SYNOPSIS: Mathematical critique of Krugman's writings.

Paul, ‘we’ actually know !

Thomas Cool

Cool@can.nl
Http://www.can.nl/~cool

April 13 1997

Krugman, Phelps, Ormerod and Heilbroner have produced forceful analyses on the current state of the economy, society and economic theory itself, and all with a distinct attention for unemployment. These authors agree on many points, but disagree on major points too. Interestingly, where these authors disagree, my own work offers new answers, on angles clearly not considered by them. My analysis solves conflicts, fills gaps, and complements on useful points. By relating my work to theirs I hope to enable these authors and their readers to plug into - what I consider - a new synthesis for (a renewed) mainstream economics.

Mainstream economics appears to accept high rates of (equilibrium) unemployment as the apparent characteristic of the modern economy. In this view, unemployment is not inefficient, but the unavoidable price to be paid for other desirables. Take for example the case that the United States has low welfare provisions, less unemployment but more poverty and many prisons, while the European Union has high welfare provisions, high unemployment, less poverty and far fewer prisons: these differences then are explained in terms of political choices for example about institutions, labour market flexibility and employability; and it is suggested that such choices are made at the efficiency frontier. Research economists however are more focussed on the question whether current policy is optimal and whether current unemployment is inefficient. The search is for a Pareto improving solution such that some can advance - notably the unemployed and the poor (underemployed) - without costs to the others.

Specifically, Paul Krugman, Edmund Phelps, Paul Ormerod, Robert Heilbroner (with co-author William Milberg) and myself have tried to supplement the mainstream approach. The first four have received a lot of attention, but did not succeed in finding a Pareto improving solution to current unemployment. My analysis has received little attention, though I must confess that I did find such a solution.

It is useful to link my analysis directly to the work of these four authors. My analysis solves conflicts, fills gaps, and complements on useful points. By relating my work to theirs I hope to enable these authors and their readers to plug into - what I consider - a new synthesis for (a renewed) mainstream economics.

In the following I will briefly restate my own analysis, present a general review and then discuss the work of each author separately. I’ll concentrate on the major issues, and then refer to that part of my own work that links to the work of these authors.

My analysis is that a wrong understanding and implementation of tax theory caused governments to accept an inefficient structure of taxes. This caused a structural shift of the Phillipscurve and the NAIRU over 40 years, in a mostly gradual process. The US accepted poverty, and Europe chose a high minimum wage and subsequent unemployment. Since poor people and the minimum wage unemployed cannot compete for jobs, there was ample room for others to demand higher wages. The (threat of) higher inflation caused higher rates of interest to fight it - also reducing investment and subsequent productivity growth.

My analysis includes the wider institutional setting. The institutional framework of economic policy making is a prime cause for the Great Depression and the Great Stagflation. Timely analyses by economists did not get sufficient attention in the policy making process, and this can be analysed as a fault in the Trias Politica.

The original analysis and first outline is given by Cool (1990). Conjunctural aspects are discussed by Cool (1994a). Structural (multiple) equilibrium states including the influence of ‘knowledge’ on social welfare are discussed by Cool (1995a), using mathematical proofs to bypass the vagarities of empirical estimates. Social psychogical forces and processes are discussed by Cool (1996a); and a focus on the constitution is Cool (1996d). Supportive on key insights, e.g. on labour supply, the balance of trade and the concept of a ‘free lunch’, are Cool (1995b & c and 1996b, e & f). Cool (1996c) is an announcement targetted at a general audience.

Dutch readers are in the advantage with books Cool (1992) (part Dutch, part English) and Cool (1994b), and summary review Cool (1995d).

The appendix contains a technical comment about my point about taxes and the Phillipscurve (and its NAIRU).

The four authors and myself have come up with different answers on the causes for and solutions to current unemployment. Table 1 reviews the different positions.

We may also note that most authors do not (explicitly) refer to each other. The reason for this may be practical, in that books that appear in 1995 may have difficulty to refer to Phelps (1994). We may also note that even though the inflation-unemployment relationship is crucial to the analyses of all four and myself, the focusses of analysis differ. Disagreement often leads to neglect rather than to explicit criticism, and it may well be that I have selected top scorers of different citation communities. However, all authors may be justified in neglecting one another. None of the four gives an essential contribution to the understanding of current unemployment. Theoretically their work might be skipped, as I did in practice while developing my analysis.

