Why Policy Debates Are Like Law School Exams, Not Economic Models
-- Posted by Neil H. Buchanan
I expect this to be my final entry in this series of posts about economic methodology. (Of course, I cannot guarantee that something new will not move me to return to these issues at some point.) Readers can find links to the previous five entries in the series via yesterday's post.
Here, I want to explore a core reason why I continue to worry about the tendency among left-leaning mainstream economists (whom I have been referring to as the Orthodox Left, led by Paul Krugman) to think that it does no harm to start with conservative models and tweak them to generate more realistic (and liberal) outcomes. I say this even though I do sincerely believe, as I have said multiple times, that the Orthodox Left has been almost completely right about nearly everything for the last seven years or so. The issue, therefore, is not about the recent past, but about lessons for the future.
An analogy might help. When I was in my first year of law school, I was sitting on the students' side of the lectern for the first time in fifteen years, having spent the intervening time as an economics teaching assistant and then professor. Unlike most 1L's, I did not think that I needed to worry about exam-taking strategies, and all of that "think like a lawyer" stuff. My many lawyer friends told me that it was all ultimately about logic and clear thought, which I had reason to think that I could at least fake. The Fall semester curriculum included the required Constitutional Law course. What joy! I loved every minute of the class, the professor and I had great talks outside of class, and everything looked like it was falling into place. I glanced over a few sample exams, reviewed my notes, and took the exam.
Oops. When grades for that course were reported, the lowest grade of my academic career stared me in the face. I went to the professor's office hours, and we went over the exam together. For each question, I asked: "OK, what did I get wrong?" Each time, the professor said, "Well, you really didn't get it wrong." Imagine my confusion. After some further conversation, it finally became clear to me that on some of the questions I had written the best right answer, while on others I had written an answer that was still right, but I had not written the best right answer. But that was not the whole story, because the professor explained to me that I was supposed to have explained, for each question, a range of possible approaches, describing the strengths and weaknesses of each, and explaining why the best right answer was truly the best.
Dorf on Law readers who have attended law school are surely now saying, "Well, duh!" This is what issue-spotter exams are supposed to do. How could I not know that? The answer, of course, is that I was purely an economist, and I had not bothered to find out that the educational enterprise in law is different. When I spoke with my ConLaw professor, I had a difficult time not being sarcastic, because all I could think was, "Wait a minute, I was right on every question, but I didn't get an A+. How ridiculous is that?" Economists, especially because so much of their self-worth is tied up in being "scientific," which mostly translates into doing a lot of math, simply look for the right answer. If you figure that out, then you win.
This is not, however, merely a matter of pedagogical differences. The law school model is based, in the first instance, on the idea that legal cases might be won on any number of bases. We "argue in the alternative," discussing different routes to victory for our clients. Even when the matter is not to be litigated, the logic of the law school approach is that whoever assesses one's arguments might surprise us, accepting what we think are the weaker arguments and never even taking the "right answer" seriously. (Think, for example, of a slam-dunk promissory estoppel claim. It would be malpractice to rely solely on such a claim, no matter how strong, given the hostility of many judges to the very existence of the promissory estoppel doctrine.) The best lawyers develop as many arguments as they can, rather than putting all of their proverbial chips on one spin of the roulette wheel.
This, therefore, brings us back to my first post in this series, where I suggested that it was a mistake for Paul Krugman not to argue against the false claim (based on orthodox production functions) that capital and labor are paid according to their respective levels of productivity. That conclusion is at the core of orthodoxy, and the orthodox right uses it to say that the government should not intervene in the economy to "tilt" the playing field toward labor, because that would inefficiently undermine the presumptively-correct market outcomes. Krugman is right that, even if we grant that capital itself is paid fairly, we still have a strong argument that capital should not be owned by so few people (however much it is paid). But what happens if we lose that argument? Where is the fallback position? And is that even the best argument?
Similarly, we can return to an argument repeatedly offered by orthodox conservative economists like Greg Mankiw, who claim that the government should simply enforce the laws and then get out of the way. Regarding Piketty's blockbuster book, for example, Mankiw wrote: "Like President Obama and others on the left, Piketty wants to spread the wealth around. Another philosophical viewpoint is that it is the government’s job to enforce rules such as contracts and property rights and promote opportunity rather than to achieve a particular distribution of economic outcomes. No amount of economic history will tell you that John Rawls (and Thomas Piketty) offers a better political philosophy than Robert Nozick (and Milton Friedman)."
Contrast this with a quotation from Tom Palley, which I have referenced in earlier posts: "The conventional character of Piketty’s theoretical thinking rears its head in his policy prescriptions. His neoclassical growth framework leads him to focus on taxation as the remedy. There is little attention to issues of economic institutions and structures of economic power because these are not part of the neoclassical framework." Palley's point is much broader than the specific question about using tax policy, because his argument further suggests that using a neoclassical (orthodox) framework, even from the left-leaning perspective, leaves one ill-equipped to respond to Mankiw. An analyst who starts from neoclassical premises has accepted the baseline of current laws (which determine the distribution of wealth), and thus cannot ask Mankiw the obvious questions: Where did those rules come from, and why are they beyond question?
