I believe that the knowledge-return from adding to the data set and performing some simple statistical test is greater than the marginal benefit of running the same old over-studied data – such as the standard macro variables for the G-5 countries – through some pointlessly more sophisticated theory or econometric technique. In the 1980s, it became fashionable to claim that the real exchange rate followed a random walk, because statistical tests were unable to reject that null hypothesis at conventional significance levels. (Analogous claims were made about all sorts of variables in macroeconomics and finance.) But these tests were typically run on a few decades of data. I argued that one would not expect such limited data sets to offer enough power to reject the random walk even if mean reversion were the right answer. Economists had forgotten the lesson from introductory econometrics “failure to reject the null hypothesis does not necessarily entitle you to assert that the null hypothesis is true.” More provocatively (n “Zen and the Art of Modern Macroeconometrics”), I alleged that economists had subtly redefined the rules for a specific reason: it was too hard in macroeconomics to find statistically significant relationships. It is much easier to fail to find significant relationships. It hardly takes any work at all. But the affirmation “my research supports the hypothesis that the exchange rate follows a random walk” sounds much more respectable and publishable than “I have been studying exchange rates statistically for a year and have absolutely nothing to say about what makes them move.”
If one is in pursuit of the right answer, one needs to cast the net wider, to encompass a century-long time series, or a panel of countries. On a priori grounds, that is how much data it should take, before the test will have the requisite power. Sure enough, when one did that, one could reject a random walk in the real exchange rate, and find mean reversion.
Many have taken to using the “black swan problem” to mean a highly unlikely event, as the sub-prime mortgage crisis of 2007-08 is interpreted to have been. The way I would prefer to define it is when an event is considered virtually impossible by those whose frame of reference is limited in time span and geographical area, but that is well within the probability distribution for those whose data set includes other countries and other centuries (or those who make appropriate use of a priori theory, as with those irrational numbers). Analysts don’t cast the net widely enough. They can’t imagine that terrorists might inflict mass casualties by bringing down a buildings (New York, 2001) or that housing prices might fall in dollar terms (US, 2007) or that an advanced economy might suffer a loss of confidence in its debt (Greece, 2010). “I haven’t observed such a thing in the past, so it won’t happen in the future.” These things had happened before, but mostly in times and places far away. What do “black swans” have to do with it? An Englishman in the 19th century who encountered a black swan for the first time might have considered it a “7-standard deviation event,” even though the relevant information had already been available in ornithology books. 4
A voyeur in politics
I inherited an interest in politics from my father. One manifestation of it, to which I am a bit reluctant to confess, is a certain Zelig-like record of being a spectator at historic political events. At my current place of work, the Harvard Kennedy School, it is effortless to meet all the world leaders, one by one. In my youth I had to work harder at it. In the winter of 1971 I worked for George McGovern in the New Hampshire primary. One day in the summer of 1973, I got up early enough in the morning to be first in line to watch John Dean testify before Sam Ervin’s Senate committee ( that Richard Nixon had tried to cover up Watergate). In the summer of 1974, I was watching from outside the fence of the South Lawn of the White House as Nixon took off in his helicopter for the last time.
Later, I managed to get “better seats” in the Washington arena. I spent many hot summers in the nation’s capital, usually at either the Federal Reserve Board, the International Monetary Fund, or the Institute for International Economics. Those visits were highly rewarding, but strictly research. Then, in 1983, Martin Feldstein asked me to work for him at the Council of Economic Advisers. It was the Reagan Administration, of which I was not especially fond. But one reason why I happily took the job was the opportunity to work with Feldstein. There was extra prestige, at least in retrospect, from the fact that the position I was filling had been held during the preceding year by both Paul Krugman and Larry Summers. (I was single, worked very long hours in those days, and was happy to fill in for two.)
Surprising as this often is to outsiders, the Council of Economic Advisers is a rather technocratic, nonpolitical outfit. Making one’s best forecast of the trade deficit and the growth rate is the same in either a Republican Administration or a Democratic one. Trying to explain the virtues of free trade in an interagency meeting is the same in either case. Putting into a Presidential speech an explanation as to why a skeptical Congress must approve a quota increase for the IMF is the same.
My best Zelig story dates from November 4, 1983. At that point I did not yet have clearance to enter the White House proper, as opposed to the Old Executive Office Building next door, where the CEA resided. Nevertheless, through a chain of coincidences, I found myself in the Oval Office for half an hour with President Reagan, Vice President George Bush, Secretary of State George Schultz, OMB Director David Stockman, and others.5 Chit chat focused on the casualties from the recent Grenada invasion and the bombing of the Marine barracks in Beirut. Nobody asked me who I was, because they assumed that, if I was there, they should already know who I was. But at one point Richard Darman elbowed Ed Meese, covertly pointed at me, and whispered “who the hell is that?” I could infer this because Meese looked over at me, and then gave Darman a big shrug. Eventually, I figured out that I was at the wrong meeting, and left.
