I saw someone the other day put forward an argument against re-distributive taxation based on a study and a second person say ‘but that study is from the Cato Institute’. A bunch of people then jumped in to object that this was an ad hominem argument, and the study should be evaluated on our own merits, even if it was from the Cato institute. This completely ordinary interaction on the internet got me thinking about the social epistemology of such arguments and why I think imperfectly rational and informed agents should not be so quick to dismiss ad hominem arguments in these situations, at least not entirely. Rather such agents must remain resolutely conscious of the structure of social incentives that shapes the discourse they are immersed in.
Consider the following thought experiment:
You’re aboard a spaceship and you crash land on a mysterious planet. To your surprise you discover an advanced technological civilisation there.
Upon discovering the locals and resolving communication difficulties, you are asked to resolve for them an ancient dispute. You, they feel, are in an excellent position to be impartial. At first you try to decline the offer, but they are very insistent- the issue simply must be resolved one way or another.
They explain what the issue is. The juice of a certain fruit needs to be distributed between them. Ancient custom has acknowledged that for various reasons, some of them are entitled to vastly more juice than others. The issue to be decided is whether a more even distribution of juice should be adopted, or whether the customary distribution should continue.
One of the powers of this juice is that it can fuel special thinking machines, and working together these thinking machines can make powerful arguments- but always working under the direction of, and for the purposes of those who fuelled them with the juice. Naturally most of the argument producing robots are operated by those with a lot of the juice, and thus most of the arguments they have produced on the topic are in favour of the existing uneven distribution of juice.
These robots however are not the only intellectuals in this society. The juice has the property of allowing its imbiber to go with less sleep than they would otherwise require. These creatures normally sleep for three quarters of the day, but those who possess the most juice can go for days without sleeping. This means that they can spend more time writing and to discussing ideas. Very often the juice-rich write and think about the topic of the optimal distribution of the juice. They most often come to come to the conclusion that the existing distribution, or a slight tweak on it, is optimal- although a minority of them do support redistributing the juice.
Considering the situation, you find yourself with a problem. You suspect that the existing literature on the optimal distribution of the juice is probably quite biased, that results which seem to support its unequal distribution are found more often and receive undue prominence due to the structural advantages which support the status quo. However you are not an expert on the juice literature, it is very confusing and contains many mathematical symbols. There are a lot of stats, and every time someone puts forward a stat, someone else says that this is actually a misunderstanding.
What should you do?
We can debate exactly how much weight you should put put on the lopsided origins of this society’s thinking about the distribution of the juice. We would all agree though, that in the absence of the cognitive resources necessary to sort through the theories and arguments in ideal detail, if one has to make an assessment on the basis of one’s limited information and intellectual powers, surely some weight should be placed on the reality that the side supporting the status quo commands greater resources, and this is bound to make its arguments appear stronger than they otherwise would.
Lifting the thin veil, if you haven’t already guessed, the juice is money. The moral of the thought experiment is that in assessing the arguments for and against redistribution you should absolutely be acutely aware that there are not a lot of think-tanks funded by poor people. Even if you happened to be a trained economist, your mental capacities are still limited, and you do not have infinite time to check all things. You should, at a bare minimum, prioritise listening to and finding material by the side of the argument that is likely to be less well funded since you are less likely to encounter such material organically than you would be if both sides were funded equally.
DePonySum 
I don’t think the Cato Institute has that much money compared to all the organizations producing papers. But the papers arguing from a different perspective from Cato may still represent a more moneyed, upper-class POV. In another of your posts you described existing political coalitions as the establishment vs the disenfranchised, but it should be obvious that the leaders of both coalitions are hardly disenfranchised. And as Jesse Singal has been pointing out about his own coalition, that leadership often seems to be completely ignorant about the actual preferences of the people they claim to represent. Bryan Caplan wrote about some research on how the preferences of the rich dominate in U.S politics. The U.S president most responsive to the preferences of the poor was George W. Bush while the least responsive was LBJ (Caplan would take that as a confirmation that democratic majoritarianism is bad and “slack” for political elites is good).
Outside the realm of policy papers, many empirical papers have failed to replicate. And when a replication project permitted people to bet in advance on which papers would replicate, those predictions were quite accurate. So instead of saying the source of a paper is irrelevant or determinant, there should instead be more bets on verifiable things like replication. And in politics we can have things like Robin Hanson’s futarchy.
LikeLike