PolySciFi Blog

Saturday, April 02, 2005


Improving Society Is a Time-Varying Multi-Dimensional Multi-Objective Nonlinear Optimization Problem

Sometimes the articles you read converge to a single point. I've excerpted several articles and see if you can guess where I'm going with this (ok, the post title might be a give away).

Stuart Buck on competing principles:
"People often accuse their opponents of being hypocrites when, in fact, they may simply have been balancing competing principles. We all do this constantly. And the mere fact that someone reaches a different balance than you, or that they decline to treat one principle alone as being absolute, does not prove that they are being hypocritical." [emphasis in original]
Professor Bainbridge on conservative/libertarian federalism brouhaha over Terri Schiavo:
"My point was that we conservatives care about federalism, but we also care about other values that contribute to the good society. Unless federalism is to be a fetishistic end rather than a means to achieving the good society, it therefore must sometimes give way to other values. Our disagreement thus seems to be a question of degrees rather of kind....

My point all along has only been that the principle of federalism is a (very important) means to the end of the good society and there are cases in which it appropriately gives way to other values. After all, aren't there exceptions to just about every rule?"
Ok here's a little takeaway from these two articles. We have several different priorites, or objectives if you will, and balancing those priorities can be maddeningly difficult. I can solve a linear multiobjective problem fairly quickly (give me an hour of your time and you can too), but society isn't linear and nonlinear multiobjective optimization problems are pretty tough.

So before jumping to the conclusion of hypocrisy, it's valuable to note that people are making their decisions on more than one dimension. So when circumstances change a little, radically different results might be "optimal."

But here's a further complication - different people in society have different objectives, and even those we have in common (which are many), we weight differentlly.

Jonah Goldberg in OpinionDuel (debating Chait's claim that liberals are empiricists and conservatives are ideologues):
" Your whole argument boils down to the assertion that outcomes you prefer are outcomes that "work." The problem is that this is an ideologically stacked deck. If a policy maximizes income rather than freedom, or vice-versa, which one do we want to say has "worked"? You want to claim that only the policy that adheres to "liberal" ends worked. That's fine. But that's an ideological judgment, not an empirical one. For example, you write:
Conservatives say over and over that we have the "best health care system in the world." But that's only true by ideological criteria--which system is the best at minimizing government intrusion.
You made this point in your original article and again here so I assume you really mean it. But do you understand how ideologically loaded this is? The only way you can say our system is the best is by ideological criteria? Jonathan, what are you talking about? America leads the world in terms of innovation. It leads the world in terms of our investments in R&D. It leads the world as a magnet for young doctors from all around the world to train. It leads the world in terms of high-end procedures performed. There's a reason people drive down from Canada for medical procedures in the U.S. There's a reason aging potentates and princes fly to America for treatment.

What you are really saying is that America's healthcare system doesn't meet your ideological criteria for what would make it the best in the world. That's fine. From your articles, I gather you're more utilitarian, believing that the best healthcare system would do the most good for the most people at the cheapest price. That's perfectly reasonable and there's nothing illegitimate whatsoever about arguing from that perspective. But you pull this nonsense — so amazingly typical of smart liberals — that people who disagree with your ideological criteria are the only ones being "ideological."

Indeed, what is wrong with "ideological criteria"? You use the phrase like it means "the enchanted land of leprechauns and harp-playing fairies" — as if ideological criteria have no empirical basis in reality, never mind moral legitimacy. Ideology is merely a checklist of priorities and principles we bring to the real world. You have an ideology and I have an ideology..." [emphasis not in the original, and I've elided some verbal taunts not related to my point]
In an email sent to Jonah after this article, a self-identified liberal writes:
"And as you point out, there's nothing wrong with using ideology as a filter. Conservatives should care about whether tax increases impinge on individual freedom, even if they do achieve the policy ends that liberals seek from them (and even if they achieve a policy end that conservatives support as well as liberal). And liberals should care about income inequality, even if a conservative economic policy puts more money in the pockets of the poor as well as the rich."
So in the US, by my estimation, we've got some 270 million different ideologies, each one weighting the objectives for society a little differently.

Then for two more complications, 1) as Thason is fond of pointing out, there are many different ways to achive the same objectives and 2) our preferences change in time (how important is that pony you wanted as a 5 year old?).

So the larger objective of improving society (an objective I think everyone shares to some extent) is actually a time-varying multi-dimensional multi-objective nonlinear optimization problem. And if that sounds hard to solve, that's because it is.

This is also why no politician will be your "perfect politician," and certainly no party will be your perfect political party. No one else weights the objectives quite the same way that you do. Sure there's some broad coalitions that come together, but these are ultimately transitory as no two people will agree on everything. (If you find someone who agrees completely with your objectives and your free variables, be afraid, they may be your Cylon replacement.)

I think that recognizing this fact is valuable for keeping political discussions civil. Your opponent probably isn't a moron and probably isn't the antichrist. He/she just has a different value system.

This fact also influences why a) I find it valuable to have different political views on polyscifi (presents other dimensions and other weightings of objectives), b)why I'm a strong federalist (to a limit) so you can move if you don't like the way society plays out around you, and c) why I'm very much against a single world government (I was very much for it as a youngster, but with me on top, naturally).

To test my belief at least about the personal multi-objective part of the problem, I've listed 10 different societal objectives which I just pulled off the top of my head. If you're willing to play along, assign a weight from 0-10 to each of these objectives a leave your weighting in the comments.

Loosely, we'll say a '10' means I'll give my left nut (insert appropriate female equivalent), and we'll say a '0' means, "Jody, why on earth did you think this was a good thing?" Use any number you like in the range 0-10, but if you want to use an irrational number, I expect that you'll give the full decimal expansion. If you think there's some ambiguity as to what one of the objectives means, assign it whatever meaning makes sense to you.

Objectives to Weigh
Beauty (or art)
Intellectual Stimulation


This page is powered by Blogger. Isn't yours?