=============================================================================== Comments on Richard Musgrave, "Devolution, Grants, and Fiscal Competition" Brad DeLong (1) The government may not have grown since 1980 as a share of GDP, and may have grown only slowly since 1960. But the civilian government has surely grown significantly faster. And it is the civilian government that people--at least Republicans--are most worried about. Defense is an unfortunate necessity. Civilian government is a vice. (2) I would move the discussion of distribution up ahead of the discussion of the shared task of provision. Distribution--even the idea that one should worry about distribution--is likely to drop out of economists' discussions remarkably early. (3) I think that the (brief) discussion of macro policy belongs in the introduction to the paper, as a single paragraph. It breaks up the flow, and the point--that macro policy can only be done at a macro level--is straightforward. (4) I like the way the hierarchy of taxes is laid out: federals can do progressive income taxes; states must rely on broad-based (but regressive) sales taxes; localities can do a version of benefit-related taxation by taxing real estate. Is there a similarly crystal clear way to lay out a hierarchy of spending programs? (5) Grants-in-aid have always struck me as very difficult to manage. The governors get the credit for the spending programs. The senators get the blame for levying the taxes. And then the governors run against the senators for office. Thus the constant temptation for the senators to grab credit by mandating exactly what the governors must do. But in that case, why not simply have a federal-level program? Your vision of public finance, though, has the federal government raising revenue through a progressive income tax, and then has much of it being spent by the states for standard fiscal federalism reasons. I have wondered over the past three years if what we are now seeing is not the beginnings of the breakdown of such a system--because of the vulnerability that the senators take the blame for the taxation. =============================================================================== Comments on Ed Glaeser, "Are Cities Dying" Brad DeLong General Comments ================ I think that this is going to be a great paper. But it the moment it reads too much like a Journal of Economic Literature paper: too many one-sentence descriptions of recent articles, and announcements that such and such is a "pressing topic for future research." I think that if an article isn't worth at least one (and probably two) paragraphs, it is not worth summarizing. And while it is fine--and too rare--for economists to say "we don't know," I think it is better to then give one's best guess than to just call for more research. The opening did not grab me. We are economists, so we take for granted the fact that we will be building an economic model of the city, and not just projecting trends. If you do keep the city summary statistics in the paper, I would use a longer time span: twenty years isn't enough. As you know better than I--but as our readers do not--the increase in the proportion of the population urban and in the size of our metropolitan agglomerations in the past century and a half has been truly remarkable. I also don't think that you will be able to avoid focusing on policy issues. Whether cities will grow or shrink is in some sense strongly dependent on how good a job city governments manage to do. A view of city growth requires as a prerequisite a view as to what governments will do to compensate for (or to increase) distortions. I would be much more explicit up front as to what my implicit (positive) model of policy is. Perhaps something about this century being the first century in human history in which most Americans and a large fraction of all humans live in cities. Will the same be true at the end of the next century? And then the answer--probably yes because ____. As presently written, the backbone of your paper's argument vanishes from the reader's attention for the middle half of the paper, only to reappear in the conclusion. I think it would be a stronger paper if you could keep the backbone more visible. And as I try to untangle the backbone of your argument, I think it goes something like this: (I) In the old days (up until 1950 or so) city growth was driven by three factors: (a) you needed to be located where transport costs were cheap, (b) you needed to produce on a large scale in order to reap economies of scale, and (c) face-to-face contacts were the best way to generate innovations and to learn how to master already-known technology. (II) In the new days transport costs have fallen by a lot--or, rather, shipping anything anywhere is now pretty cheap. In the new days economies of scale are much less important: there is something in this idea that we have moved from an age of mass production to one of flexible customized production. And information and communications technologies raise the possibility that face-to-face contacts will become less important. (III) So is there still reason to expect the growth of the city