Table A: Different positions

At a lower level, when we look into details, then there are more points of overlap. An analysis of a practical economic problem (in this case unemployment) of course must have an econometric substratum in order to be taken seriously. Table 2 contains three technical issues, the shift of the Phillipscurve and the influence of technology and globalisation in the model. Here economics would advance if the four authors could convince each other (allow me to add: of my analysis).

It also appears that some of the differences originate from the styles of analysis, which styles also have to do with roots. Ormerod, Phelps and myself have econometric roots, Krugman’s first love was history (see Krugman (1993)), and Heilbroner is clearly a literary economist (‘though’ summa cum laude, Harvard 1940). It is important to identify these different styles.

I like to use econometrics in the way Jan Tinbergen did. It should be technically sound, but not fancy for reasons of its own; it should be relevant for a serious problem, and communicated to the general public in a responsible, modest but still clear manner (even if clarity makes it sound immodest). I also am very much interested in philosophical aspects (what Heilbroner calls the ‘vision thing’), which however is not quite the style of Tinbergen. It appears that the various authors do not share all these qualities in the same degree. Taking these criteria to classify the four authors and myself gives Table 2. The names in the table are in alphabetical order.

Actually, Table 2 summarises the discussion below on style and content.

Table B: Comparing on style and content

The world should be very grateful to Paul Krugman for explaining economic essentials, and not only for these explanations themselves but for his choice of words as well. Krugman’s writing are a display of fact & logic and scientific argument and humour & good will: a quality blend that one hardly ever sees. I can only presume that you have read these books, and then continue my line of reasoning.

My thesis differs from Krugman’s in one major respect. He claims that "we don’t know" about the causes of the productivity slowdown - whereas I claim that ‘we’ do.

The following Krugman quotes are useful - and testify of his intellectual honesty:

1. "I find that almost anything having to do with taxation is better than a sleeping pill". Krugman (1993)
2. "But let me cut to the chase: the real answer is that we don’t know." (1994b, p5, his italics)
3. "The key objective of the supply-side tax reduction was to lower marginal rates, that is, the rates that people pay on any additional income they make. That makes economic sense: marginal rather than average rates determine the incentive to work and invest." (1994b, p155)
4. "I’m not an expert on taxes." (Said in a public exchange following his Tinbergen Lecture 1996, to be published by the Dutch "Koninklijke Vereniging voor Staathuishoudkunde" - Royal Dutch Association for Political Economy)

These points are relevant for understanding:

