Economic Growth: Evolution, estimation controversies and suggestions for a new direction.

Ph.D. Project, City University, 2006-2009 Economic Growth: Evolution, estimation controversies and suggestions for a new direction. Benjamin H. Mitra-Kahn* *City University, London The New School for Social Research, New York Benjamin@mitrakahn.com www.mitrakahn.com Economic growth is at the heart of economic analysis, and is the policy objective of most development strategies. However, the current method of accounting for growth, as this thesis will argue, misses the conceptual, theoretical, and therefore empirical mark, in estimating growth. Initially, a historical review of national accounting methodology will establish how concepts like ‘the economy’ and ‘growth’ has been (and is) context specific and fluid. The second section will show that there is an underlying value theory in modern national accounts, which means that estimates of GDP growth are not an expression of output growth. Further, production estimates will be shown to be internally inconsistent, and not equal to income or expenditure, while a historical review investigates how the production, income, expenditure trinity arose. A review of American, British and Indian GDP methods will clearly demonstrate how GDP values from different countries can not be treated as identical variables, useful for cross-comparison. The final section will propose an alternative angle from which to approach economic growth, based on the concepts of access, sustainable expansion and the utilization of output. This is done in an attempt a re-orient the way we approach economic growth, so while I agree that GDP was “one of the great inventions of the 20th century” 1, this thesis will expose its hitherto concealed internal flaws, arguing for an advance into the 21st. Proposed Structure For a first section, I would like to show that; while the study of economic theory and national accounting is separated in the current economics environment, historically this was not the case. Modern economics, post 1946, has separated the job of the national accountant and that of the economist. The composition and calculation of a system of national accounts is seen as predominantly the domain of accountants, while macro -and micro economists should focus on model building. This de-linking of theory and accounting in economics has had two effects: Initially it led to a latent acceptance, and entrenchment, of Keynesian equilibrium conditions, and neo-classical utility in the national accounts, and thereby the majority of models and theories practiced both in the orthodox and heterodox schools of thought2. Particularly due to the need for empirical verification, the national accounts always play a role in epistemological claims of 1 2 Bureau of Economic Analysis (2000) An example of this is the Computable General Equilibrium modeling approach, where the underlying structure of the model comes from the National accounts and Keynesian macro balancing equations, while neo-classical single agent optimization is often used to describe sector and agent behavior, while epistemological claims is through an appeal to general equilibrium analysis (Kahn 2006, forthcoming) 2 models. Second, the de-linking has led economists to assume that the national accounts carry with it only a minimum of theoretical baggage and this has indirectly acted as a barrier to any form of any investigation into the structure or theoretical content of the accounts. This first chapter will provide a historical perspective of the evolution of national accounting and ‘economic growth’ as a concept to be measured quantitatively. This will illustrate how ‘national income’ is a fluid concept which has evolved over time, where each mode of measurement has been context specific3. One should add that such a review of methodological history has not been attempted yet, while literature on the history of national accounting itself is available4. The second section will outline a set of critiques against the current framework and its method of accounting, focusing on the three key areas, which Stone and Meade emphasized, as their key rationales for their organization of the national accounting framework. This was to “provide estimates of the national income in the various forms, which are most useful to the economist”, to facilitate the “international comparison of national income” and to allow “cross checking of various methods of estimating the national income” (Stone & Meade, 1941, p. 216). Taking each issue in turn; by exposing the cost approach to national accounting, and the implicit value theory herein (see Kahn 2005), I will show how the measure of aggregate expenditure is unable to distinguish between changes in output and changes in the cost and firm structure in an economy, meaning identical production in two countries can result in different GDP estimates depending on cost structures and changes in the number of registered firms. This indicates that GDP is an internally inconsistent variable, when it tries to act as a proxy for economic growth. Secondly, the method for estimating GDP in different countries varies, in the sense that certain variables and categories are omitted by some countries5. Add to this, the complication that the national accounts are based on domestic registered firms’ accounts only, and that accounting 3 Examples include Quesnay’s (1758) circular flow of income for France where the agricultural sector was the only surplus producer, as France’s agriculture sector was superior to her competitors, the UK and The Netherlands. These latter countries followed King (1696), in counting stocks of goods and income, with a later emphasis on trade with the colonies, due to their superiority on this side of the economy. 4 Studenski (1956) wrote a defining work on the history of national accounts, but leaves out methodology, and any possible raison d'être of the individual authors, for their varied approaches. 