given its costs--congestion, crime, pollution, and so forth? (IV) Yes because the city still retains substantial productive advantages: face-to-face meetings are for some purposes infinitely more effective than telecommunications-mediated ones. (V) IV is reinforced by the fact that the diseconomies of urban life do not grow that rapidly with size, and that there is good reason to think that technological progress will continue to reduce the diseconomies of scale. The problem is that your evidence for (IV) is somewhat weak. The piece of evidence that I find most convincing is the price of houses between San Jose and Palo Alto: in the industry that ought to take to telecommuting the fastest, the premium for living near (and working in) the center of the industry is the highest. On the other hand, we are running an adequate journal without making you come to editorial article conference meetings? I think you need to reinforce the case for (IV) to make your backbone truly convincing. For point (V), you start out well. The section on congestion starts with the good point that commuting costs that are the primary disamenity of congestion--paying high housing prices and crowding together in tall buildings are ways of reducing commuting costs at the expense of some other margin. That seemed to me well-put. And commuting costs rise with size, but only slowly. Rather than continuing with the backbone of your argument, however, you then quickly ran through pollution, crime, riots, poverty, segregation, and anonymity. Crime you have to deal with: the literature does say that it is important as a determinant of city size. But the rest have a much looser linkage with your main argument. I would postpone them to the conclusion. I would start the conclusion with the prediction that telecommuting is not going to make the city go away. And I would then consider what the continued existence and growth of the city means for pollution, riots, poverty, segregation, and anonymity. One final remark. It takes me less time--about twenty minutes less, even in rush hour--to get to downtown San Francisco than it took a Berkeley resident in 1935, before the Bay Bridge was built, to get downtown. I live ten miles east of Berkeley. If commuting time is the primary disamenity, then there is a sense in which Greatest San Francisco today is not yet back to the total net commuting costs of the pre-WWII era, when ferries moved tens of thousands of people a day across a bridgeless bay. Are we justified in treating the distribution of population across the United States as reflecting a process in equilibrium? Perhaps the invention of the intra-urban "interstate" highway, the mortgage interest deduction, the mortgage loan guarantee, and the air conditioner vastly decreased the disamenities of suburbia, especially sunbelt suburbia. It seems to have taken generations for the real estate developers, the local politicians, and the city residents to recognize the consequences. Have they full recognized them even yet? Is there any warrant for the belief that the marginal sunbelt suburbia resident would be indifferent between Cupertino and Minneapolis, in a full-information sense? A bunch of your paper depends on the assumption that the people who have not yet moved to the sunbelt have made the right decision from their point of view. More Detailed Comments ====================== Page 2: How much of the urban nominal wage premium is simply compensation for higher prices in the city? Page 4: Health risks from cities are a thing of the past, no? I think that even raising the point that in previous centuries cities were sinks of disease and death is likely to confuse readers, especially as you raise the point in a somewhat cryptic way. Page 5: What is an MSA in New England? How are its boundaries determined? And what about situations in which MSAs but up against each other. Greater San Francisco feels like one metropolitan area--but I think it is three MSAs. Can you explain the Census Bureau's logic in splitting Greater New York and Greater San Francisco among MSAs? Page 10: Surely there are better examples of the fineness of the division of labor than parapsychologists? Page 11: Perhaps you could devote a little more space to summarizing Jaffe, Trajtenberg, and Henderson? It looks like the most solid piece of statistical evidence that we have. Page 14: Gaspar and Glaeser always struck me as a model badly in need of microfoundations. To say that "if face-to-face is not a substitute for but a complement to telecommunications, then it is ambiguous" is from one perspective to beg the question. Why can't e-mail and videoconferencing be modified to generate some random contacts? Page 16: I think (I could be wrong) when you say that higher housing costs... reflect the demand for the urban amenity and should not really be thought of as a cost of city living, you are slightly confused. Higher housing costs do not reduce utility below that of living in Northwood, North Dakota: urban amenities and urban job opportunities do compensate. But higher housing costs--and the fact that if the city grew housing costs would be even higher and would no longer be compensated