During this period, Feldstein popularized the notion of the twin deficits: that the then-new large US trade deficit was the result of a large budget deficit. The analysis was an implicit rebuke to those who had foolishly predicted that the tax cuts enacted in 1981 would lead to smaller budget deficits and higher national saving, rather than the reverse. Others in the White House and the Treasury rejected our forecast in the 1984 Economic Report of the President, that the trade deficit would continue to rise, let alone our diagnosis as to why. It made front page headlines when Secretary Regan responded to a question in congressional testimony by confirming that, so far as he was concerned, the ERP could be thrown in the trash. I couldn’t have been more pleased, though Feldstein was under enormous pressure, with frequent press reports that he was about to be fired and later that the CEA would be abolished. (Our forecast proved right on target next year.)6
With Feldstein, my array of mentors was pretty much complete.7 After his term on the CEA, Feldstein returned to Harvard and the Presidency of the National Bureau of Economic Research, with which I became increasingly involved. Later, he decided to divide the NBER’s International Studies program into a trade half and an International Finance and Macroeconomics half (IFM). Our forged-in-fire relationship was perhaps one reason why he asked me to be the Director of the IFM program. This position has helped me ever since to keep my fingers on the pulse of what is the hottest new research in the field and who are the young researchers doing it.
The position also made me a member of the NBER Business Cycle Dating Committee, which officially declares the starting dates and ending dates of US recessions. I came on board at the beginning of what turned out to be the longest economic expansion in American history (1992-2000). So for 9 years I could joke that I was on the best sort of scholarly committee: one that never had any reason to meet. But then came the recession of 2001. We dated the peak of the preceding expansion – that is, the beginning of the 2001 recession – as coming in March of that year, and the trough – the end of the recession -- in November.
Part of the job of being on the BCDC is being good-natured when observers react to our announcement of a business cycle dating point by questioning the need for the Committee, housed at a non-government research organization, the NBER. Most of the teasing takes one of two (mutually inconsistent) lines of argument. One is that everybody knows that a recession is defined as two consecutive quarters of negative GDP growth; so who needs the more complicated and less easily quantified procedures of the BCDC? The other line of argument is that “everybody has known for a long time” that the country has been in a recession, so it is ridiculous for the Committee to announce it so much after the fact.8 One rebuttal to both of these criticisms is that the relevant economic statistics come out with lags, are subject to major revisions, and often give signals that conflict with each other. Official GDP fell in the first and third quarters of 2001, but rose in the intervening second quarter. So if we had followed the simple two-consecutive-quarters rule of thumb, then we would not have found a recession at all. (We factored in other indicators, including job loss, to reach our judgment.)
At the time when we announced that the 2008 recession had begun with a peak toward the end of 2007, the government estimates still reported that the official GDP measure of output was actually higher in both the first and second quarters of 2008 than the last quarter of 2007! (We based our call on other indicators, such as job loss and the national income measure of output.) Much later, the Commerce Department revised its statistics, as it always does. The current estimates reassuringly show that GDP was in fact lower in both of the first two quarters of 2008 than in the last quarter of 2007. Even though our announcement of the beginning of the recession was greeted as long overdue when we made it, we would have had to wait another year and a half to get that crucial revision from the Commerce Department. Dating the ends of recessions is even tougher, by the way. The biggest headache in the last three recoveries has been that employment has lagged far behind the other indicators.
Incidentally, some Americans vaguely think that the terrorist attacks of September 11 caused the 2001 recession or the disappearance of the budget surplus that President Bush had inherited in January of that year. Of course both were in fact well underway before.9 But I don’t blame Bush for the 2001 recession.10 It usually takes a new president awhile before his actions have effects on current conditions, whether for good or ill.
I joined the faculty of the Economics Department of the University of
California at Berkeley – just a few miles from where I grew up – in 1979. I spent most of the 1980s and 1990s there. I loved walking to work, down the hill, along rose-lined paths and past redwood trees. I grew to enjoy teaching classes of 200 or 300 students.
When I first arrived, the Economics Department happened to be unusually short of faculty members who were close to me in either age or field. But eventually I was joined by Barry Eichengreen, Maury Obstfeld, Ken Rogoff, David Romer, and Christy Romer, who were close colleagues in both the personal and professional senses. And Andy Rose at the Business School. One of my few regrets in life arises from the circumstance that after they all came, and just as the Economics Department had been restored to its status as one of the top half dozen in the national rankings, I left to move East. Since the year I left, others in the Berkeley department have reaped an avalanche of Nobel Prizes and Clark Medals. My mixed feelings about having left derive, not from that, but from the good friends and colleagues that I left behind. And the landscape of my native state. I miss the redwood, live oak, and bay trees; the mountains; and the view of the water. No matter how lost you get in the Berkeley hills, you always know which way is West.