5. See the appendix and e.g. Cool (199c), for starters, on my analysis on taxes.
6. Krugman (1994a) makes a big issue of productivity.
   Comment: Quite correct.
   Note that I am rather sure about the explanation of and cure for the productivity slowdown, but that my certainty derives from mathematical proof and trained intuition, and not from an econometric model exercise on the (world) economy. First see the appendix below and my papers. Secondly, my analysis does not invalidate what others have said on the shift to the service economy - and the difficulties of measurement - etcetera. Finally, there are relatively new insights to take into account.
   One of the ideas that I would have liked to look into, but have had no time for, is, that the return on consumer investments (like home improvement for the elderly) may be larger than that on financial stock ("savings"), and that this return is not adequately accounted for (also as a tax base).
   Another idea, also emphasised by Phelps, is that real rates of interest are high (anyway). A major cause is that Central Banks have to be tough, given the reduced competition on the labour market. Another cause is that the government doesn’t dare to raise marginal rates given the current misconception about taxes; so governments borrow (at a higher rate) what actually should have been taxes. Subsequently, investors buy government bonds and grow lazy and spoiled about taking risk (that otherwise would have spurred productivity).
7. Krugman (1994b) p186 onward discusses East Germany and its relation to the downfall of the European Monetary System. The story is familiar: the then-existing policy paradigms of the EMS forcing a recession in Europe when Germany raised its interest rates. Krugman suggests that exchange parities should have been adjusted before the markets forced this. He suggested that preoccupation with fixed rates seduced policy makers to adopt the Maastricht Treaty on the EMU: "(...) by early 1993 political and economic stresses had made the solemnity of Maastricht seem almost comic. If there is a lesson here, it is that serious and dignified men and women in impressive international meetings may have absolutely no idea what they are talking about." (p192).
   Comment:
   This is too quick. When Germany decided that wage earnings in the East should be equal to those of the West (to reduce migration), it should also have decided to let wage costs reflect productivity. This is a better approach than parity adjustment; and known at the time, see my work and the Financial Times editorial "Time for Mr Kohl to act", July 26 1991.
   In the same way, EMU can still aspire at monetary stability, and this can be done when countries use their tax structures (thus, structure as opposed to level only) to balance wage costs with productivity. Even though EMU is not a logical beauty, and East Germany still suffers from a wrong policy mix, the gut feeling of EMU - one economy, one means of payment - was admirably correct. This is even clearer given my work on taxes and their influence on wage costs.
   Note that many top economists make fun of EMU instead of providing answers of how to deal with the policy challenge. This is not so professional.
   One possible answer is the following. With one rate of interest for the EMU territory, and rates of inflation differing by regions (countries), real rates will tend to differ. Some markets will be interested in the real rate instead of the nominal rate. So loans indexed to the local inflation rate might suit many, for example Dutch government and Dutch pension funds, for part of the portfolio.

The following points are only interesting:

8. Krugman makes a point that income developments are fractal. Laywers get much more than cleaners, but top lawyers get much more than average lawyers.
   Comment: Ditch ‘fractal’. It still is a lognormal distribution.
9. Krugman (1994a & b, 1996a) suggests that international influences are less important, due to the size of proportions, than commonly thought. Yet, he himself (1996b) comes with the ‘parable of clocks’: international fluctuations may get into phase, similarly like clocks.
   Comment:
   So, though fluctuations may only be the cream on top of fundamentals, there still is a new research topic.
   Note too that the Great Depression and the Great Stagflation were OECD phenomena and more than ‘cream on the top’; these may be traced to the Trias Politica, see Cool (1996d).
10. Krugman (1993): "I had some trouble getting that paper published - receiving the dismissive rejection by a flagship journal (the QJE) that seems to be the fate of every innovation in economics".
    Comment:
    Cool (1995a) contains an appendix showing the silly arguments that the EER used for rejection of the body of that paper; and Cool (1996a) extends in general. People in responsible position have the awkward tendency to start criticising before asking questions. They fail to see that their criticisms can be formulated as questions - which then are a reason for publication. And they are insulated against protest to this injustice. I recently came upon some beautiful comments by Bellman (1968) on the evolution of scientific ideas. Note, though, that Krugman’s wonderful books since 1990 have only been made possible since my analysis has been blocked from general attention: so that is a form of comfort.
11. Note: With respect to Table 2, I’ve hesitated about classifying Krugman as having less roots in econometrics. His credentials as a technical economist are quite adequate. But, my experience with econometric modelling has been extensive and will not easily be copied. Also, I don’t particularly like the topic of taxation myself either, but it only by going through the details of a complete model (too) that I came upon that explanation. Though, Paul may make me regret this classification.

 

Phelps

 

Phelps (1994) is as creative as the others, but also the technically most advanced author who also presents econometric tests for some of his conjectures. His book is impressive.

My first reaction in 1994 to Phelps’s book was guided by his explanations in plain English. Given those explanations, his study dropped in my priority list. My attitude is (in line with Tinbergen and Keynes) that substance comes before technique. So it may come as a surprise to the reader that as an econometrician I did not jump to the occasion to comment on Phelps’s techniques and tests. But of course, had I had more time, I would have studied those pages too. And of course it is still appreciated that Phelps has produced these technical pages. They have affected his style, and they allow for wider tests at a later stage. Indeed, for the purposes of this paper, I have looked into the estimation sections more deeply. My comments below however remain preliminary, since, indeed, I have not fully read all chapters.