On that territory, the left's task is not impossible, of course, but it is much more difficult if one cannot challenge the baseline. As I put it in a recent post: "Conservative economists (and, to be clear, many liberal economists who buy into the basic framework of modern economics) take for granted the body of laws that allow a modern economy to function. When criticizing a policy as 'inefficient,' the unstated assumption is that the other laws and policies are the baseline from which we can measure deviations from efficiency."
Again, it is still possible to win arguments with conservatives, even when working from their baselines. Sometimes, in fact, it is embarrassingly easy, which is what Krugman is talking about when he mocks the half of the orthodox economics profession who cannot even get the basic IS/LM analysis right during a liquidity trap that lasts for years.
Krugman, however, apparently wants to believe that it is always that clear and easy. When he took a swipe at Jamie Galbraith's critique of Piketty's book, Krugman wrote: "And what’s going on here, I think, is a fairly desperate attempt to claim that the Great Recession and its aftermath somehow prove that Joan Robinson and Nicholas Kaldor were right in the Cambridge controversies of the 1960s." But Galbraith was most definitely NOT talking about the Great Recession and its aftermath. He was talking about the debate over the concentration of wealth, which was Piketty's subject, and saying that Piketty's stripped-down orthodox approach gives away that game. Krugman himself has noted with some amusement that, in the midst of the ongoing damage from the Great Recession, the Big Debate about economics is a death match about long-term trends in wealth concentration. Debating about Piketty is not debating about short-term macro.
And if we try to fight the inequality debate only with Piketty's tools, which means accepting the production function-based orthodox theory of distribution, then anything that people like Krugman argue will be deemed "inefficient," as a baseline matter. One can overcome that baseline, but the infrastructure of neoclassical theory is a rather large presumption to rebut. And given that we actually know that production functions are logically incoherent (thanks to those British Cantabrigians whom Krugman dismisses, especially Robinson and Kaldor), the baseline should definitely not be that, absent a "special case," capital and labor are paid their appropriate shares of national income.
This is a very real debate, with on-the-ground consequences for policy, and the theories that we use will determine the arguments that we can make. Just because we can see a way to get the "right" answer via one path does not mean that it is wise to give up others, especially when those others pose far fewer hurdles. This is like law school, not an economic theory course.
I expect this to be my final entry in this series of posts about economic methodology. (Of course, I cannot guarantee that something new will not move me to return to these issues at some point.) Readers can find links to the previous five entries in the series via yesterday's post.
Here, I want to explore a core reason why I continue to worry about the tendency among left-leaning mainstream economists (whom I have been referring to as the Orthodox Left, led by Paul Krugman) to think that it does no harm to start with conservative models and tweak them to generate more realistic (and liberal) outcomes. I say this even though I do sincerely believe, as I have said multiple times, that the Orthodox Left has been almost completely right about nearly everything for the last seven years or so. The issue, therefore, is not about the recent past, but about lessons for the future.
An analogy might help. When I was in my first year of law school, I was sitting on the students' side of the lectern for the first time in fifteen years, having spent the intervening time as an economics teaching assistant and then professor. Unlike most 1L's, I did not think that I needed to worry about exam-taking strategies, and all of that "think like a lawyer" stuff. My many lawyer friends told me that it was all ultimately about logic and clear thought, which I had reason to think that I could at least fake. The Fall semester curriculum included the required Constitutional Law course. What joy! I loved every minute of the class, the professor and I had great talks outside of class, and everything looked like it was falling into place. I glanced over a few sample exams, reviewed my notes, and took the exam.
Oops. When grades for that course were reported, the lowest grade of my academic career stared me in the face. I went to the professor's office hours, and we went over the exam together. For each question, I asked: "OK, what did I get wrong?" Each time, the professor said, "Well, you really didn't get it wrong." Imagine my confusion. After some further conversation, it finally became clear to me that on some of the questions I had written the best right answer, while on others I had written an answer that was still right, but I had not written the best right answer. But that was not the whole story, because the professor explained to me that I was supposed to have explained, for each question, a range of possible approaches, describing the strengths and weaknesses of each, and explaining why the best right answer was truly the best.
Dorf on Law readers who have attended law school are surely now saying, "Well, duh!" This is what issue-spotter exams are supposed to do. How could I not know that? The answer, of course, is that I was purely an economist, and I had not bothered to find out that the educational enterprise in law is different. When I spoke with my ConLaw professor, I had a difficult time not being sarcastic, because all I could think was, "Wait a minute, I was right on every question, but I didn't get an A+. How ridiculous is that?" Economists, especially because so much of their self-worth is tied up in being "scientific," which mostly translates into doing a lot of math, simply look for the right answer. If you figure that out, then you win.