Member of the Council
In 1996, Joe Stiglitz, who then was Chairman of President Clinton’s Council of Economic Advisers, asked if I was interested in being a Member of the Council. This is a political appointment – not a staff job like I had had at CEA 13 years earlier. Thus it requires nomination by the President and confirmation by the Senate. The procedures for clearance and confirmation are among the many processes in Washington that are thoroughly broken. I don’t think the public understands how many top positions in policy-making are empty at any given moment, usually for the silliest of reasons. The Senate did not give me a hard time, in large part because we were in the midst of the strongest expansion in US history. I was sworn in by Vice President Al Gore a mere eight months after taking up residence.
There are three Members of the Council. The Chair is overall in charge. The other two divide up responsibility for issue areas. I had international economics, macroeconomics, and a few areas of microeconomic policy.
The main role of the CEA is presenting to the president and to others in the government, through the “interagency process,” what, in its view, the field of Economics has to say about the policy issues that need to be decided at the time. A hundred policy issues arise every month. The Council does not have any built-in constituencies, in the way that the Agriculture Department has farmers, Commerce Department business, Labor Department labor. Thus its influence is only as big or as small as the president or others choose to value its advice. Where most agencies have many “line responsibilities” – things that won’t get done if the agency doesn’t do them – CEA has only a few. On an annual basis, CEA writes the Economic Report of the President. On a daily basis, the Council writes confidential evening memos to the President explaining the official economic statistics that are to be released early the following morning. President Clinton got our memos on almost a daily basis, so great was his thirst for facts and figures. I know that some other presidents have been much less interested in such details.
We also had something called the Weekly Economic Briefing of the President. As soon as I arrived, I was struck how the WEB went into detail, such as explaining conflicting scholarly studies regarding the success of a school voucher program in Milwaukee. I was sure that this was more than the president needed to know. But I had not yet learned how different this president was from the one I had worked for in the 1980s. The next week, Clinton cited the conflicting evidence over the Milwaukee experiment in a campaign debate on national TV. After that, we kept the facts, figures, charts, and analysis flowing.
One “line responsibility” of the CEA is to lead the process, twice a year, to forecast the rate of economic growth and the other key macroeconomic variables that feed into the making of the federal budget. The Treasury and OMB (Office of Management and Budget) are the other two agencies that participate in the “troika.” I was fortunate to be there during a period when the economy repeatedly surprised all observers with good news on all fronts. It was a little embarrassing that the economists in the Administration kept under-forecasting economic growth. And unemployment, which macroeconomists had long said probably could not go below 5% without pushing up inflation, did so in 1997, and eventually went even below 4%. Every time we sat down to prepare a new forecast, some of the participants wanted to rely on the historical statistical relationships, while others argued that there had been a fundamental shift in the parameters due, in particular, to Information Technology. At the time, the latter sort of thinking was called the “New Economy.” Now it is called the “internet boom”, or even “internet bubble” – though it is important to realize that the economic performance was genuine and originally based on fundamentals11, even if the stock market got carried away by dot.com-mania, as it clearly did.
My approach was “Bayesian”: every six months, if the growth rate had again remained above traditional estimates of “potential” and the unemployment rate had again remained below traditional estimates of the “non-accelerating-inflation rate of unemployment,” with no signs of inflation, we would again adjust our estimates of those parameters just a little. But we would not throw in the towel, and jump the estimates discretely.] I told the staff that the year in which the government adjusted its estimates sharply in the optimistic direction would be the year that the stock market crashed and the economy entered recession. In the event, that is precisely what happened after we left. I am convinced that the grossly over-optimistic forecasts made by the government in January 2001, not just for the short term but for the long term as well, were a major reason why President Bush was able to convince the public that the budget surpluses he inherited not only would continue in the future, but would be so large that they required huge long-term tax cuts.
Two issues took up more of my time while I was on the Council than any other one topic or two topics. One was the emerging market crises that hit East Asia in 1997 and Russia in 1998. The other was the Kyoto Protocol on Global Climate Change, which was negotiated in November 1997. The first involved issues that were familiar to me. The parallels to the international debt crisis that began in 1982 were greater than most observers realized. The second issue was unfamiliar to me, and required a lot of hurried studying up, followed by a hundred inter-agency meetings.