The major comments are:

12. Phelps (p374-375) is sceptical about how politicians abuse economics, and about how economists themselves react to (new) ideas.
    Comment:
    Talk to Krugman, and study my analysis on the Trias Politica.
13. Phelps: "There is already a moral-philosophical case for employment subsidies targeted at the low end of the wage scale to bring the rewards for work not having a high scarcity value more nearly in line with the requirements of econmic justice." (p366) and he seems to approve of proposals also made by Dennis Snower.
    Comment:
    I even show that these measures cost nothing and are Pareto improving.
    Do you agree that there may be an ‘equilibrium’ in your sense, but inoptimal ? (See below.)
14. Chapter 18 contains a ‘concise postwar economic history’.
    Comment:
    The reader is invited to compare that history with my amendment to the Bruno & Sachs story, see Cool (1996c).
15. Phelps catalogues monetary aspects as temporary (‘high frequency’) and nonmonetary aspects as structural (see p4 and 335).
    Comment:
    I agree that it is valuable to look at nonmonetary effects. But the major issue is the Phillipscurve, a relation between unemployment and inflation, and thus it is difficult to neglect monetary policy. When Central Banks have a wrong theory, and cause the rate of interest to rise, then this should be in the model.
    On page 314, the acceleration of prices (change of inflation) is introduced in a Phillipscurve in an ad hoc manner.
    Similarly, on page 329 the possible influence of Bretton Woods is discussed, and Phelps remarks that this system allowed for adjustable pegs - but then misses the point that the pegs were pretty fixed in practice.
    No doubt, Phelps will agree that the whole story contains both elements.
16. Phelps uses the calculus of variations, and his marginal tax rate is ¶ T(y)/¶ y.
    Comment:
    This is proper in this theoretical development, but it should be replaced by a dynamic marginal rate when the theory is translated to the real world. In the appendix below it is explained what I mean by this, and it is shown that this dynamic marginal rate may be close to the average rate.
    Curiously, Phelps’s econometric exercise uses average rates (p 314 & 318), and finds a contractionary relationship. In a sense, this supports my analysis, which allows lower average taxes and thus lower unemployment. However, I think that the estimated equation is too simple for the true model.
17. Turnover costs appear to be very effective in one of the major models.
    Comment:
    That would mean that a simple subsidy would have huge effects. This does not seem realistic. The huge effect comes from the homogeneous labour assumption, and it is more appropriate to assume heterogeneous labour, see Cool (1994a and 1995c).
18. "The shifts and long swings in unemployment are an equilibrium phenomenon, not a matter of misperceptions or misforecasts and consequent wage-price misalignments" (p vii). Phelps then uses "(...) the equilibrium case in the expectational sense of the term: the case of correct expectations about the course of the economy." (p1)
    Comment:
    The Moon falling on and past the Earth - and expecting to fall so - is a story of disequilibrium and of equilibrating forces but also of equilibrium. What you use is just a matter of perception and of words. More important is the inoptimality of present unemployment.
    Phelps writes on optimality: "(...) much of what we measure as unemployment reflects job rationing, hence is involutary and imposes private and social net burdens (...)" (p viii, see also Phelps p9).
    Thus note that there is another concept of the "natural rate" (NAIRU), namely the market clearing rate.
    Even when expectations are correct - even when happens what you predict - then you can still be unhappy about that and look for change; and thus there can still be forces towards the clearing rate. Fulfillment of expectations is not the only utility that you are after. Phelps’s emphasis on the expectations definition suggests that his analysis is incomplete.
    Inoptimality may also have causes in the political structure, a point that gets less attention by Phelps regardless of his comment on p374-375.
19. Phelps: "A worldwide increase of public expenditure (...) was not found to be expansionary (...) The same is true of a worldwide increase of public debt. (...) Prudence requires putting aside the Keynesian approach for the time being in favor of taking up the structuralist approach." (p330)
    However, the page before: "(...) the economy is so complex an organism, so to speak, that it would be naive in the extreme to imagine that, at long last, the true macroeconomic model of equilibrium unemployment determination had been discovered. A question that permanently looms over any such research as this is whether the results interpreted as favorable to the theory are in reality the expression of some mix of other theories, some likely to be old and some not yet known." (p329)
    Comment:
    I fully agree with the statement on page 329 but think the statement on page 330 overdone. The body of neoclassical thought is too big and strong to be replaced by a mostly ad hoc econometric exercise. This is hubris !
    For starters: government expenditures rose as a result of unemployment benefit payments. So there is a positive relation between unemployment and expenditure. Secondly, "Keynes" is much more complex than the simple idea that deficits would reduce unemployment. Macro-economics aspires at wise management of economic development, only occasionally using deficits to reduce unemployment. (What politicians do, is another story.) One needs a more complex structural model to disentangle the various relationships, instead of a two-equation reduced form estimate as Phelps does.

