Economics and Reality (Economics As Social Theory)

Rating:
Summary: Keynes was not a critical realist,but a logical empiricist
Review: There are two claims made ,either implicitly or explicitly,by critical realists(Lawson,Hodgson,Runde,Fleetwood,etc.)in general, and by Lawson specifically in this book,that are simply incorrect.The first claim is that J M Keynes was a critical realist.Keynes was an early logical empiricist upon whom R.Carnap based his formal Logical Empiricism.All of the major points made by Keynes in the A Treatise on Probability(1921;TP)-the importance of intuition,a priori reasoning,Kant's synthetic a priori,the logical theory of probability,the weight of the evidence(argument),the equal importance of experience and reason in making decisions, the principle of indifference ,the special nature of frequency theories of probability versus the generality of the logical approach to probability,were all upheld in Carnap's 1969 defense of his philosophy.Carnap got one thing,and only one thing, wrong.He,like so many others,fell for the canards spread by F Ramsey ,in his 1922 and 1926 book reviews of Keynes's TP ,that Keynesian probabilities were "nonnumerical" and "nonmeasurable"(Lawson is a prime example of the "so many others").He failed to recognize that Keynes's basic approach to probability estimation was interval valued.In general,probabilities are indeterminate or imprecise or indefinite intervals which have a lower and an upper bound.Nowhere in his book ,or in any other book written by any other critical realist or economist, is the above connection noted or even alluded to in a footnote.This ignorance about the basic nature of Keynes's philosophy calls into question the accuracy and reliability of practically every statement made by Lawson about Keynes ,not only in this book,but in the entire corpus of Lawson's published writings.The second error of both Lawson and his fellow critical realists is the explicit and/or implicit claim that Keynes did not provide a general theory of employment in his 1936 work,The General Theory of Employment,Interest and Money(GT).This claim is the result of the pervasive mathematical illiteracy,ineptness and innumeracy prevalent among critical realists(institutionalists,post keynesians,cambridge keynesians).The following summary of the technical aspects of Keynes's GT can be duplicated by any reader,economist or critical realist included, of this review who knows how to integrate(take the antiderivative of)derivatives.First,go to the pages of the GT where Keynes presents derivatives.These pages are pp.55-56,ft.2,pp.280-286(chapter 20),pp.304-306(chapter 21),pp.271-278(chapter 19),and pp.114-117,126(chapter 10;Keynes's definition that Y=PO on p.209 refers back to chapter 10 and the Y-multiplier model.The other pages listed above all refer to Keynes's D-Z model).Keynes's expected aggregate supply function Z(Z1+Z2)=g(N)=P +wN,where P equals expected profit,w equals the money wage,and N equals aggregate employment.Keynes's expected aggregate demand function D(D1+D2)=f(N)=pO,where p is the expected price level of aggregate output and O,real output,is a function of N.Keynes has just specified an aggregate production function(p.283,GT).The actual or current aggregate demand function is Y(=C+I)=PO=bY+(1-b)Y,where b is the marginal propensity to spend on consumption goods(mpc), 1-b is the marginal propensity to spend on investment goods(mpi),and Y is a function of N.P,in the context of the Y-multiplier model,is equal to the actual price level of aggregate output.Setting D=Z,one obtains the aggregate supply CURVE,a locus of all possible expected prices and expected profits. Assuming that Say's Law holds,set the elasticity at the bottom of p.116 of the GT equal to the elasticity at the bottom of p.283 of the GT.The following result is obtained:w/p=mpl/(mpc+mpi),where w/p equals the expected real wage and mpl equals the marginal product of labor.If,and only if,mpc+mpi=1 will the classical and neoclassical labor market clearing equilibrium condition ,needed to obtain an optimal result,occur.There will be no involuntary unemployment,only voluntary and frictional unemployment.If mpc+mpi is < 1,involuntary unemployment will occur since as (mpc+mpi) falls,the ratio on the RHS will rise.To maintain the equality requires that the money wage rises or that the money wage rises by more than the price level.The LHS ratio must rise.This simply means that labor as a whole can do nothing in this situation to LOWER the money wage ,since this would violate the necessary first order condition for optimality THAT THE W/P=MPL.Obviously,government(public)goods,export goods,and import goods can be divided into consumption and investment goods,in which case Keynes's general optimality condition does not need to be expanded.It can ,of course ,be expanded in the following way.Let mpg,mpe,and mpm equal the marginal propensities to spend on public goods,exports,and imports.Keynes's general theory would then be summarized and read as follows:w/p=mpl/(mpc+mpi+mpg+mpe-mpm).Unless mpc +mpi+mpg+mpe-mpm=1,involuntary unemployment will occur.Finally,the strange objections made by Richard Kahn and Joan and Austin Robinson that Keynes ignored imperfect competition are easily met by multiplying the LHS(w/p) by the elasticity 1-[1/(Es)]and the RHS by the elasticity 1+[1/Ed].Until Lawson and/or his fellow critical realists deal correctly with Keynes's philosophy,approach to probability , and his general theory of employment, important side issues, like the proper role that methodological considerations should play in economics if it is to become a science,the misuse of econometrics(addiction to the normal probability distribution),the relative importance of a particular set of evolving labor ,political,social,financial and economic institutions over time,etc., will offer no challenge to the reigning neoclassical school at all.

