===============================================================================

Comments on John Gruber, "Policy Watch: Medicare and Medicaid"

Brad DeLong


A few comments and questions:

(1) Are you counting real or nominal Medicaid spending? I would prefer (a) real spending (b) divided by total U.S. population, and alternatively (c) deflated by the total U.S. poverty population.

(2) Move the point that you are only talking about one-third of Medicaid spending to page one. A lot of our target audience has no clue that Medicaid is also a program for the disabled and for the elderly poor.

(3) Why the Medicaid program creep starting in 1984? In a time of great fiscal stringency, who decided that the Federal Government should expand its financial commitment to fund medical care for children? Why didn't attempts to expand Medicaid spending get turned back by those to whom the budget dollars had previously been committed?

(4) Did the expansions in Medicaid in the 1980s really lower child mortality by one-twentieth? What else was at work in the 1980s? Even with low child mortality to begin with, this strikes me as a very large number? Am I off base?

(5) Does Congress have anything to do with increasing take-up? Isn't increasing take-up a policy that can be carried out, administratively, by the states and by HHS? Why haven't more states undertaken "baby your baby" programs?

(6) Do states and HHS have the power to--administratively--require copays and deductibles for Medicaid enrollees for non-preventive visits? If so, why hasn't HHS moved in this direction already?

(7) Your view is that Medicaid program creep is likely to continue. Is this the best way to expand health coverage for the uninsured?

===============================================================================

Comments on Chari, "Robert Lucas: Architect of Modern Macroeconomics"

Brad DeLong


(1) Do you think that "Expectations and the Neutrality of Money" has been more influential than "Econometric Policy Evaluation: A Critique?" I still teach the "islands" model of "Expectations," but I have the sense that fewer and fewer other people do--and that very few people believe that the driving force behind the business cycle is that people confuse nominal price changes due to inflation with changes in the real prices at which they sell and buy. 

If you think that Robert Lucas's main contribution has been the substantive contribution of dynamiting the foundations of macroeconometric models, and making economists change the way they analyze macroeconomic policy, then the "Econometric Policy Evaluation" paper is the key to Lucas's career. 

On the other hand, the methodological contribution from the invention of dynamic, stochastic general equilibrium models has been immense. How much of this is due to "Expectations and the Neutrality of Money"? What were its precursors? Who beforehand almost--but not quite--invented the dynamic stochastic general equilibrium model? If you think that the substantive conclusions of "Expectations and the Neutrality of Money" are outmoded and of only historical interest given the progress of real business cycle theory, then the key to Lucas's career is "Expectations" as a methodological model. 

On this there seems to be a pretty sharp split within the economics profession. Real business cycle types think that Lucas's key innovations are methodological and that his key paper is "Expectations." Non-real business cycle types tend to think that Lucas's key innovation is his way of thinking about how expectations need to be incorporated into models--both general equilibrium and non-general equilibrium, both stochastic and perfect foresight--and tend to see "Econometric Policy Evaluation" as the key to his career.

Chari is in the first camp. But I think the article would be strengthened if he noted the existence of the second camp, and explained why he found himself in the first.

(3) On page 5, Chari writes that "The central theoretical breakthrough of the last fifty years is that economists now think of 'equilibrium' as a rest point in the space of decision rules." This is a very nice way of putting the methodological innovation, and of placing Lucas's contribution in perspective. I suggest moving this insight way up to the front, and drawing parallels between Lucas on the one hand and Nash, Arrow, and Debreu on the other at several points throughout the paper.

(4)I have never understood Sims's critique of the Lucas critique. Sims is a very smart person. Yet he has come to the conclusion that the Lucas critique is not relevant, and that VAR (and macroeconometric models that close to replicate VARs) are fine for macroeconomic policy evaluation. Is it possible on page 7 to give a sketch of why Sims has not found the Lucas critique to be convincing?

(5) I'm not sure that there was a consensus among macroeconomists in the 1960s that the Phillips curve was really there. In his review of Keynes's General Theory, Jacob Viner wrote that Keynes's policies would produce full employment only as long as the printing press could keep ahead of the business agents of the labor unions, with the implication that the printing press's lead was not guaranteed. Hayek's The Road to Serfdom has a passage on how Keynesian policies can push employment higher than equilibrium only if workers are deceived about the value of their real wages, and workers cannot be deceived forever. Friedman and Phelps did not work in a vacuum.

(6) I think pages 11, 13, and 14 could be cut without hurting the article.

(7) I'm not sure that it is correct to say that the "evidence is quite mixed" as to whether anticipated and unanticipated changes in the stock of money have different effects on the U.S. economy.

