Missing the Trees for the Forest

One person's "overarching narrative" is another person's set of Bayesian priors.

Take, for example, your pollution discussion. An economics textbook will tell you that there is an ideal level of taxation, yes. However, it won't tell you about regulatory capture, mission creep, the Hayekian knowledge problem, etc. There is always a correct set of contextual data to be used to interpret and resolve problems in isolation, yes - but determining what this set is and how we should interpret the probabilities of various events occurring is pretty much always going to invoke an overarching narrative unless you really, really think that this screws everything up.

But as an economist - a Masonomics student, no less - I'm inclined to see a greater harm in "markets fail, so assume a benevolent social planner and imagine what policies she could implement" approach to solving economic questions than the harms of dirtying one's self in the morass of historical context. This is why social science is hard - and why it should be hard. It's not that we should be indifferent to the injection of ideology into these debates, but that it's liable to create greater harm if we try to avoid all questions which cannot yield precise analytical answers which are self-apparent to reasonable minds.

Hopefully I'm not strawmanning your point here - maybe you're primarily trying to explain how ideology mucks things up, but I also get the impression that you think it's reasonably feasible to avoid questions that lend themselves to ideological answers... I definitely would disagree with such an assertion.

I have found a pattern: when people consider an idea in isolation, they tend to make good decisions. When they consider an idea a symbol of a vast overarching narrative, they tend to make very bad decisions.

If I recall correctly, these are generally called Grand Narratives. And it is one of the agendas of post-modern philosophy to try to undermine and eliminate them in favor of more nuanced ways of thinking about things.

Admittedly, most "po-mo" folks aren't as selective as we'd like, since they, often, consider rationality one example of a Grand Narrative.

"What drives someone to hear about a reasonable change in cereal advertising policy and immediately think of a second Holocaust?"

you have to one up everyone else commenting on a subject in order to get attention. attention = potential higher status.

A culture where a popular feminist once said that all sex was rape

Misquote.

This post exhibits a common fallacy in both economics and politics. I'm not sure what, if anything it's traditionally called, but for the moment I propose the "We are all on the same side" fallacy. Let me explain.

"Let's say we are considering a law mandating businesses to lower their pollution levels. So far as I understand economics, the best decision-making strategy is to estimate how much pollution is costing the population, how much cutting pollution would cost business, and if there's a net profit, pass the law. Of course it's more complicated, but this seems like a reasonable start."

Actually, no. Non-owners of a polluting business will rationally prefer that the business is to be made to stop polluting, even if that costs that business more dollars than they are willing to pay for clean air/water/etc. Classical economics suggests that in this case the business owner should pay the populace for the right to pollute, but that's never been practical or seriously considered.

In reverse, owners of polluting businesses prefer to pollute even if the benefit to them of polluting is far less than the cost to everyone who suffers their pollution. This is the classic example of an externality in economics.

Politics, especially democratic politics, is the process by which people with differing and opposed interests compete to try to put their interests first. There is no one rational best choice for everyone. Different people have different wants and desires that cannot always be mutually satisfied.

None of this negates your discussion of how people use broader narratives to understand (or misunderstand) specific stories. Sometimes, as in the Duke Lacrosse case, it's simply a matter of fact. And other times, though rarer than most people think, there is one obviously better solution for everyone; and we should all just rationally choose it.

But most of the time you need to distinguish the rhetoric and propaganda from the genuine interests of different parties. Take your example of "there would be seemingly good idea to regulate something that clearly needed regulating". Maybe for the good of society something, call it a "boonlicket" to avoid triggering patterns on real products, needs regulating because fly-by-night producers are highly incented to produce dangerous boonlickets instead of spending ten cents more per unit. Or maybe retailers are incented to sell expired or damaged boonlickets instead of new ones. And let's suppose that everyone who uses a bad boonlicket loses their little toe; and that this is a real problem and it's really happening in the marketplace. Lots of folks are now walking around without little toes. (I'm being deliberately extreme here to make it really obvious that regulation is called for.)

Nonetheless retailers and manufacturers of good boonlickets may still not accept regulation because following the regulations will cost them a lot of money, even if it saves millions of little toes. (Then again, in some cases good manufacturers may actually request regulation if it puts the bad competitors out of business, and restores consumer confidence in boonlickets as a class; but for the moment let's assume the costs of regulation are really, really high; or maybe the bad boonlickets only sometimes cut off a little toe.)

