There’s been a good deal of debate on this blog about whether it’s appropriate to look at the motives of people who are making particular arguments (for example Lomborg and the environment), or whether you should take them at face value and respond to the arguments directly. Much the same debate has been taking place among US econobloggers, including Brad de Long and Arnold Kling. You can get started with Brad here.
My view is that motives matter. It’s very difficult to conduct a reasoned discussion with someone if you know they will lie, or distort the truth, whenever they get away with it. Hence, it’s important to distinguish between honest disagreement and propaganda and necessary to respond differently to one than to the other.
Update This post from Tim Lambert illustrates the point perfectly. It concerns an article by well-known pro-gun academic John Lott purporting to prove the correctness of some claims made by rightwing windbag Rush Limbaugh about media coverage of black quarterbacks in the NFL. As Lambert says
Even if all the data is correct and his regressions have been correctly calculated his analysis is not in the slightest bit persuasive. The reason is that his behaviour in the coding errors case suggests that he just keeps trying different models and just cherry picks the one that gives the result he wants. Would Lott want to get a result that supports Limbaugh? Well, check out this Mary Rosh posting:
You have got to download this paper. Lott has done an amazing piece here. Fits in perfectly with Rush Limbaughâs program today.
Tim’s site has more on the Lott/Rosh saga and the bizarre parallels with the case of antigun researcher Michael Bellisles. Ted Barlow at Crooked Timber has more on this (warning: Turn irony detectors on full before reading this post). Finally, for more on the problem of cherry-picking, aka “data mining”, read here, here and here. To spell it out, statistical results from someone you can trust to play by the rules are worth discussing, those from someone known to engage in data-mining/cherry-picking are not.
John Quiggin 
I was going to disagree… but then I thought it through. I think it’s all about trust.
When I’m trying to make up my mind about an issue that’s beyond my area of expertise (ie most issues) I have to rely on the judgment of people with expertise. I need to find people who are:
1. able to find and interpret the information available; and
2. can be trusted not to exploit my (and other people’s) ignorance about the subject matter to further their own interests (or moral agenda).
In an abstract sense it’s the same problem I have when I visit the doctor, dentist or auto mechanic.
When ‘experts’ betray our trust they put evidence based policy making as a whole at risk. The more often people are denied information, given evidence for only one or two sides of an issue, or have the evidence twisted to fit somebody’s agenda the less likely they are to trust expert judgement in future and the more likely they are to make their minds up according to partisan allegiance.
Cynicism is corrosive.
Good post Don, you summed up a few of my thoughts on the issue, especially re. information assymetry and the depleted reserves of trust.
I don’t think that motives matter, it’s the quality of arguments that matters. I like reading blogs written by ppl I will generally disagree with but can make a decent argument, without grinding ideological axes.
Ideologues tend to be quite bad at making coherent arguments, which is good ’cause it makes them easy to spot and then ignore. If you cannot ignore them, just resort to pointing out any fallacious arguments they make, that should usually scares them off. Or you could grind the argument down to a fundamental point of disagreement, and leave it at that.
John,
It’s a difficult question.
Generally I believe that the better course of action is to focus on arguments rather than motive simply because anyone’s objectivity can be called into question without too much effort.
For example, it could be validly argued that any information supplied by the US Government on the situation in Iraq should be ignored or heavily discounted given that it is not an objective observer. It is just as easy to dismiss any statement or argument from Greenpeace on the basis that it has a political agenda to further.
Similarly, given Henninghamâs work should we discount most media commentary on the basis that most of itâs practitioners are “slanted” to one side of the spectrum?
The politicisation of the media, some arms of science, community groups, churches etc. leaves us ordinary folk distrustful of what were previously considered reliable and independent purveyors of information.
Knowing the motives of the protagonists in a debate is very useful. For the expert, it saves time because it gives clues on where to look for dubious assumptions, fudged figures or indeed evidence of data mining. For the non-expert, who is easily persuaded by a polished-sounding argument, knowledge of motives is a good antidote against the seductive power of statsitics and jargon.