by urban amenities--constrain the growth of cities. Page 16: The point that it is commuting costs that are the primary disamenity is very well-put. Page 17: You don't seem to have much to say about pollution, save that cities are not significantly more polluted. I don't think it deserves a subsection--just a sentence attached to the "commuter" subsection. Page 18: Reword your sentence: "Almost 50 percent of the effect [of city size on crime rates] can be attributed to a greater concentration of female-headed households in cities." Page 19: You need to say a little bit more about the hedonic estimates of the "cost of crime"--what assumptions are necessary for them to make sense, and so forth. Page 21: If (as you maintain) cities expand everyone else's economic opportunities, why should cities ever create poverty? If you are going to mention the possibility you need a mechanism. If you don't have a mechanism, you shouldn't raise the possibility. If William Julius Wilson were here, he would say that the poor live in cities because their parents lived in cities, and because the demand for unskilled labor has fallen. Over the next two generations, he would say, the ghetto poor are likely to slowly move out of cities (except to the extent that infrastructure and support services truly do make city life more comfortable for the poor than country or suburban life). The non-immigrant poor do not move as fast as other Americans, but they are still mobile--far more mobile than, say, Europeans. Page 22: Why do you think that the costs of segregation for segregated minorities will continue to rise? Page 23: For a lot of people, urban anonymity is a good thing: it means that your father (or your uncle Fred) can't tell you what to do. To the extent that we value individual freedom of choice (that is, think that uncle Fred probably would not do as good a job choosing for you as you would do), this needs to counterbalance any negative effect of "anonymity." =============================================================================== Comments on John Quigley, "Urban Diversity and Economic Growth" Brad DeLong Comments ======== I had five observations: (1) The first ten pages or so are very nice--and table 1 seemed to me to summarize a lot of information very concisely. (2) I found myself less excited about table 2: I am not sure that the math does much more than the simple statement that "the line of thought beginning with Dixit-Stiglitz makes it plausible that diversity and variety in consumer goods or producer inputs yield external scale economies even in the absence of all supernormal profits." You could save about four pages--and get to the synthesis of empirical results more quickly. (3) By the end of the paper, you have reported increasing returns or substantial externalities (i) to urban manufacturing, (ii) to urban educational levels, (iii) to the paper trails of innovations, (iv) to diversity of industrial structure, and (v) to the simple concentration of activity. Are the observed disamenities of cities--that it costs a fortune to live in San Francisco (or, indeed, anywhere between San Mateo and San Jose save for East Palo Alto), and that urba crime is relatively high--enough to account for the fact that cities are not even larger than they are? Or would we all be better off if we lived in Greatest New York? Is the failure to externalize spillovers the reason that, say, California has two big cities rather than one? (4) Perhaps we should have invited Jaffe, Trajteberg, and Henderson to contribute? I think there results are interesting enough to be worth at least a full page, and perhaps more... (5) Readers of your paper are going to wonder what happened to the urban crisis. If cities are so great, why is Washington D.C. bankrupt? Even if the low-density auto-centered cities of the sunbelt have a very bright future, what about the cities of the northeast and midwest? Is the "urban crisis" really just the result of the automobile, the superhighway, and the mortgage interest deduction lowering desired density, and leaving the city centers relatively undesirable--hence the place to which the poor are driven? Perhaps all that is needed is a paragraph at the start saying that this paper is about what is right with cities, not what is wrong with them. Perhaps more is needed. I am not sure. =============================================================================== Comments on Charles Wyplosz, "EMU: How and Why It Might Happen" Brad DeLong (1) The opening is quite good: the story of how the move to full capital mobility entailed either the end of stable exchange rates within Europe or the end of European nations' national monetary policies is quite convincing. And if you have abandoned an independent monetary policy, why not make it obvious (and credible) that you have done so? Thus an economic rationale for the single currency, in addition to the political rationale for the euro as a symbol that the process of European unity is still moving forward. But there is one thing I don't understand about the negotiations that led to Maastricht. Why the German demand that the euro be as strong as the DM? Didn't Helmut