Less important comments are:

20. "The natural rate moves!" (p vii)
    Comment:
    The book suffers from the emphasis on the novelty of this idea. However, the nonconstancy is part of its definition, and this was not so revolutionary, in 1994. For example, it was a common notion to me in 1989/90 when I generated my analysis, and Phelps (p xii) mentions a 1979 paper by Jeffrey Sachs. But note that the book reflects a 20 year research project, e.g. Phelps discusses on page ix early models of the early 1980s that assumed a constant NAIRU. So it may well be that some researchers settled for constancy, and that it was a struggle for Phelps to get rid of constancy; and we should be tolerant of struggles like this. But, objectively, the emphasis on a non-novel idea is out of touch with modesty.
21. The opposition of "structuralism" to "neoclassical" (p14-19) is rather constructed, and not modest again.
22. "(...) historical evidence that unemployment is (or was) trendless (...)" (p x)
    Comment:
    Agreed.
    Note, though, that my analysis is that due to differential indexaton of taxes and subsistence, there is a trend in a component of unemployment (namely, minimum wage unemployment, and poverty (underemployment)).
23. On technology: "the theory averts any implication that secular productivity growth puts the equilibrium unemployment rate on a trend (...)" (p xi)
    Comment: Talk to Heilbroner.
24. "(...) the present study is the most comprehensive econometric model of unemployment to date" (p 313).
    Comment:
    Well, there is Lawrence Klein’s Project Link, there is .... etcetera.
25. Phelps (p352) relates to Jude Wanniski, an ‘amateur fiscal theorist’ who wrote ‘an interesting book’, and dismisses him as a serious thinker. On p353 Phelps speaks about ‘professional theorists in the supply-side movement’ without mentioning names.
    Comment: See Krugman (1994b).

 

Ormerod

 

The book’s name "The death of economics" is not inviting to serious research. One may appeal to a "The King is death. Long live the King !" approach, and indeed Ormerod’s last chapter "Economics Revisited" seems to suggest this. But this is so round-about and distractive ! Why first make people believe that you want to get rid of economics, and then tell them that you have a better economic analysis ?

This way of presentation also gives too much credit to decisions makers. Politicians and economic advisers who believed in those theories are presented as misguided persons, and victims of failing theories of old. Just as anybody can make errors. However, the proper story is that illusions and ideological views have been maintained in the face of contradictory evidence, and against the advice of renowned economists. Ormerod’s presentation obscures this evidence and its meaning. The proper story, that Ormerod misses, poses the question of reform in the structure of economic decision making.

Agreed

I agree with Ormerod: "The whole challenge of economic policy is to shift the attractor points around which the economies move, and hence the whole solution path of the economy over time." (p208)