5 The USA, for example, has no government capital account, meaning it has no government depreciation deducted from GDP, as opposed to the UK which does. As such any direct comparison of the growth estimates is inappropriate at best. 3 standards vary (substantially6) from country to country. Means cross country comparison of GDP, are in reality the comparisons of a wide range of variables, common only in name. A critical review of the growing literature on efforts towards mending and/or extending the accounts will be undertaken, with a focus on the work of Summer, Heston7 and Godley. This is to illustrate that while these developments are conceptually important and interesting, they carry the same underlying flaws as the aggregate accounts. Thirdly, GDP accounting uses production rather than consumption as a point of departure, and later applied a trinity of equality between income, production and expenditure. It will be argued that this perspective derives from Mill8 and his contemporaries, and is transformed into a trinity over the next 90 years. Whilst the theory argues for individual accounts, in statistical reality the macro variables are not ‘cross checked’, but are derived from the production (cost) account using macro balancing conditions. Implicitly the accounts postulate that registered firms’ costs and profit margin are the most relevant factor in a definition of economic growth. This means that any ‘cross checking’ is done between the original company figures, and derivatives of this initial account – (a tautological exercise). Further, income of households is not estimated independently, and there is no rationale why changes in company costs plus profits (which is a hazardous value in itself) should approximate to income changes. We should note that Kuznets (1941), in his tome on national income and accounts stated that income, output and expenditure could never equal, theoretically or empirically. It is in fact curious how the ethos of the macro economist, that ‘the macro economy is more than the sum total of its parts’ is ignored when it comes to accounting for that macro economy. The aim of these two preceding sections is to provide a solid understanding and critique of the current national accounting framework. Each section serves as individual contributions to knowledge on their own merit, and further they serve as a rhetorical device for the third section , which aims to present an alternative angle for estimating ‘national income’ by using a concept of output utilization and access, rather than the cost approach. 6 For example, UK accounting rules on the treatment of Goodwill writes it off instantly, while US firms can retain goodwill on their accounts for longer periods. 7 See for example Summer & Heston (1991, 1999, 2003) 8 See for example Mill 1844, Essay II. 4 Taking the stance that for countries whose economies are not permeated by market activity or where income and goods are not distributed evenly across the economy (both in a geographic and monetary sense), it is of more interest to the policy maker, investor, economist and politician to define improvements in the economy by how much its populace can and has consumed, in terms of the quantity of output, and not its cost. I.e. rather than look at how much hospital beds, rice or electricity cost firms last year, it is more interesting to investigate how much was consumed and where (or by whom). This is not to say that the economy’s cost structure is not important, but it will not be the centrepiece of the concept of growth, rather it is one component of the multifaceted economy under analysis. What is central is to find a common estimator for changing heterogeneous output, in a changing demography, and I propose this to be ‘usage’, in terms of ratios of what is being offered in the economy, against what is being purchased, consumed and utilized. Literature Review Over the last 30 years, authors have continuously expressed their dissatisfaction with measures of economic growth, but even alternative approaches adhere to the cost approach with a focus on production as defined by expenditures for their measure of national income. A review of the literature, will highlight how the critiques of GDP as a welfare measure, generally argue about the non-inclusiveness of the variable and the accounts, but fail to question the issue of whether GDP is representative of economic growth, and what this means. These approaches can be categorized into three distinct blocks. 1. The composite indices: Over time, multiple focuses and composite indices have been suggested to either complement or replace GDP. The most famous contributions have been The Purchasing Power Parity measure, Physical Quality of life Index, The Human Poverty Index and the Human Development Index9. They all suffer from a similar dependence on GDP as either a proxy variable for growth or as the key variable in the index, leading to their effective refusal as an alternative to GDP10 2. The alternative aggregates: To deal with the issue of measuring aggregate economic activity and growth from a purely production perspective – which was found to be unacceptable by many 9 10 See Lan (2002) on PPP, Morris (1979) on PQLF, Seers (1972) on HPI or UNDP (1990, 1993) for HDI. Ravaillon (1997) on the HDI, McGillivray (1991) on the HPI, Sen (1983) on entitlements or Lall (1983). 5 economists. They responded by developing multiple ‘alternatives’: The Measure of Economic Welfare (MEW), Economic Aspects of Welfare (AEW), Index of Sustainable Economic Welfare (ISEW) and the Genuine Progress Indicator (GPI)11. However, none of these depart from the traditional national accounting framework, rather they pick and choose the components they wish to include or exclude, while making general assumptions about the nature of land ownership and arbitrary ‘observations’ like the cost of commuting. These aggregates, although interesting, are unable to vary much from GDP, and still rely on the framework, making them untenable as alternative measures. 