The problem is that Romer's writing style is not exactly ideal for what he's doing. He's writing blog posts like academic papers - very careful to specify definitions, cover all bases, and speak in a dry, neutral tone. That makes sense when you're writing a paper, because you're writing to a highly selective audience - i.e., the tiny handful of people in the world who are specialized enough to care about your research topic, and experienced enough to know all the background. It also makes sense because of the way papers get published - it's a long process, and you can't easily go back and change things, write addenda, write quick follow-ups, respond to arguments point-by-point, etc. Blogging is a whole different animal. Romer could get his message across more effectively, I think, if he would just take greater advantage of the format.
For example, let me attempt to translate a Paul Romer post into blog-ese. The post is called "Freshwater Feedback Part 1: “Everybody does it”". That's an OK title, though I would have gone with "justification" instead of "feedback" ("excuse" being a little too antagonistic). Anyway, I love this post. But it could still use some translating. Here, let's begin.
You can boil my claim about mathiness down to two assertions:
1. Economist N did X.
2. X is wrong because it undermines the scientific method.
#1 is a positive assertion, a statement about “what is …”#2 is a normative assertion, a statement about “what ought …”And now, in blog-ese:
Basically, my "mathiness" claim comes in two parts. First, I'm trying to expose some things I see some economists doing. Second, I'm trying to explain why I think those things are wrong, because they undermine the scientific method.Romer:
As you would expect from an economist, the normative assertion in #2 is based on what I thought would be a shared premise: that the scientific method is a better way to determine what is true about economic activity than any alternative method, and that knowing what is true is valuable.
In conversations with economists who are sympathetic to the freshwater economists I singled out for criticism in my AEA paper on mathiness, it has become clear that freshwater economists do not share this premise. What I did not anticipate was their assertion that economists do not follow the scientific method, so it is not realistic or relevant to make normative statements of the form “we ought to behave like scientists.”
In a series of three posts that summarize what I have learned since publishing that paper, I will try to stick to positive assertions, that is assertions about the facts, concerning this difference between the premises that freshwater economists take for granted and the premises that I and other economists take for granted.Blog-ese:
I personally believe very strongly in the scientific method. I thought that pretty much all economists believed in it. But in my conversations with economists who are sympathetic to the freshwater folks I criticized in my "mathiness" paper, I realized that a lot of them don't actually believe in the scientific method at all. Or at least, they don't believe economists ought to follow it!
As you might expect, I was pretty surprised to find this out. Surprised, and dismayed. So I'm going to write three posts about my conversations with these people. I'll try my best to stick to the facts. I just want to explain how these people think differently from the way I do.Romer:
In my conversations, the freshwater sympathizers generally have not disagreed with my characterization of the facts in assertion #1–that specific freshwater economists did X. In their response, two themes recur:
a) Yes, but everybody does X; that is how the adversarial method works.
b) By selectively expressing disapproval of this behavior by the freshwater economists that you name, you, Paul, are doing something wrong because you are helping “those guys.”
In the rest of this post, I’ll address response a). In a subsequent post, I’ll address response b). Then in a third post, I’ll observe that in my AEA paper, I also criticized a paper by Piketty and Zucman, who are not freshwater economists. The response I heard back from them was very different from the response from the freshwater economists. In short, Piketty and Zucman disagreed with my statement that they did X, but they did not dispute my assertion that X would be wrong because it would be a violation of the scientific method.Blog-ese:
The freshwater sympathizers generally don't deny that the freshwater people do the things I accused them of doing. Instead, their response was basically "Everybody's doing it." Also, they told me that by calling them out on it, I was giving aid and comfort to the forces of darkness.
For right now, let's talk about the defense that "Everybody's doing it."Romer:
Together, the evidence I summarize in these three posts suggests that freshwater economists differ sharply from other economists. This evidence strengthens my belief that the fundamental divide here is between the norms of political discourse and the norms of scientific discourse. Lawyers and politicians both engage in a version of the adversarial method, but they differ in another crucial way. In the suggestive terminology introduced by Jon Haidt in his book The Righteous Mind, lawyers are selfish, but politicians are groupish. What is distinctive about the freshwater economists is that their groupishness depends on a narrow definition of group that sharply separates them from all other economists. One unfortunate result of this narrow groupishness may be that the freshwater economists do not know the facts about how most economists actually behave.Blog-ese:
First, a pessimistic note. I've come to realize that freshwater economists are just...different. I'm more and more convinced that the freshwater people are used to playing politics instead of doing science. I'm talking about academic politics, of course, not national politics - you know, where "the battles are vicious because the stakes are so low."