No-one is saying that it’s enough just to attack the motives of one’s opponents. At the end of the day an argument will only win on its merits.
John was quite right to warn his readers of Lomborg’s affiliations and track record. Unfortunately he confused the issue with his non-sequitur about the Danish Government’s cuts to foreign aid. But the rest of it was justified: if Lomborg tries to boost his credentials in the eyes of people concerned about third-world poverty, by pretending he is driven by a similar preoccupation, John is doing me a service by exposing the pretence.
I don’t think it’s necessary to disregard information supplied by a source known to be biased. It’s often possible to make the most of unreliable information when you know in what way it’s likely to be unreliable. In many cases (intel about Iraq being a good example) there isn’t much impartial information out there. In a more post-structural vein, I’d say that pretty much everything is infected by motives and biases, so the best we can do is to be up front about where we’re coming from, and try to discern the same information from those we listen to.
Information only becomes completely worthless when someone is straight-up fabricating it. Cherry-picking leads to somewhat useful results when you know the picker’s motives.
This is an interesting argument. I think we can safely say that (1) studies of motives re public and intellectual figures etc are legitimate, as are (2) studies of the positions, arguments and works of the same. The difficulty comes with (3) the relationship between the two.
Obviously, (2) may remain valid or at least worthy, regardless of (1), as can the opposite. At the reductio ad absurdum for (3), think of Ern Malley’s poetry, some of which live on as valid regardless of them being produced as a function of Australia’s most famous literary fraud. Think also of the zillions of words devoted to destroying Marxism via character assassinations of Karl himself. On the other hand, only a fool would take the word of a known liar on a matter of fact, particularly on a matter in which said liar had an interest. Still, even in this case, the possibility that the liar is being truthful can’t be entirely ruled out in a vacuum, only made extremely improbable.
In off-the-cuff pro tem sum, studying motives is relevant in its own right, in the context of understanding political and intellectual activity, and in assisting in establishing truth probabilities … but, at the end of the proverbial day, if the target is the position, argument or work itself, that’s where the case finally has to be made.
James Farrell: “At the end of the day an argument will only win on its merits.” You’re joking, right?
And isn’t is more than a little hypocritical of Lomborg-bashers to argue he’s using data selectively, pretending he has nobler motives than he really has etc. when they have done or are doing exactly the same thing?
Lomborg deserves kudos for showing some environmental scientists have the same relationship to truth as executives for tobacco companies. The latter mislead and deceive for profit, the former for green political beliefs.
John, your “here here and here” links don’t work; I think they link to a Movable Type page that you can see and we can’t.
Fixed, thanks (JQ)
Well… I’ve thought it through again and now I’m going to disagree. I don’t think motives matter.
John writes: “My view is that motives matter. It’s very difficult to conduct a reasoned discussion with someone if you know they will lie, or distort the truth, whenever they get away with it.”
I agree that it’s difficult to conduct a reasoned discussion with someone who won’t play by the rules. But this has nothing to do with motives – it has to do with behavior.
There are intellectual positions I don’t agree with that I can defend without having to lie, distort the truth, or be overly selective with the data. And there’s nothing ethically dubuious about doing this – any university lecturer who can’t present her or his students with alternative theoretical perspectives in an unbiased way shouldn’t be allowed near undergraduates.
There are all kinds of motives for engaging in argument over public issues. Some people get paid for taking a particular position (eg lawyers, some think tank experts), others are morally motivated (eg some environmentalists), while others like to display their intellectual prowess (eg some academics). But as long as they can be trusted to “play by the rules” their motives don’t matter.
It seems to me that the important thing is not the character of individuals but the design of institutions. There needs to be some way that norm compliance can be monitored and those who play fair are rewarded. Academics use peer review, journalists police each other and it’s taken for granted at many news outlets that if you fabricate quotes and ‘witness’ events that never took place (eg Stephen Glass) that your job is at risk and your reputation trashed. Even think tank experts need to be careful – look what happened to Michael Warby when he used one of those internet rumors in a newspaper column.
As a media consumer I can’t check everything I read. As a research consumer I can’t go around asking for the original data and running my own stats. What I want is a set of institutions that helps ensure that most of what I read is trustworthy.
