User:Mary Ann Magadia/sandbox/Maan

Jesus Felipe new article content ...

Jesus Felipe (born 22 June 1962, Madrid, Spain) is Advisor to the Chief Economist in the Economic Research and Regional Cooperation Department of the Asian Development Bank (ADB)), Manila, Philippines. He is the Managing Editor of the Asian Development Review.

Career and Research
Jesus Felipe has been with ADB since 1996; before becoming Advisor to the Chief Economist, he held different positions in the Economics and Research Department and in the Central and West Asia Department. He also held academic positions at the Georgia Institute of Technology and the Hong Kong University of Science and Technology. He holds a Ph.D. from the University of Pennsylvania.

Jesus Felipe has had a long and successful career with the ADB, where he has worked on assignments in countries as diverse as the [|Philippines], Pakistan, China, and [|Kazakhstan]. He has traveled extensively and advised governments on how to prepare medium and long-term development plans. He has vast knowledge of the region’s key [|emerging economies], and a deep understanding of the [|economic policies], business behavior and politics behind why some countries move up in the development ladder while others do not.

His research interests spread across areas such as long-run growth in Asia, the dynamics of structural transformation (see Key Indicators for Asia and the Pacific 2013 Special Theme Chapter on Asia’s Economic Transformation: Where to, How and How Fast?), industrial policy (see book on Development and Modern Industrial Policy in Practice, Issues and Country Experiences), inclusive growth and full employment (see book on Inclusive Growth, Full Employment, and Structural Change, Implications and Policies for Developing Asia), the impact of technology on employment, productivity, technological progress, the functional distribution of income, business cycles, the path of profit rates, and Asia’s potential growth (see Asian Development Outlook 2016, Asia’s Potential Growth). He has published extensively in academic journals on different aspects of Asia’s development, productivity and structural transformation; as well as books on labor markets, structural transformation and industrial policy, several of which have been highly cited. Critique of the aggregate production function Jesus Felipe has spent over two decades studying the empirical foundations of the neoclassical aggregate production function (APF). His analysis is well recognized to have undermined the credibility of econometric estimates of aggregate production functions (0r variants of it), and of total factor productivity (TFP) calculations. https://en.wikipedia.org/wiki/Production_function https://en.wikipedia.org/wiki/Total_factor_productivity

Background to this work By the 1970s, it was very clear that the concept of an APF was very problematic, to say the least: (i) the Cambridge Capital Debates had basically ended. Samuelson (1966, QJE) had clearly indicated that Joan Robinson and her colleagues were right: except in strange contexts (one commodity world), the notion of an APF (including Samuelson's 1962 own attempt to rescue it), was 'wrong'; (ii) Franklin Fisher's work on the conditions to aggregate micro production functions into an aggregate production function (work that had begun in the 1940s with Klein, Leontief, May, Nataf) had also delivered very clear conclusions: the conditions to aggregate micro productions functions into an aggregate production function (with neoclassical properties) are so stringent that one cannot believe that actual economy satisfy them. For practical purposes, Fisher argued, what all this means is that the representation Y=A*F(K,L) of a sector or an economy is simply NOT TRUE, not even as an approximation (as Fisher proved in 1968). On this, see the survey on aggregation by Felipe and Fisher (2003) in Metroeconomica. https://en.wikipedia.org/wiki/Cambridge_capital_controversy Samuelson, 1966: https://www.jstor.org/stable/pdf/1882916.pdf Samuelson, 1962 http://theme.univ-paris1.fr/M1/hpe/HPEM1-TD7.pdf https://en.wikipedia.org/wiki/Joan_Robinson https://en.wikipedia.org/wiki/Franklin_M._Fisher Source: http://economics.mit.edu/files/1382 (ii) Felipe, Jesus; Fisher, Franklin (2003). “Aggregation in Production Functions: What Applied Economists Should Know.” Metroeconomic 54:2 & 3: 208-262. http://riscd2.eco.ub.es/~josepgon/documents/Felipe_Fisher.pdf Despite the above, the profession remained oblivious to all this. Stiglitz and Ferguson, among others, argued that while all the above was fine, what mattered was the empirics, that is, that if one estimates an aggregate production function and it 'works', then why should one throw it? Forget about the aggregation problems! (source for this para) The accounting identity critique Interestingly, the Cobb-Douglas (1928) paper, the first econometric estimation of an aggregate production function, had a very poor initial reception within the profession. It was severely criticized and questioned. It was not until Solow's (1957) work that this literature and research program, i.e., estimation of production functions, and TFP estimates, etc., took off. https://en.wikipedia.org/wiki/Cobb%E2%80%93Douglas_production_function Solow 1957 http://faculty.georgetown.edu/mh5/class/econ489/Solow-Growth-Accounting.pdf However, curiously, the same year Solow published his seminal paper, Phelps-Brown (1957, QJE) published a paper showing that all aggregate production functions Y=A*F(K,L) do, is to trace the accounting identity according to which value added (Y) ≡ wage bill (W) + profits (P) (GDP from the income side in the national accounts), or Y ≡ W+P = wL+rK (where w is the average wage rate, r is the average profit rate, L is employment and K is the stock of capital). One needs a bit of patience to understand Phelps-Brown paper (language of the 1950s), but the story is clear. Phelps Brown http://qje.oxfordjournals.org/content/71/4/546.abstract In 1963 (Review of Economic Studies), Hebert Simon and Ferdinand Levy used a similar argument (in retrospective it is the same one as that of Phelps-Brown, but presented in a different way) to show the isomorphism between the identity Y≡wL+rK and the Cobb-Douglas form. Simon (1979, Scandinavian Journal of Economics) showed the argument clearly and extended it to the constant elasticity substitution (CES). He thought the issue was so important that he dedicated a paragraph to it in his Nobel acceptance lecture. Samuelson (1979, JPE) also used the same argument to question the Cobb-Douglas (1928) estimates: all aggregate production functions do (in that case the Cobb-Douglas form) is to reproduce the income accounting identity. Hence no deep meaning (none at all) should be attached to econometric estimates. Simon and Levy http://digitalcollections.library.cmu.edu/awweb/awarchive?type=file&item=33638 https://en.wikipedia.org/wiki/Isomorphism https://en.wikipedia.org/wiki/Constant_elasticity_of_substitution Simon 1979 http://www.journals.uchicago.edu/doi/pdfplus/10.1086/260806 Cobb-Douglas 1928 http://www2.econ.iastate.edu/classes/econ521/orazem/Papers/cobb-douglas.pdf Shaikh (1974, 1980) understood all this and 'trashed' Solow's (1957) exercise. Solow (1974) did not like it and replied. Shaikh tried to counter Solow’s reply but the editor of the Review of Economics and Statistics decided not to print Shaikh's counter to Solow. It is worth reading Shaikh's 1980 full reply (chapter in a book). Shaikh 1974 http://www.anwarshaikhecon.org/sortable/images/docs/publications/aggregate_production_functions/1974/1-humbug.pdf Shaikh 1980 http://www.anwarshaikhecon.org/sortable/images/docs/publications/aggregate_production_functions/1980/1-humbug2.pdf Solow 1957 http://faculty.georgetown.edu/mh5/class/econ489/Solow-Growth-Accounting.pdf