Either way, the manufacturers and retailers of the bad boonlickets are going to strongly oppose the regulations because they'll lose everything they make on the bad boonlickets. Now of course they can't go around saying they want to sell boonlickets that cut off peoples' toes to save $0.10 a unit, so instead they try to obscure and obfuscate the problem. They talk about "Small government! Freedom! Capitalism!" though they don't actually care about any of that stuff. They just want to continue selling the toe-cutting boonlickets.

To understand what's really going on in these debates and disputes, you need to look beyond what people say. Ask yourself who benefits? Who pays for that benefit? How much do payers pay? How much do beneficiaries receive? A classical economist would identify this as a transfer of wealth from one group to another, though I think wealth isn't a full description of what's happening here. But it is a transfer of utilons from one party to another; and politics is one way of distributing a society's gross utilons. It is by no means obvious or accepted that we should simply accept any policy that results in a net increase of utilons across the population, or reject any any policy that results in a net decrease of utilons across the population. The real question for politics is "Who gets what?", not "How much is there?" Politics is about distribution, not maximization.

I came to the wrong conclusion in the Duke Lacrosse case, even though I didn't think our society is all that racist. I messed up because I just assumed the authorities wouldn't do what they did unless they had good reason. I have since changed my assumptions.

I'm afraid I have to take issue with your Cheerios story in the linked comment. You say of the 4% cholesterol lowering claim, "This is false. It is based on a 'study' sponsored by General Mills where subjects took more than half their daily calories from Cheerios (apparently they ate nothing but Cheerios for two of their three daily meals)." You link to http://www.askdeb.com/blog/health/will-cheerios-really-help-lower-your-cholesterol/ but that says nothing about how much Cheerios subjects ate.

I found this article that describes the 1998 Cheerios research that is the foundation for the claim: http://findarticles.com/p/articles/mi_m0813/is_8_32/ai_n15691320/ . It says that participants ate 3 cups of Cheerios per day, while control subjects ate 3 cups of corn flakes. 1 cup of Cheerios is about 100 calories, so 3 cups would be 300, far less than "more than half their daily calories" for any reasonable adult. Further, this article goes on to report that LDL (bad cholesterol) in the Cheerios group fell from 160 to 153, which looks to me like 4%.

Furthermore, my understanding is that the FDA's complaint is not with the accuracy of Cheerios' claim; it is that it is making such a claim at all, even if truthful. The FDA has a lot of rules about what kinds of health benefits products are allowed to make. It is not enough that a claim appears to be correct; the question is the depth and strength of the evidence behind it. For specific health claims, the FDA basically requires full blown clinical trials, as with drugs. According to the LA Times, http://latimesblogs.latimes.com/shopping_blog/2009/05/fda-warns-general-mills-over-cheerios-cholesterol-claims.html , "The FDA allows some health benefits of foods to be advertised but within strict limits. For instance, a company can say that a diet low in saturated fat and high in fiber-rich foods such as fruit, vegetables and whole grains may reduce the risk of heart disease."

There are reasonable questions to be raised about what policy we want to have for regulating health claims. But demonizing Cheerios and General Mills does not facilitate rational discussion of the issues. Saying that their claim is false, and exaggerating the amount of Cheerios which was eaten in their study, only serves to put Cheerios in an unjustifiably bad light.

And BTW I typically eat 600-1000 calories for breakfast, often cold cereal. Sometimes I eat Cheerios but usually I mix two or three different cereals. 300 calories of cold cereal is not difficult for me. The hard part is holding myself back to only eat that much.

This is... impressively level-headed, like orthonormal's post was. The net result might be to shame me into shutting up on those hot-button topics, which would be a good thing. We really had better stop and move on. Although the last couple days have been an obscene karma mint for many commenters including me.

I agree with the logic of this analysis, but I have a problem with one of the implicit premises: that "we" should care about political issues at all, and that "we" make governmental decisions. I think this is wrong, and its wrongness explains the seemingly puzzling phenomenon of jumping from tree to forest.

There was no need for anyone beyond the jury to have an opinion on the Duke lacrosse case. We weren't making any decisions there. I certainly wasn't, anyway. So of course when people do express an interest, it is for entertainment and showing off only. They may think it is for other reasons, but it is essentially a form of social interaction, part of the status game that we are all playing. And this game is played better with big issues than with small ones.