I want institutions that don’t let people get away with lying, distorting the truth, and cherry picking the data. I want the rewards for playing by the rules to exceed those from breaking them. If I have to go around reading people’s minds to discover their motives I’m stuffed.
I concede that “motives” is not the right word for what I’m talking about. “Trustworthiness” is much closer to my central concern.
Motive remains, nonetheless an interesting and valid field of inquiry … outside of economics, that is … if I’m allowed to say so on John’s blog. The hedonist calculus, in its infinite variations, is useful, and economics is fascinating, and working out the relative extent to which the discipline is useful is one of the great modern arguments. But I always feel nine-tenths of the most interesting questions has been lost with the simplifying assumptions as to motive. Can this assumption be dispensed with in some way? Why not call it another surrogate identity, just for the sake of realism, and surely then everything else in economics can just continue as is? Still brilliant, but perhaps with less totalising pretensions … that cause so much unnecessary angst. Just an idle thought, passing by.
John,
Some of us will be on picket lines tomorrow. Opponents of the strike will question our motives, so we need especially good arguments if we are to convince them that our resistance to the higher eductation funding proposals is not purely self-interested.
You undertook to put your paper from last week’s conference up here. Any chance of seeing it, or a condensed version of the argument, today?
It’s not a simple matter of bias leading to selective editing. I’ve noticed that left wing stuff used to be dominated by a sort of unconscious bias that didn’t even let things in in the first place, while right wing stuff used to have a bias in favour of objectivity so awkward stuff got through without actually being properly assessed. These days the fashion is shifting, so that former Marxist stuff has evolved into things like “Spiked” that are being fairer to the facts, while right wing things like Quadrant are insensibly becoming two dimensional and cartoonish.
However this change is still not fully developed. In the September 2003 IPA Review there is an article by Gary Sturgess, “To Join Interest with Duty”, pointing out the strengths of corporate private partnerships in carrying out government functions. He’s following the old pattern: he advances a plausible argument with a lot of merit (though I think he overstates it); and, he does bring in some supporting facts – quite a few of them.
The catch is that he hasn’t actually checked which side of the argument his cases support. That is, he really does have a respect for the facts, and he brings them out so we can see for ourselves. Only, I happen to know that some of them do not support his position after all! So, he really is being intellectually honest, except that he hasn’t finished the job. The meta-arguments for dismissing his argument are wrong, since we actually can check him out.
I won’t list all his examples, just those I happen to know about and which point the other way to what he supposes:-
– He rates the British East India Company as a success, since it “from 1834 until 1858, had no business other than that of public administration.” Actually, its failures from that perspective led it to be greatly constrained in the beginning of all that, with the for-profit mechanisms being replaced with salaried public servants; and, its final failure was the way it led to pressures feeding the Indian Mutiny, and it was taken out of the government business after that, precisely because it was seen to have failed. The things he considers its successes – the whole civil service thing – were in fact transitionally imposed on it while it was being phased out.
– Leicester Square is assessed as a successful company town. In fact London landholdings were not usually corporate – it almost certainly was not a company town, though I have no direct confirmation of this.
– Bournville was not a successful company town resting on commercial motives, but rather an expression of the Quaker ethos of the Cadbury family, expressed through its family business – not a company town at all.
– Pullman in the USA was no success at all; in fact the course of the Great Pullman Strike showed just precisely how not to do it. The company had rented out homes to its employees, which aggravated things when it had to stop paying them enough; they got evicted as well as losing out when Pullman had to cut back. The original paternalism backfired vastly. (In fact, this is one of the incidents that showed me the externality mechanism of downsizing that spreads costs and encourages unemployment; Pullman had internalised those costs but tried to cope as though it hadn’t, which shows just how impractical it is to run a business on true costs when everybody else is passing the buck – they were demonstrating bad management.)
– Garden cities like Welwyn are usually considered failures, since they didn’t deliver what was promised and weren’t even implemented in the form originally intended, with a ring of garden cities around a centre. The garden city idea didn’t hit its target.