3. The Socio-Economic Approach: Typified by Sender12; this approach avoids GDP and aggregate measures altogether, and looks at socio-economic data sets to establish whether countries have improved on the micro level. While providing useful snapshots of literacy rates, mortality rates and physical infrastructure, it is impossible to establish whether the population as a whole are better off, and if so, where and why this is coming from. Further, the micro surveys are unable to give us any notion of the consumption and productivity of the population, or any sort of consistent variable or set of variables by which to measure this. As such, they are more of a complement to a wider macroeconomic understanding of economic growth, as opposed to a stand-alone measure. The growing literature on empirical work within national accounting often attempts to fix problems without distinguishing between root causes in the accounting framework itself and other factors. For example Summer & Heston (2003) show how disaggregating price indices regionally provides a better understanding of real income figures, and while their critique that aggregation across overly large sample may compromise the applicability of those figures regionally, they fail to recognise that their income values are company costs being deflated by market price changes, not necessarily improving the actual income estimates. Similarly Heston, Deng, Feenstra and Timmer (2004, p.2), adjust expenditure figures in respect to Terms of Trade and conclude that “differences between real GDP measured from the expenditure and output-side can be substantial, especially for small open economies”, but rather than ask why or how these accounts may diverge, the tacit conclusion is that nominally they must equal. 11 See Nordhaus & Tobin (1972) and Stewart (1974) for MEW, Zolotas (1981) for AEW, Daly (1988) and Daly & Cobb (1989) for ISEW or Cobb, Halstead & Rowe (1995) for GPI. 12 See Sender (1999). 6 There are in fact few papers which attempt to tackle the national accounts head on, and explicitly take on the underlying theories in the accounting system. Ogle (2000) shows how the national accounts carry with them a Keynesian heritage, but Kahn (2005) illustrates how a value theory can be teased from the framework, and this value theory renders the analysis of productivity growth in an economy powerless to observe actual productivity changes. It is shown that with zero change in output, a rise in the number of industry intra-trade boosts GDP, while efficiency gains lowers the GDP estimate – the latter being inconsistent with any theoretical notions of an efficiency, growth trade-off. Cross country GDP estimates appear to be identical and comparable, but they have vastly different methods of calculation, meaning that two different countries’ GDP rates, are two different types of estimate, leaving meaningful comparison impossible (Ruggles 1999), and Kahn (2005) shows that identical output growth leads to different GDP estimates for different countries. Comim (2001) discusses how cross checking is not undertaken in the accounts, as the income, expenditure and output accounts are not estimated individually, but are rather derived from the same expenditure account from firms. Ruggles (1994) showed how the commodity flow is estimated from a production view, whereby the “Household sector accounts are derived residually” (P. 491), and consumption is a simple derivative of this derived income (Ruggles 1949, p.15). Savings and investment are set to equal, while “the actual transactions of households were never incorporated as part of the national accounts” 13 Proposed Methodology The first section will require extensive use of original source material, working from William Petty, through King, the physiocrats, the Italian school and the marginalists, up into the modern national accounting method. I would like to expand on my 2005 master’s dissertation, where I trace how the methodology of the national accounting framework, throughout time, has been made to fit with political, economical, and other agendas. With some indications that the current consensus is not so different in its root causes. This approach requires a historical analysis, using the sources available at the British library, and I would like to take this analysis one step further in looking at the sociology of the economists who established the modern framework (Keynes, 13 Ruggles & Ruggles, 1949, 1991, 1995 [p.65], quote cited in Kahn 2005 [p.26] 7 Kuznets, Stone, Loveday and Meade in particular), and this is possible through their recorded correspondence via HM Treasury, Cambridge University and The League of Nations. The second section will focus on the critiques of the underlying structure of accounts, which is a mixture of theoretical questions, in defining the value theory implicit in the accounts, and the epistemological claims and evolution of macro balancing identities. With technical arguments surrounding specific accounting practices both on a local and national level for each country, and for how the national accounts are composed. A comparison of UK, India and the US accounting methods, using UN guidelines together with nation specific manuals for national accounting, will trace out the individual characteristics of each country’s national income accounts, and their theoretical and empirical flaws. Again sources are available in the greater London area through the University of London libraries and the British library (particularly their India section). For the third section, I will propose one approach to accounting for economic growth, through analysing the expansion and diversification of output, placing an emphasis on that which is produced domestically, and