The sad thing is that the freshwater folks think they're no more political than everyone else. They really do believe that "everybody's doing it." They're wrong, but that's what they think.Romer:
In my informal conversations, no one from the freshwater camp has articulated exactly what they mean by the adversarial method, so I’ll try to fill in the blanks here. (It would be helpful if someone who supports this view indicated whether I have characterized it accurately.)
In an equilibrium where everyone follows the adversarial method, each side tries to make the best possible case for its position. What we might call the rules of evidence are that an advocate cannot make statements that are false, but it is to be expected that an advocate will withhold information that does not support the advocate’s position.Blog-ese:
So if the freshwater camp doesn't follow the scientific method, what do they follow? I call it the "adversarial method". The adversarial method is where each side tries to make the best possible case for its position - not lying, not saying anything false, but omitting any evidence that might support an alternative point of view.Romer:
As always, it helps to be specific. One of the issues that I raised in a conversation concerned the support (in the mathematical sense) of the distribution of productivity in the Lucas-Moll model. Their assumption 1 states that at time 0, the support for this distribution is an unbounded interval of the form [x,infinity). In response to objections of the general form “this assumption is unacceptable because it means that everything that people will ever know is already known by some person at time 0,” Lucas and Moll present a “bounded” model in which the support for the distribution of productivity at time zero is [x,B]. Using words, they claim that this bounded model leads to essentially the same conclusions as the model based on assumption 1. (I disagree, but let’s stipulate for now that this verbal claim is correct.) I observed that Lucas and Moll do not give the reader any verbal warning that because of other assumptions that they make, the support for the distribution of productivity jumps discontinuously back to [x,infinity) at all dates t>0 so it is a bounded model in only the most tenuous sense.
When I suggested that this omission concerning the behavior of the support of the distribution was a sign of something wrong, the response I got was that in presentations of the paper, members of the audience and/or discussants pointed out that this is how the support behaves. This implied that it was evident to someone who digs into the math that the support for the distribution is bounded only at date 0 and not at any other date. Because this result could be inferred from a careful examination of the math, the authors met their obligations to the reader. Because the support is bounded at date 0, they did not, technically, make a false statement when they use the word “bounded” to describe the model’s behavior. Beyond this, they should not be expected to say in words anything that would weaken their argument.
I defined mathiness as a combination of words and symbols in which the meaning of the words diverges from the meaning implied by the symbols. One of the ways they can diverge is if the words selectively disclose only some of the formal results. Calling their model “bounded” is an example of this type of divergence.
The freshwater sympathizers agreed, for example, that Lucas and Moll strategically refrained from verbal disclosures about some of the properties of the underlying mathematical formalism. Where we disagreed was whether this was a sign of behavior by the authors that is wrong. In effect, their response was caveat emptor; this is what all economists do.Blog-ese:
For example, I was talking to someone about the Lucas-Moll model that I complained about earlier. Lucas and Moll assume that at time 0, the distribution of productivity has support over the interval [x, infinity) - in other words, that everything that can possibly be known is already known by someone out there. That's totally unrealistic. In the real world, there are a bunch of things that nobody knows yet.
So then Lucas and Moll claim that they have a "bounded" version of the model, where at the beginning, only a finite number of things are known - the productivity distribution starts out with support [x,B] at time 0. They say this model leads to the same conclusions as the earlier one (I disagree, but that's a story for another day). But here's the trick that Lucas and Moll pull...as soon as time starts, the distribution of productivity instantly goes back to [x, infinity)! In other words, in the microsecond after time begins, humanity discovers an infinite number of new things!
If that isn't mathiness, I don't know what is.
The thing is, Lucas and Moll stick this assumption in there with zero warning to the reader. If you don't look incredibly closely and dig through the math, you never even notice that this is what they're doing. To take a crazy-sounding assumption like this and just bury it in the math and hope no one notices it - well, that's mathiness, right there.
So I complained about this to a freshwater sympathizer. (S)he told me that, well, this issue came up in seminars, so people definitely noticed it. And if people noticed it, (s)he said, I shouldn't be worrying. Did I expect Lucas and Moll to go out of their way to point out this flaw to readers?