This is not, however, merely a matter of pedagogical differences. The law school model is based, in the first instance, on the idea that legal cases might be won on any number of bases. We "argue in the alternative," discussing different routes to victory for our clients. Even when the matter is not to be litigated, the logic of the law school approach is that whoever assesses one's arguments might surprise us, accepting what we think are the weaker arguments and never even taking the "right answer" seriously. (Think, for example, of a slam-dunk promissory estoppel claim. It would be malpractice to rely solely on such a claim, no matter how strong, given the hostility of many judges to the very existence of the promissory estoppel doctrine.) The best lawyers develop as many arguments as they can, rather than putting all of their proverbial chips on one spin of the roulette wheel.
This, therefore, brings us back to my first post in this series, where I suggested that it was a mistake for Paul Krugman not to argue against the false claim (based on orthodox production functions) that capital and labor are paid according to their respective levels of productivity. That conclusion is at the core of orthodoxy, and the orthodox right uses it to say that the government should not intervene in the economy to "tilt" the playing field toward labor, because that would inefficiently undermine the presumptively-correct market outcomes. Krugman is right that, even if we grant that capital itself is paid fairly, we still have a strong argument that capital should not be owned by so few people (however much it is paid). But what happens if we lose that argument? Where is the fallback position? And is that even the best argument?
Similarly, we can return to an argument repeatedly offered by orthodox conservative economists like Greg Mankiw, who claim that the government should simply enforce the laws and then get out of the way. Regarding Piketty's blockbuster book, for example, Mankiw wrote: "Like President Obama and others on the left, Piketty wants to spread the wealth around. Another philosophical viewpoint is that it is the government’s job to enforce rules such as contracts and property rights and promote opportunity rather than to achieve a particular distribution of economic outcomes. No amount of economic history will tell you that John Rawls (and Thomas Piketty) offers a better political philosophy than Robert Nozick (and Milton Friedman)."
Contrast this with a quotation from Tom Palley, which I have referenced in earlier posts: "The conventional character of Piketty’s theoretical thinking rears its head in his policy prescriptions. His neoclassical growth framework leads him to focus on taxation as the remedy. There is little attention to issues of economic institutions and structures of economic power because these are not part of the neoclassical framework." Palley's point is much broader than the specific question about using tax policy, because his argument further suggests that using a neoclassical (orthodox) framework, even from the left-leaning perspective, leaves one ill-equipped to respond to Mankiw. An analyst who starts from neoclassical premises has accepted the baseline of current laws (which determine the distribution of wealth), and thus cannot ask Mankiw the obvious questions: Where did those rules come from, and why are they beyond question?
On that territory, the left's task is not impossible, of course, but it is much more difficult if one cannot challenge the baseline. As I put it in a recent post: "Conservative economists (and, to be clear, many liberal economists who buy into the basic framework of modern economics) take for granted the body of laws that allow a modern economy to function. When criticizing a policy as 'inefficient,' the unstated assumption is that the other laws and policies are the baseline from which we can measure deviations from efficiency."
Again, it is still possible to win arguments with conservatives, even when working from their baselines. Sometimes, in fact, it is embarrassingly easy, which is what Krugman is talking about when he mocks the half of the orthodox economics profession who cannot even get the basic IS/LM analysis right during a liquidity trap that lasts for years.
Krugman, however, apparently wants to believe that it is always that clear and easy. When he took a swipe at Jamie Galbraith's critique of Piketty's book, Krugman wrote: "And what’s going on here, I think, is a fairly desperate attempt to claim that the Great Recession and its aftermath somehow prove that Joan Robinson and Nicholas Kaldor were right in the Cambridge controversies of the 1960s." But Galbraith was most definitely NOT talking about the Great Recession and its aftermath. He was talking about the debate over the concentration of wealth, which was Piketty's subject, and saying that Piketty's stripped-down orthodox approach gives away that game. Krugman himself has noted with some amusement that, in the midst of the ongoing damage from the Great Recession, the Big Debate about economics is a death match about long-term trends in wealth concentration. Debating about Piketty is not debating about short-term macro.
And if we try to fight the inequality debate only with Piketty's tools, which means accepting the production function-based orthodox theory of distribution, then anything that people like Krugman argue will be deemed "inefficient," as a baseline matter. One can overcome that baseline, but the infrastructure of neoclassical theory is a rather large presumption to rebut. And given that we actually know that production functions are logically incoherent (thanks to those British Cantabrigians whom Krugman dismisses, especially Robinson and Kaldor), the baseline should definitely not be that, absent a "special case," capital and labor are paid their appropriate shares of national income.
This is a very real debate, with on-the-ground consequences for policy, and the theories that we use will determine the arguments that we can make. Just because we can see a way to get the "right" answer via one path does not mean that it is wise to give up others, especially when those others pose far fewer hurdles. This is like law school, not an economic theory course.