(8) I have never understood why Kydland and Prescott (1982) was generally viewed as a more promising research program than Lucas (1972). Lucas's model has an observable driving variable--the stock of money--to which output, employment, and prices respond. Kydland and Prescott made no effort to specify an observable driving variable: there is no correlation between patents and output to rival that of money and output. Lucas (1972) had problems accounting for persistence. Kydland and Prescott (1982) had problems accounting for the lack of evidence of strong intertemporal substitution in microeconomic studies of individuals. It seems to me (and to Lucas) that Lucas's research program is still the more likely to bear fruit. Would it be possible to spend a little more time on the supersession at Chicago of the monetary-misperceptions research program by the real-business-cycle research program.

(9) Re: "Should the Federal Reserve raise interest rates in the summer of 1997?" I think that it would be better to jump right into the dynamic-consistency and the policies-as-rules-of-behavior literature here, and cut out most of page 27. Surely it is not illusory to think that economists can offer sensible policy advice. What Lucas has done is not to convince economists that "we are over our heads" in offering advice, but he has changed the way that economists go about offering advice. 

The right way to answer questions is now thought to be to reformulate them in linked terms as questions about the design and application of policy rules: "is raising interest rates in the summer of 1997 part of a rule for the conduct of monetary policy that will lead to good outcomes on average?" and "is the rest of this good rule for the conduct of monetary policy in place?" The answer can have all kinds of subtleties. For example, an action that is part of a good rule may be inadvisable if political forces are likely to bring to power a party that will not follow through on that particular rule. 


================================================================================

Jane Waldfogel, "Understanding the Family Gap in Pay for Women with Children"


Brad DeLong


A few comments and questions:


(1) I take it that the bottom line is that the U.S. is but retarded relative to other industrial economies in terms of pro-natal and pro-child policies. Are there any prospects for this to change? Is FMLA an exception, or the beginning of a shift?

(2) The coexistence of relatively advanced equal opportunity legislation with poor female relative wage performance could exist because the laws are a very weak tool. What evidence is there that the laws have worked?

(3) It is not clear to me from your numbers that the family gap is rising. Am I misinterpreting them? It seems to me that the family gap--women with vs. women without children--is more or less constant. The NLS and the CPS seem to carry somewhat different messages. Which do you trust?

(4) Just because no research to date has tested the hypothesis that employers discriminate against women with children doesn't mean it isn't true. Surely there is indirect evidence that would bear on this possibility?

(5) Your major story seems to be that maternity leave increases worker attachment to the firm, and increased attachment preserves the value of good matches between employers and employees. How large would these effects be--increasing the fraction of women covered by maternity leave by x% would have what effect on female wages? And can we tell if the U.S. firms that now offer maternity leave are those in which the value of the match is important? Is there a possibility that extending maternity leave would have little impact on female wages?

(6) Some more discussion of how you reconcile Gruber (1994) with the other studies would be welcome. Is maternity leave different? If so, why? If not, did Jon see things in the data that aren't really there?

(7) Nice conclusion, starting at the bottom of page 14. I would move much of page 13 to the front of the paper, and start the conclusion perhaps at the top of page 14.


================================================================================

Robert Brazelton, "The Economics of Leon Hirsch Keyserling"


Brad DeLong


A few comments and questions:

(1) When did Keyserling start graduate school? When did he leave graduate school? What, exactly, did he do after he left graduate school? And how did Truman find him a logical choice? The paths by which someone became a member of the CEA in the late 1940s are interesting. Yet this paper in its present form does not discuss them.

(2) What was the Wardman Park Hotel's "Monday Night Group"? How did Keyserling become a member of the same circle as Clark Clifford? Was Clifford his connection to the Truman Administration?

(3) Why quote from Hamby? Why not quote from Keyserling? Direct quotes always give much more of the flavor of the individual.

(4) I don't understand how Keyserling could believe in countercyclical policy but not in tightening in an inflation. Countercyclical policy is tightening in an inflation, and loosening in a recession.

(5) Is there a better way to express Keyserling's belief in economic growth? What policies did he think promoted long-run economic growth? Policies to lower interest rates and stimulate private investment? Government provision of infrastructure? Gap closing to eliminate slack between actual and potential output? And if gap closing, how can you tell if the gap has been closed?

(6) I think that the triangle metaphor is not an effective one. Where did Keyserling use it?

(7) The general lesson from the Nixon price control experience is that peacetime price controls don't work: they create the inflationary pressures that they are supposed to remedy. What did Keyserling think of the Nixon price control experience?

================================================================================

Charles Holt and Suan Laury, "Voluntary Provision of a Public Good"

Brad DeLong


I think that this article works well. I was impressed that you were able to complete fifteen rounds in less than twenty minutes. How many students were in the class? 

I was confused at one point. You say that the black cards are provided so that contributions can be kept secret. But you also talk about an agreement that "everyone would contribute until someone did not." Were contributions public in this round?

================================================================================

Waverman and Sirel, "European Telecommunications"

Brad DeLong


A few comments and questions:

(0) The paper needs an introductory summary to tie it together.