It is one of the ironies of working in government that one can sometimes find far more room to influence policy in an area where one knows nothing, than in areas where one is putatively a world expert. While I, like most economists, was leery of the high economic costs if greenhouse gas emissions were to be cut very suddenly, I eventually became committed to the Kyoto Protocol.12 I thought that its design -- particularly the provisions for international trade in emissions permits and for trading off among the six greenhouse gases -- offered the best hope for addressing the environmental goal in an economically efficient way. The Protocol left a lot out, to be sure. The three biggest gaps that remain to be filled are full participation by all countries, a mechanism for setting emission targets well into the future, and any reason to expect countries to comply with their commitments. These are issues that I have done research on over the years subsequently.
I left the CEA in 1999. Rather than returning to Berkeley, I accepted a job offer from Harvard University’s Kennedy School of Government. Ten years earlier I hadbeen leery of moving from an Economics Department to a School of Public Policy. One obvious reason for the move in 1999 was that, by that stage in my career, I had developed some interest in participating in the public policy debate, which would be easier to do from Harvard Kennedy School than from the West Coast. Another reason is that I no longer thought I would be giving up much to get these benefits: I could continue to collaborate on research with Andy Rose via email, and could have lunch with all the excellent economists in the vicinity, just as easily at HKS as at Berkeley. To list only some of those who are located intellectually in international economics and physically in my building: Ricardo Hausmann, Robert Lawrence, Dani Rodrik, and (when not running Chile) Andres Velasco. But an advantage of Kennedy School is that it is in fact easy to partake from the elusive grail of interdisciplinary communication, for example at the faculty lunch seminar. Further, being a senior Professor at Harvard is a charmed status, as is being an undergraduate at Swarthmore, a graduate student at MIT, and an assistant professor at UC Berkeley.
At Berkeley, like any Economics Department, my teaching had been to either undergraduates or Ph.D. students. I still have some of both kinds of students at Harvard; but most of my students at the Kennedy School are masters students, who are in between. I like teaching these classes. In an Economics Department, there is a wide artificial gap between teaching undergraduates and teaching Ph.D. students. On the one hand, undergraduates like to hear about the real world, but there is a limit on how far you can go in terms of theory (though Harvard undergraduates, whom I teach in a course cross-listed in the Economics Department, are very smart). On the other hand, with Ph.D. students, they can do the math, but you are doing them a disservice if you talk about the real world and thereby give them the impression that if they do the same they will be able to write a thesis or get an academic job. The classes I teach now at Harvard Kennedy School are the best of both worlds, for me. I can mix theory and the real world. The combination is what they need and yet is intellectually satisfying for me.
I have always made sure that I lived walking distance from my place of employment. I live in Cambridge and either walk or bicycle to work, often noting when I pass over the spot near Harvard Square where George Washington took command of the Continental Army in 1775. My son’s elementary school is four blocks from our home. I enjoy walking him to school.
One of several big benefits of achieving tenure, and then Full Professorship, is that one can choose to work on whatever seems most interesting rather than whatever is most likely to demonstrate technical prowess and be published in the top journals. For me, this freedom included branching out in terms of subject matter, beyond the study of exchange rates and international financial markets. First, I ventured into other parts of macroeconomics, including, for example, the coordination of monetary and fiscal policy when different policy-makers believe in different models. Then, I ventured into other parts of international economics, including, for example, the circumstances under which the “trade-creating” advantages of regional free trade areas outweigh the “trade-diverting” disadvantages.
During the second half of my research career (so far!), I have ventured further afield still, into questions such as why some countries are able to achiever higher incomes than others and whether trade is bad for the environment.
Pontificating about big-think issues such as globalization has its role to play, if one wants to communicate with non-specialists or even to influence the public debate. But, as any academic knows, one doesn’t get articles published in refereed journals by writing judicious surveys of the literature or offering policy recommendations. One must, rather, contribute some incremental new methodological innovation, whether theoretical or econometric. (Preferably the outcome is to show why some other author is wrong; but one should at a minimum fill a supposedly “much-needed gap in the literature” [sic].)
Embarrassingly, a single econometric idea underlies quite a few of my journal articles over the last ten years, even though they appear in different sub-fields of economics. It has to do with geography.
I have been fascinated by geography my whole life, since before I got interested in economics in college.13 Although this must be a reason why I decided to specialize in international economics 30 years ago, at that time international economics had virtually nothing to do with geography. I am not talking about the reality that scholars who primarily specialize in a particular region of the world rank lower in the academic pecking order than those in the other sub-fields, which operate at a greater level of abstraction and generalization. I am talking about all the standard theories of international economics which sought to predict, say, the trade patterns or growth rates of countries, and that dealt with the set of actual real countries when it came to empirical analysis, and yet featured no role for such fundamental geographic variables as distance, landlockedness, language, or historical relationships. Rather, countries were disembodied points, which lacked any spatial coordinates and possessed only capital stocks, labor forces, productivity levels, money supplies, and a few other variables.