Rating:
Summary: graduates and faculty in agricultural economics should read
Review: For me the book is very good. I do not take it as a show-off of erudition but as a thorough explanation of why recognition of subject and context matter if we are to address problems more meaningfully. The authors goes into indicating a 'broad criteria and strategy', as giving a recipe is not consistent with critical realism, the philosophy of social science proposed as appropriate to move on to leaving a traditional lack of sense in much economic research.

Rating:
Summary: A nice work that doesn't communicate
Review: Those readers that are suspicious of the methodological foundations of modern economics will definitely find a lot of interesting material in Lawson's book. There is no doubt that Lawson's work (which is a product of more than 10 years of research at Cambridge) is deep and raises fundamental questions. One big problem, however, is that the book assumes that the reader has a substantial amount of knowledge about philosophy of science and this is surely not the case as far as the "representative" economist is concerned. Too bad. Lawson's message deserves to be made available to a larger audience. This would only be accomplished, however, if Lawson had chosen to cut the amount of philosophical blah-blah-blah and give more direct to the point examples of what he has in mind. He preferred to demonstrate his erudition, though.

Rating:
Summary: I probably could disagree more, but not much
Review: Tony Lawson is a "critical realist" who argues that mainstream economic theory is incorrect, the basic problem being that (a) it follows the "deductivist" method rather than having a seriously empirical, inductive, historical-descriptive encounter with economic reality; and (b) it doesn't take methodology seriously, which it should. Much of the book is involved in developing economic methodology, and using the product of this development to critique deductivist economics and provide a basis for an alternative.

I disagree completely with (b), and only partially with (a). Lawson never argues at all why we should study methodology. He just believes it is obvious. But, when you study economics, you learn of lots of people who have contributed, none of whom was a methodologist, and most of whom never studied or took seriously methodology (except perhaps in their old age, long after their contributions). Natural scientists learn just enough methodology to get started researching, and pick up some more through tacit knowledge (learning from colleagues). But, mostly it's intuitive. Just as you don't need to know linguistics to speak perfect English, you don't have to know methodology to be a great researcher. And obversely.

The important point is: the only successful critique of a theory is a theory that predicts and explains better. Methodological critiques carry no weight at all with scientists, natural or behavioral. I believe that is both correct and as it should be, and the book gives me no reason to think otherwise.

As to the first point, Lawson seriously understates the value of "deductivist" models. From the Physiocratic Tableau, to Cournot's model of duopoly and Ricardo' theory of rent, to models of supply and demand, thence to the Edgeworth Box and ultimately to general equilibrium theory and its Fundamental Theorem, deductive arguments have been incredibly important in explaining why market economies work and why incentives are important. We would not have the foggiest idea of how to assess economic arrangments unless we had studied these models.

Lawson is correct in saying that most of the deductivist theory found in the journals is just drivel, but that is true in natural science too, and even in mathematics. We only know ex post what is important. For instance, the past few decades have given us principal-agent models, which are incredibly important for understanding market economies, and game theory, which is immensely revealing and used in policy-making, structuring of government and private auctions, and industrial regulation, to name a few applications.

Lawson is also especially critical of econometrics and economic forecasting. He is wrong on both counts. First, he only discusses time series macroeconomic forecasting, when most econometrics is used on cross-sectional data, and it works quite well. Even in time series, while the grand hopes of the 1970 to make accure macro forecasting models never materialized (and are no longer pursued by theorists), econometric forecasting is probably one of the most sought-after skills an economist has, and is used widely in business and government.

So, what is left of Lawson's critique? Nothing. I have ordered his new book, to see if it is any better, but I don't have high hopes.

This does not mean that mainstream economic theory is in great shape and cannot be critiqued. Far from it. The point is, however, that each generation of young economists sees the errors in the previous, and the enterprising ones transform research by moving to repair the problems. Criticizing mainstream economics is difficult because it is a moving target--always changing to meet new circumstances.

Lawson does not look at new developments in economics at all. His book was published in 1997, but it deals with the economics of the 1950-75 period (e.g., the author's extended dialog with Frank Hahn, whose work is of this period). No game theory. Very old-fashioned econometrics. No experimental economics. These are the moving forces of contemporary economics! Lawson devotes much space to how economics cannot be experimental, whereas experimentalist are now in the ascendancy in economics (witness the Nobel prize given to Daniel Kahneman and Vernon Smith two years ago).