you do nothing. This can occur--in fact, often does occur--when there is, in fact, a possible positive outcome b for SAL. Which is to say--just because there is a chance that you might improve SAL, is no reason to go around staking resources--let alone large resources--on an event.

Let me put it another way. If an Administration is arguing that one side is all SAL and the other side the Prince of Evil, that our foreign policy is to support people who are SAL, and that by intervening we might be able to improve the SAL of the world--even if you uncritically accept all of those assertions, that is still not a case for any particular action one way or another.

If all of those assertions are themselves questionable, well, the decision is even further from resolution.

Two footnotes:

1. The Kelly formula is widely misused and misdescribed in gambling circles. My own rule is: don't believe anything you read about it, unless it's in TeX.
2. "Power to do good is the true and lawful end of aspiring." Francis Bacon.

4 Comments:

Anonymous Cheryl Rofer said...

I'm not sure that anyone but a physicist would be able to follow this entire post through.

But I think that what you're saying is that working at the margins over time is more effective and robust to potential failure than the big spectacular push up the middle.

I think I've mixed economic and martial metaphors there.

12:23 PM  
Blogger Andrew Foland said...

Yes, I have Tanta disease. I am sure this could be boiled down into many fewer paragraphs.

I just keep hearing how "the US must do X because that is the side of good and right" without any discussion of how to manage finite resources to maximize the total amount of goodness done over time. And that latter is, to me, the true moral imperative.

I guess what I'm really after is that I'd rather have a systematic way of thinking about problems, like "how to do good", even if that framework is beset by uncertainties and doubts, than no system at all.

No system at all makes you feckless, inconsistent, and ultimately less effective.

8:52 AM  
Anonymous Cheryl Rofer said...

Agreed. And there is, for example, systems analysis to do this more or less mathematically.

I've been looking at what seems to pass for energy policy lately. Can't find anything that measures up to your preferences. I guess I'm not surprised, just disappointed.

9:55 AM  
Blogger Owner Earnings said...

For entertainment purposes only, here is the fraction of its asset value that Lehman lost (dividing the 91 cents on the dollar lost on the $130BB outstanding bonds, into assets of around $800 BB):

Very good point.

And that's just the banks.

Word verification? U serious?

11:39 PM  

Post a Comment

<< Home