Well...yeah, I kind of did.
Science shouldn't be a caveat emptor situation. If you're a scientist, you have a duty to point out the weak points in your arguments instead of forcing people to go hunting around for them.Romer:
For me, the most interesting thing to emerge out of these conversations was the realization that the difference between us, which seems like an intractable difference in values or beliefs about right and wrong, may simply reflect different inferences the facts. We agreed about what about what Lucas and Moll did but we disagreed about what other economists do.
One way to characterize the underlying disagreement about what is wrong is to say that we are both commenting on strategic interaction between economists that takes the form of a repeated, multi-player prisoner’s dilemma. In this game, following the scientific method corresponds to cooperation; following the adversarial method corresponds to defection. My claim is that economics is characterized by a trigger strategy/reputational equilibrium that sustains the scientific method. In calling attention to defection by the freshwater economists, I am following a strategy that is equilibrium play in this reputational equilibrium.
In contrast, the freshwater economists believe that we are already in the noncooperative adversarial equilibrium, so it is wrong to express disapproval of economists who are simply engaging in the type of behavior that is rational in that equilibrium. The freshwater economists might agree that in some first-best sense, an equilibrium based on the scientific method would be preferable, but they apparently believe that we are not in such an equilibrium; that it is not possible to get back to such an equilibrium; and that even if we did, it would not be possible to sustain it.
If so, what looks superficially like a deep and intractable disagreement about values or morality may simply reflect disagreement over the facts about what most economists do. When the freshwater types say “everybody is following the adversarial method,” what they may honestly be saying is that “everybody I know is following the adversarial method and they all believe that everyone else is doing this too.” But because freshwater economists have so sharply isolated themselves from the rest of the profession, they may be generalizing from an unrepresentative set of observations.
To me, the facts seem to be that freshwater economists are following a coordinated strategy based on the adversarial method yet that many other economists are still committed to the scientific method.Blog-ese:
So what explains the difference between my approach - which I think is the mainstream approach in econ - and this adversarial freshwater strategy? I have a little pet theory as to why they do what they do.
Basically, scientific dialogue is like a repeated, multi-player prisoner's dilemma. Following the scientific method is "cooperation"; following the adversarial method is "defection". If you know about repeated multi-player prisoner's dilemmas, you know about trigger strategies and reputational equilibria.
I've always followed the scientific method because I thought I was in a cooperative reputational equilibrium. I pointed out the weaknesses of my theories without being prompted, because I thought everyone else would do the same for me. But the freshwater folks didn't share my beliefs. They think they're in a noncooperative adversarial equilibrium, where everyone will always defect, defect, defect. They might wish that everyone followed the scientific method, but they think no one does, so they better not follow it either. They're simply carrying out their rational best responses, given their cynical, pessimistic beliefs about the world.
So these freshwater folks aren't bad guys. They're just cynical. We don't disagree about values, we just disagree about the facts. They've isolated themselves from the rest of the profession - walled themselves off inside their little freshwater bubble - so they don't realize that in the outside world, people are playing a much friendlier, more cooperative game.Romer:
If I could persuade the thoughtful freshwater types of just one point, it would be that they owe it to themselves to assess the factual basis for their belief that all economists follow the adversarial method instead of the scientific method. Everyone they talk to may honestly believe this, but a bit of empirical inquiry may reveal that this is a fact about whom they talk to, not a fact about the behavior of most economists.Blog-ese:
If I could tell the freshwater economists just one thing, it would be that the rest of economics is doing things differently. Really. We're out here being honest with each other, trying to get to the truth together, not politicking for our own pet theories. We're being scientists. You can too. If you get outside your bubble, you'll see I'm telling the truth.Anyway, so that's my translation of a Paul Romer post. The cool thing is that, due to the magic of blogs, if I've translated anything incorrectly or incompletely, I can just go back and correct it!
Also, remember that I'm just the translator. I don't necessarily endorse the things Romer says here. Personally, I think his theory about the freshwater folks sounds a bit far-fetched - it relies on some pretty persistent heterogeneous beliefs. Surely the freshwater folks wouldn't take too long to realize that they were doing things very differently than others...
Update: Romer's more recent posts are adopting a more bloggy tone, and - in my opinion - are much more fun and readable!