Disagreed

26. He claims that there is a new analysis of unemployment moving around an "attractor" (that itself can move).
    Comment:
    This attractor is nothing else but the NAIRU. It is true that it can be clarifying to shift from the conventional parlance to the parlance of chaos theory, but it is not revolutionary as claimed. The same immodesty as Phelps. See Cool (1994a).
27. He defends the macro-economics approach, e.g. on using a rather simple relation between inflation and unemployment.
    Comment:
    Defence is fine, but the correct approach still is based upon micro-foundations. See Cool (1994a).
28. Ormerod writes: "The distinguishing feature of chaotic systems is that their behaviour is impossible to predict in the long run (...)"
    Comment:
    The word "chaotic" means "deterministic looking like random" in mathematics. Above quote is only true for (systems of) equations with a random term somewhere.
    "Chaos" has the connotation "random" in the public mind, so it might be best not to use the term in books for the general public.
    Ormerod gives much attention to uncertainty, and the way that he presents it carries with it the suggestion that nothing can be done about unemployment. Though uncertainty is important to macro-economics indeed, it however is not really relevant for his main thesis that something could be done about unemployment.
29. He claims that the 1950s were a special period of reconstruction, in the sense that the success of these years is not easily repeated.
    Comment:
    In my analysis, the conditions of economic success can be influenced, and similar results achieved again. The mood of optimism would follow the results, rather than conversily (though there is feedback too, of course). See Cool (1995a) and (1996c).
30. Ormerod: "So what can be done ? One solution to the problem of high European unemployment, for example, is work-sharing." (p207) To achieve this, he appeals to social values.
    Comment:
    But work sharing is not necessary (see my work in general), and less easy to achieve anyway (see Cool (1995c)).
31. Ormerod: "But perhaps the most important point of all, linked though it is to the underlying mathematics, must be stated in words, for it is a question of moral values. The concept, rampant in the free-market philosophy of the 1980s, that there is no such thing as society is one which, if it is allowed to persist, will prevent the creation of full employment regardless of the form which economic policy takes." (p211)
    Comment:
    There is little use in discussing whether there is or is no "society", since it would seem to be a matter of definition. If a government would choose not to solve unemployment, then this should be accepted in a democratic society. It is a different thing that we now can show a solution to inefficient unemployment, since that is a matter of logic and intellectual honesty.

Heilbroner & Milberg (1995) are very wordy and imprecise - and the many words are used for hyperbole instead of exactness. It is very easy to get irritated.

There are only a few points that I agree with, but even these points are formulated vaguely and annoyingly, and my comments are guarded. Also, to reduce the irritation, I only usefully comment mainly on chapters 1 and 7:

1. Heilbroner: "(...) Keynesian theory can be judged a success (... when allowance is made for ...) bargaining power of labor." (p57) and "Stagflation has come to an end with the political and economic events of recent years. The bargaining strength of labor in the advanced industrial countries has been threatened in part by the rise of international competition." (p59)
   Comment:
   Advanced nations are ‘service countries’, and see Krugman on "international competition". Bargaining power is a very important variable, but has more to do with the level of wages than the (inflationary) rate of change. Taxes are neglected. With unemployment and poverty so large, we are only at the low inflation asymptot of the Phillipscurve, and stagflation is not dead yet. Curiously, Heilbroner’s book is motivated by social problems, but the problem is declared dead ! In other words, he doesn’t see that his problems are caused by stagflation.
2. "(...) the extraordinary combination of arrogance and innocence with which mainstream economics has approached the problems of a nation that has experienced twenty years of declining real wages, forty percent of whose children live in "absolute" poverty, and which has endured an unprecedented erosion of health, vacation, and pension benefits. (reference) The commitment to full employment legislated in 1946 has been "honored" in these socially destructive years not by vigorous employment-generating programs such as the reconstruction of its cities, but by redefining "full employment" as a higher level of unemployment." (p6)
   Comment:
   Agreed on the concern, disagreed on the rest. Do not mix up politics with economics. See Krugman’s description of how policy fashions drifted from economics proper. Also, there were serious questions regarding the causes of unemployment, and these questions cannot be played down so so easily and derogatory.
3. "It is the legitimacy of the public sector within capitalism that lies at the core of the contemporary crisis of vision." (p120)
   Comment:
   He is too vague on this, so he might as well be wrong. But agreed in principle, see my advice to adapt the Trias Politica, Cool (1996d).
   In general, Heilbroner doesn’t clearly distinguish between economists as scientists (who have all the time of the world to doubt) and economists as policy advisers (who also have to take into account that decisions have to be made here and now).
4. "(...) the mark of modern-day economics is its extraordinary indifference (to the connection between theory and reality /TC) At its peaks, the "high theorizing" of the present period attains a degree of unreality that can be matched only by medieval scholasticism." (p3-4)
   Comment:
   Yeah, for "peaks": that may be. It is good we have those peaks.
   