Kohl reunify Germany by promising the sun, the moon, and the stars to the eastern laender? Didn't he hope to finance a considerable amount of this wealth transfer to the German east by inflation? Wouldn't Kohl be much happier had the DM been weaker in the 1990s? Why did this preference not make itself felt in the EMU negotiations? (2) The description of the Maastricht process is quite good as well. But once again there is something I do not understand. Why include the debt-to-GDP criterion for "convergence" if no one is going to make it? Does the debt-to-GDP criterion exist so that the heads of government can decide to abandon the whole process in the spring of 1998? Readers in the United States will have a hard time following and understanding this particular fiscal convergence criterion. They will see individual European nations as more-or-less equivalent to American states. American states do not run up their debt to large multiples of GDP under the belief that eventually the Federal Reserve would feel compelled to monetize the New York State debt after a default. The last defaults of U.S. state governments came 150 years ago. State governments in the United States are very fearful of excessive borrowing lowering their debt quality ratings. Why the fear that things will be different in Europe? I do not understand how to make this clearer, because it is unclear to me (a) why people worry about debt-to-GDP enough to make it a convergence criterion, and (b) why people would then go ahead and commit to EMU even though this criterion is unmet. (3) It seemed to me that your discussion of country-specific shocks could be improved. There are four kinds of country-specific shocks. The first kind is the self-inflicted shock: Ronald Reagan comes to power and fiscal policy goes insane. The EMU straitjacket would presumably diminish the likelihood of such a shock in any European country. The second is a (rational or irrational) shift in investors' animal spirits as to how attractive a country is as an investment location. Here again EMU would diminish the magnitude of such shocks. The third is a shock to world demand or supply in some particular industry. And the fourth is the residual--shocks to domestic baseline consumption, or whatever. It seems, to me at least, that the discussion from page 8 to page 11 could be improved by making more distinctions. (4) A counterexample to the case of New York City in 1975 is the case of Washington D.C. in the past two years. A lot of money has been pumped into D.C., but at least so far the central government has not wielded its power to demand massive institutional reform very heavily. Of course, a national capital is a special case... (5) I am not sure that we know enough to speculate about the impact of the euro outside of Europe. I would probably drop the section on "EMU and the Rest of the World." But my preference is not a strong one at all. (6) Why do populist politicians everywhere stress their opposition to EMU? Why don't they attack the central bankers, political oligarchs, and economic ruling classes who have benefited in the past from the absence of a common European currency? I do not understand the populist political economy of EMU at all well. (7) What is likely to give if EMU goes forward? Will the unemployed burn the European Monetary Institute? Will Europe undergo another Great Depression? Herbert Stein says that: "If something can't go on forever, it will stop." But how will it stop? =============================================================================== Comments on Martin Feldstein, "The Political Economy of European Monetary Union: Political Reasons for an Economic Liability" Brad DeLong (1) My first reaction is that this is not turning into as much of a "debate" as I had expected. Charles Wyplosz seems definitely "wet"--or at least suffering from "rising damp." So this paper's stress that EMU is, ultimately, a political move in a very long-run game between Germany and France that may (or may not) save us someday from yet another Franco-German war is, I think, very helpful. That said, it seemed to me that pages 3 through 12 were somewhat longer than they ought to be. Could the laying-out of the politics of EMU be compressed somewhat? It is clear that the internationalists see EMU as a next step that has to be taken if they are ever to get to a European foreign policy. It is less clear (to me at least) that they are correct. I wonder if the point that the internationalists think that EMU is a political necessity could be made less discursively, and then the paper could move on. (2) Konrad Adenauer certainly believed that Germany's soul could be saved only if it were centered on the Rhine, looking west; and that Germany was damned should it wind up centered in Berlin, looking east. Perhaps German elite belief in EMU is as simple as trying to keep in balance--that reunification was a step to the east, which needs to be followed by a step to the west? My view has always been that the French expectation that France and Germany would form an inner caucus within the European Union has a good chance of coming to pass, so my bet would be that the wishful