Likewise with poor Cheerios. (It's funny - I wrote a semi joking rant last night defending Cheerios, and as a result I now find myself quite favorably disposed to the little yellow box; an effect we have often discussed and warned against.) I don't need to have an opinion on what the FDA should be doing. They aren't asking me. Nobody's asking me. At best I can vote for a President who can appoint an FDA commissioner and perhaps set policy, but my influence on this process is infinitesimal.

So once again, if I do take an interest it will be as part of a social game, not because it's something I can do anything about.

This effect is the fundamental reason why ideology rules in politics. It's because our beliefs don't matter, so we adopt them just for fun and for a competitive edge. We don't seem to recognize this, perhaps because believing ideologies are important helps us win the game. But it explains why people are quick to see little things in the context of big issues.

The more time I spend thinking about it, the more I come to realize that Narrative Is the Enemy, at least where attempts to see and reason clearly are concerned. One heuristic has proven surprisingly useful time and time again, in efforts of rationality as well as creativity: don't try to deliberately tell a story.

I gave an example: the government's request that Kelloggs stop making false claims about Cheerios

Cheerios is made by General Mills, not Kellogs.

I like the concept of narratives (make it "affective narratives"). Every problem looks like a nail. This bias doesn't necessarily determine what gets decided, as the factions can fight over the outcome, but it results in discarding all the relevant considerations, leaving only the sufficiently polarized ones. The perception of facts gets distorted without an agreed-upon affective bias for a specific cause. A cult-anticult field that doesn't necessarily consist of a pair of organized cults. When the cults are organized around the poles, it gets worse, but the first stage is like spontaneous fluctuations from sanity to both sides along the given direction.

This possibly creates an approach to the question posed in Do Fandoms Need Awfulness?. If there is a controversy, it creates factions, which start to argue in an adversarial mode, creating an affective narrative that draws attention to a few simple properties of the object of fandom, raising them to absolute, which appeals to the people who value these properties.

I get the feeling sometime that people tend to be "blunt objects", that there is a tendency to see one issue that does have some importance, or a few, and then go slamming against that issue and things that resemble it. Then your slamming becomes an issue and other people start hammering against your whole position and related views. If this system works at all, it hopefully works by this pounding back and forth settling somewhere close to where we think things ought to be, based on the relative 'objective' merits that give a bit more fuel to one side or the other. Being built to be successful packhunters and foragers on the savannah and not the perfect disputants, I think people are applying what limited resources they have or think they have in attempting to settle fine points with large unwieldy views, the sort of broad conceptions that did the job in the savannah. Hopefully by policing ourselves and/or increasing our own ability to handle lots of nuanced information at once we can apply more precision to things. Then instead of flinging the issue around like a tetherball, we could carefully place it in a measured and thoughtful position to best satisfy our collective desires. Anyway just some thoughts and an analogy I find viscerally enjoyable.

I agree.

In fact, I think the "vast overarching narrative" is often something very basic. In many cases, I think a given policy is seen simply as "punishing" a group, and is resisted by people who think that group deserves more respect.

For instance, why are so many people against increasing taxes on the wealthy to cut deficits? Are they really worried that these people will lose incentive to work hard, and the economy will suffer? Or that markets for luxury goods will crumble? I suspect what many really feel is that successful people deserve praise and admiration, while increasing their taxes puts them down -- the tax increase is seen to say that their wages were not fairly won. And perhaps many on the other side feel so strongly about it because they think wealthy people are overpaid and otherwise get too much admiration, and thus should be taken down a notch.

The "vast overarching narrative" is simply "wealthy people get too little / too much respect."

Same with govt regulation. Do you think society gives large corporations too much or too little respect? (That is, do you think they have too much power and wealth, or do you think people tend not to appreciate how important they are toward empowering and creating wealth in our society?) Perhaps differences in the answer to that simple question explain why many take a set of certain positions on a wide range of questions of regulation, the practical implications of which are actually very diverse.

Thanks for sharing your thoughts on this. Even though I don't think there was anything really new in this post, I like it anyway because it finds additional evidence and helps build the body of evidence to support the idea that politics is the mind killer.

As far as actually talking about bias in general, I don't think this post adds anything that hasn't already been said, for instance in "Politics is the mind-killer". In terms of the current discussion, I thought the idea was to drop it and move on.