   "Analysis has thus become the jewel in the crown of economics. To this we have no objection. The problem is that analysis has gradually become the crown itself (...)"
   Comment: well, that is an overstatement. Is the suggestion that all economics now is a "peak" ? Besides, did you really look at the practical work at the relevant institutes ?
   Heilbroner misses the point that my analysis is fine work in the mathematical tradition, and that it is neglected by many (by him too). Rather than downgrading all math, he should highlight the work that matters, and state the reasons why it matters.
5. Heilbroner sees the following causes for unemployment:
1. "On the domestic front, they include a technology of rampant automation that has created severe employment strains in all advanced countries (...) The result is prospective increasing dependency on government-financed programs of unemployment relief or public works." (p120-121)
2. "Meanwhile, on the international front, (...) "globalization" of production carries unsettling implications for all advanced capitalisms, including the lowering of social, environmental, and labor standards (...)" (p121)
3. Other issues are volatility of financial flows, demography and immigration, ecology and nationalism & terrorism.
   
   Comment: This is bad economics. See Krugman & my work.

Heilbroner’s book is recommended on the back-flap by Lester Thurow as "essential reading". He and his readers are advised to read Krugman on Thurow.

There is a final caveat. With my European background it is easier for me to see the value of government involvement, cost-benefit analysis and policy analysis. I am not familiar with the American academic situation, and it may be that Heilbroner really has a case that these aspects are underappreciated in the US.

All authors advice their colleagues, policy advisors and politicians. All however accept the current institutional setting of economic policy making, and accept that their thoughts get less unbiased attention than could be useful. Cool (1990, 1994b, 1996d) however advices a constitutional amendment for an Economic Supreme Court, and explains that the lack of sufficient checks and balances is a major cause for the tragic economic record of the last century. When experts know of Pareto improving possibilities, then policymakers have too much freedom to neglect this. Policymakers have too much freedom to pursue their own pet theories even in the face of contradictory evidence.

Technically, the tax issue goes as follows. Following the hypothesis of optimisation, marginal tax rates are important. Econometric research however tends to show that average tax rates tend to be most important. At best a paradox. The following is an explanation.

With T(y) the tax associated with income y, the marginal rate commonly is computed as ¶ T(y)/¶ y. However, translating theory to practical economics of dynamic developments, requires at least a tax function T(y, a), with a the tax parameters. The proper marginal rate would be (take the total differential, divide by dy ):

dT(y, a) ¶ T(y, a) ¶ T(y, a)
(1) ----------- = ------------ + ------------ da / dy
dy ¶ y ¶ a

With economic growth from one year to the other, tax rates are normally adjusted. Tax adjustment cannot simply be neglected in policy simulations. This may be best seen by looking at the "dynamic" marginal rate (DMR) over discrete points in time:

(2) DMR(y) = (T(y, a) - T(y_-1 , a_-1 )) / (y - y_-1 ) = D T / D y

Balanced growth can be defined, for our purposes, such that tax rates are adjusted fully with income such that the dynamic marginal rate equals the average rate:

(3) D T / D y = T_-1 / y_-1 Û D T / T_-1 = D y / y_-1

If growth rates are equal in each year, the DMR and the average rate are constant over time.

These relations also hold for individuals. Thus the "marginal rate" is less important, and the "average rate" (dynamic marginal rate) is more important, as already shown by practical research.

It is of interest, of course, what happens when the economy (and individuals) deviate from the balanced growth path. For example, taxes are indexed on inflation instead of income. Based upon a wrong analysis of incentives (thus, marginal instead of average rates), OECD policy has been to adjust tax exemption only for the cost of living and not for the general rise of productivity, income and well-being. This caused the Phillipscurve to shift. See Cool (1994a, 1995a & b and 1996b & e) for more technical developments and references to other authors.

The point here is, though, that above properties have been neglected in common textbooks, research and policy advice. Taking account of these properties and including them in current mainstream economic models, explains the Great Stagflation.

Ewp references are to the Economics Working Papers Archive at the Washington University at St. Louis: http://econwpa.wustl.edu. See also http://www.can.nl/~cool.