thinking that is taking place is taking place in Frankfurt and Bonn. But I have no solid evidence to point to. (3) When I talk to Europeans, the benefit that many of them stress that they expect from EMU is not that they will no longer have to exchange money but, instead, that their governments will no longer be able to pursue unrealistic and unsustainable macroeconomic policies. They argue that some reduction in the homeostatic mechanisms of interest and exchange rate shifts that keep the economy in balance is worthwhile if the rules of EMU keep governments from pushing economies out of balance. For the past quarter century attempts at fixing exchange rates in order to constrain governments have been miserable failures, because exchange rates that can be fixed can be un-fixed, and governments and speculators know this. Perhaps a single currency system would be different? It seems to me that the question of how (or if) governments will behave differently after EMU could be profitably considered before the section on the effects of EMU on cyclical unemployment. (4) How could Mitterand possibly assure the French public that the ECB would not make European (and thus French) monetary policy? Did he know what he had signed? The charter of the ECB looked airtight to me, no matter what "stability council" might be set up. Am I naive? (5) I cannot understand the "division of responsibility for general exchange rate policy between the ECB and the ECOFIN Council." But, then, I never understood how American monetary policy could be conducted by the Federal Reserve and exchange rate policy conducted by the Treasury. It seemed to me that the Secretary of the Treasury occasionally talked about the exchange rate, and that markets did or did not react depending on whether they did or did not think that he had enough points with the FOMC that it would avoid making him look like a fool, or think that he had insight into what the FOMC was going to do. It seemed to me that the Secretary of the Treasury occasionally bought or sold dollars for DMs, but did so on the same scale as a middle-rank trader at Goldman Sachs, and to call this an "intervention" rather than a "speculation" was to abuse language. It seemed to me that this "division of responsibility" was largely a matter of the Federal Reserve not making it painfully obvious that the Treasury Secretary had little more influence on exchange rate policy than did Ted Truman's most junior staf economist. Am I wrong? (6) It seemed to me that the fear that EMU will retard structural reform in Europe is not clearly right--that it is as likely to go the other way. I think that Alan Greenspan believes that "structural" reform in Europe is next to impossible as long as people look to the central bank to reduce unemployment without requiring reworking the social insurance state. It is possible that once the ECB is in place, European governments will face up to their obligations to reform the social insurance state to boost employment. It is also possible that an ECB could make a grand bargain with the EU--you do structural reform and we will boost aggregate demand--that no individual European central bank that seeks a stable exchange rate could make in the past. =============================================================================== Comments on Robert Inman and Daiel Rubinfeld, "Rethinking Federalism" Brad DeLong (1) You bring the discussion of Heller-Pechman style "revenue sharing" to a close at the top of page 3 without its punchline. Presumably we don't think that this ended well? Certainly the general revenue sharing funds got eaten away pretty rapidly, as senators began to wonder why the governors should get all the credit. (2) One does hope that the more decentralized direction is being charted using considered principles of federalism. But one knows that that is not the case. I would be somewhat stronger here: that the move in a decentralized direction is not likely to turn out that well because it is not being guided by considered principles of federalism. (3) I found the discussion of the principle of economic federalism quite good. I found myself wondering, however, if some of the later discussion couldn't be folded into this part of the paper--along the lines of "cooperative federalism helps solve this difficulty, but accentuates this one." It is not clear to me that it is worth cycling through the four principles one after the other. (4) The principle of constitutional federalism seemed to me to move in several directions at once--toward participatory democracy, toward reducing the size of the cental government in order to create more space for citizen action, toward reducing the government to preserve individual liberty, and so on. I would be inclined to prune back this section a lot, on the grounds that it takes us well beyond the economics of federalism (although into territory that any policy assessment will ultimately have to cover). (5) The TANF section worked well, I thought. In fact, I thought it worked well enough that if I were you I would move it to the front of the paper. Do the introduction; talk TANF; and then ask "what kind of federalism is this?"