Felipe and McCombie’s work and contributions to the accounting identity critique Jesus Felipe and John McCombie (University of Cambridge) teamed up in the mid-1990s, and since then have shown in a series of papers that the neoclassical research program based on the notion of an aggregate production function is empty: one can rewrite the expression Y≡wL+rK into a form like Y=A*F(K,L), and specifically a Cobb-Douglas, CES, etc. John http://www.landecon.cam.ac.uk/directory/professor-john-mccombie Have you ever asked yourself how it is possible that a simple Cobb-Douglas form (or even the slightly more complicated CES or translog forms) can explain the complicated production relationships of an economy? Do we really believe that these simple forms can explain over 90% of the variations in GDP of an economy (with millions of firms doing millions of different things with supposedly different production functions, capital stocks, types of labor, etc.), as many estimates have shown? In 1997, Greg Mankiw, for example, wrote that his faith in Solow's model derives from the high R2 obtained when he estimates it. I think one has to be really naive to believe this. The 'only' explanation is that all these forms do is to mimic an accounting identity. https://en.wikipedia.org/wiki/Gross_domestic_product Cobb-Douglas and translog https://en.wikipedia.org/wiki/Cobb%E2%80%93Douglas_production_function Together, Felipe and McCombie have shown in dozens of papers that what Cobb-Douglas (1928), Solow (1956, 1957), Mankiw-Romer-Weil (1992), Hall-Caballero-Lyons (estimation of mark-ups and returns to scale), millions of TFP exercises and discussions of whether a theory of TFP is needed or not (Prescott, Hall, Jones), convergence regressions, labor demand functions (the negative relationship between wage rate and employment) did, are particular cases of the identity Y≡wL+rK. All many authors have done (estimate regressions, calculate TFP, etc.) starting from Y=A*F(K,L), can be done by starting from the identity Y≡wL+rK. But keep in mind that this expression is an 'accounting identity', and hence no conclusions can be drawn. http://jesusfelipe.com/cat1 Mankiw et al 1992 http://eml.berkeley.edu/~dromer/papers/MRW_QJE1992.pdf

The reason behind their argument is that a production function is intrinsically a relationship among physical quantities of output (e.g., cars per worker or per hour; number of tables, etc.) and inputs. At any level of aggregation, however, one has to use value (dollars, euros) to add up the different physical quantities, i.e., 300 million dollars (Y). But once this is done, one is trapped by the fact that the L, K and Y used in econometric estimations are linked through the identity above, i.e., they are the same variables). Touché. This problem does not occur if one uses physical quantities. But virtually all work that has been published uses value data. Felipe and McCombie feel that all the neoclassical growth and productivity research program is 'not even wrong', paraphrasing the physicist Wolfgang Pauli. Their works is compiled in their 2013 book The Aggregate Production Function and Measurement of Technical Change: ‘Not Even Wrong.’        https://en.wikipedia.org/wiki/Not_even_wrong http://www.e-elgar.com/shop/development-and-modern-industrial-policy-in-practice https://en.wikipedia.org/wiki/Wolfgang_Pauli