(1) What proportions of government revenues have historically been provided by European telecom company profits?

(2) Why did the EC decide that telecommunications competition was important? Was it driven to do so because of the WTO? Was there a separate ideological victory of pro-competition tendencies?

(3) Why are the authors reluctant to add up the revenues in France and Germany? Even given that most money is earned in non-competing markets and thus that the Europe-wide "share" of the total ECU 112 billion is not a "market share" in the industrial organization sense, it is still interesting to note the relative and absolute sizes of the different companies in the industry--even though should competition emerge, these sizes would probably change radically and rapidly.

(4) I find it curious that British Telecom still has such a large share of the market in Britain. I find it somewhat bizarre that British regulators criticize the U.S. for mandating local access. The claim that "U.S. requirements for competition at the local level... ensure that alternative facilities will not be built" seems to me to be precisely the point: we do not need to have two (or three, or five) sets of wires running into each building. Why do the British think differently?

(5) Why should potential competitors risk their money either in the "privatization but not immediate competition" or in the "some privatization..." markets? In both cases the regulators would seem to still have a strong incentive to ensure the victory of the national champion. And is there enough of a fringe market to supply profits to competitors?

(6) My draft ends at the beginning of the "strategic alliances" section...


================================================================================

Harris and Kraft, "Telecommunications Act"



Brad DeLong


A few comments and questions:


(1) How high up into the executive branch did serious substantive discussions of the telecommunications act extend? I can readily believe that, say, Joe Stiglitz at the CEA had strong, substantively-informed views. I have a more difficult time believing that others at the top of the executive or the legislature had much of an agenda besides pleasing important contributors and being associated with a successful bill.

(2) Talk a little bit more about what the MFJ was. It won't be familiar to most of our readers. Similarly, most of our readers will not understand terms like "interLATA"; they may or may not understand what RBOCs are. 

(3) The summary of the history seemed to me to be quite smooth. About halfway through page 4, however, the article shifts into another mode: a brief review of a few recent important competitive developments. One way to handle this would be to make it another section, to signal readers that the tone of the paper is changing. As things are now, I was surprised to run across this list of developments, and uncertain what I was supposed to take away from it.

(4) Surely the MFJ was an act of "legislation" as important as the 1996 act or the 1934 act?

(5) Section II was good. There is a lot of information in it. Section II is also hard to crack: I think we will lose a lot of our potential readers somewhere in the middle of it, on one of the "The FCC also..." phrases.

I am not sure how to fix it: how to keep the information density of section II high, and also to make it more accessible. But I think that some work is needed...

(6) "In the long run it does not matter what the regulators do..." Surely regulators determine when the long run arrives? And to start section III with that promise does not fit with the statement, half a page later, that "The Act served to reinfore many of these problems and create new ones by attempting to impose competition via extensive regulation on a system where existing retail rates were drastically out of balance without first correcting the imbalance." Which is it? Did the Act create problems, or is the Act irrelevant?

(7) The conclusion of the paper seems to be that the Act failed. If it was supposed to set out an institutional framework of property and access rights to the network so that competition can now flourish, it wound up regulating much too much. This point needs to be moved up to the front of the paper. And the end of the paper should address the question of what is to be done: given that states are going to keep POTS cheap, what can be done to improve the economic efficiency of the telecommunications industry?

================================================================================

Pablo Spiller and Carlo Cardilli, "The Frontier of Telecommunications Deregulation"


Brad DeLong


A few comments and questions:

(1) What is the "Group of Five"?

(2) Readers will ask: "How can telecommunications not be a natural monopoly? If the largest network refuses to allow others to connect to it (or charges an arm and a leg for connection services), then people will migrate to the largest network." To say that "increased modularity" is now the most important factor is, some will say, to presume that the government is already providing all competitors with interconnection rights at reasonable prices. Do you agree? If not, why not? If so, then what is the difference between your view and the belief that "competition needs to be engineered by the regulators"?

(3) Why doesn't the British system count as equal access? How is having the default carrier for BT customers be BT different from having the default carrier for MCI subscribers being MCI?

(4) Will the components that must be unbundled be the same in a decade as they are now? Or will networks be sufficiently different that the unbundling requirements will have changed as well?

(5) Section II--what the countries have done--seemed to me to work very well.

(6) Section III--relative performance--seemed to me to suffer from the lack of an overview at its start. You have to read to the end before you can start putting the whole picture of relative performance together.

(7) I take it that the bottom line is that Chile and New Zealand made mistakes by only vaguely specifying their interconnection rules, and that the others did better. If so, move this up to the front: that competition flourishes as long as interconnection rules are well-specified at the beginning.

(8) Do we want local loop competition? It's expensive to run a lot of extra wires to houses. And it is unclear what else you can use the wires for, while data can always be found to flow down trunklines and through switches.

(8)