that which is consumed domestically. This way, one re-defines economic growth as the improvement in use, access and utilization of goods and services in the economy, as opposed to the aggregate costs of firms. Dividing the quantity (not costs) offered & produced with that which is utilized, purchased and consumed, means the measure becomes a set of ratios, which can be aggregated for individual sectors, regions or analyzed individually. The producing sector would be disaggregated by their method of production, the expansion of non-renewable extraction (e.g. mining), renewable output (e.g. crops) and unlimited input (e.g. wind-farms), comparing each to it’s input-output ratios and the cost structures of the relevant firms, we can estimate improvements in efficiency and production. The financial accounts that arise from here can be traced through to estimate the circular flow of money through the economy, giving an idea of the its velocity, the level of monetization and which groups spend or earn more. The stock of re-producing capital (e.g. livestock) and private capital is taken into account, and changes herein is added to the output increases which are then included in consumption calculations. A stock of natural resources remaining (and re-planted) is included to give a sense of the efficiency changes in extraction vis-à-vis inputs, and for a sense of sustainability of production. 8 This chapter will first require a framework for a set of accounts to be arranged, which would have to take into account country specific institutions and industries. For the US data and surveys are available through the US Census Bureau, the Bureau of Labour Statistics and the Bureau of Economic Analysis. Similarly for the UK, the Government Statistical office will be able to supply data and surveys from multiple sources. For India, some fieldwork is necessary as one has to travel to New Delhi to acquire the data from the Indian Statistical Institute, the Registrar Generals Office, while the Centre for Policy Research and the Indian Council for Research on International Economic Relations (ICRIER) have both recently worked on economic growth and development, which I learnt, doing some preliminary research during my time there this past summer. The methodology proposed stands aside from the current approach to economic growth, due to its initial rejection of GDP as a proxy for growth, and it proposes a real alternative measure of ‘Consumption ability’, based on quantity of output, and the cycles this produces in the economy. Further it would be beneficial to have an appendix on the process by which each country’s national account is composed, as a guide and explanation to other investigators. This Ph.D. can be expected to take between three and four years, with 14 months for the two opening chapters, which attempt to first offer a historic appreciation for the mutability of concepts like ‘economic growth’, and thereby allow the reader to see that GDP is not a definition written in stone, at the beginning of time. Then the consistent critique of the current framework will show that the method which served so well as a war budget planning tool in the 1940’s, is not consistent, nor relevant when applied to the economic world, and its problems 60 years after. It is with these considerations that I argue for a re-prioritization, and a change in the way we approach entities like ‘production’, ‘the economy’ and ‘economic growth’. With an expected semester in New Delhi, and another gathering data for the UK and US, while building a framework. The remaining time will be spent writing up the thesis, and the third section, which offers one path for an alternative concept of economic growth. This measure emphasizes first what people can access and consume, and then explicitly looks at the flow of money and the stock of resources and how these are employed and extracted in the economy. This is offered in contrast to the traditional method of aggregating the costs of registered firms, while imputing and deriving missing flow balances, based on the concept that an economy is commensurate to one big firm, where the parts add exactly to the sum total. This alternative statistic centres on consumers, and not simply costs, shifting the focus of growth analysis from a mentality of ‘people for goods’ to ‘goods for people’. 9 References: Bureau of Economic Analysis (BEA); 2000; GDP: One of the Great Inventions of the 20th Century; Survey of Current Business; (January) Cobb, Clifford, Ted Halstead and Jonathan Rowe; 1995; The Genuine Progress Indicator; San Francisco; Redefining Progress Comim, Flavio; 2001; Richard Stone and Measurement Criteria for National Accounts; History of Political Economy; Annual Supplement to volume 33: “The age of economic measurement”; Vol. 33; pages 213-232 Daly, H. E.; 1988; On Sustainable Development and National Accounts; in Collard, C., Pearce, D. and D. Ulph (eds.); Economics, Growth and Sustainable Environments: Essays in Memory of Richard Lecomber; Basingstoke; Macmillan Daly, Herman and John Cobb Jr; 1989; For The Common Good: Redirecting the Economy Toward Community, the Environment, and a Sustainable Future; Boston; Beacon Press Heston, Alan; Feenstra, Robert; Deng, Haiyan and Marcel Timmer; 2004; Estimating Real Production and Expenditures Across Nations: A Proposal for Improving Existing Practice; University of Pennsylvania; http://pwt.econ.upenn.edu/papers/FeenstraHestonTimmerDeng.pdf ICRIER; Krishna, K. L.; 2004; Patterns and Determinants of Growth in Indian States; ICRIER working paper 144; Indian Council for Research on International Economic Relations ICRIER; Virmani, Arvind; 2004a; India’s Economic Growth: From socialist rate of growth, to Bharatiya rate of growth; ICRIER working paper no. 122; ICRIER ICRIER; Virmani, Arvind; 2004b; Economic Performance, Power Potential and Global Governance: Towards a new international order; ICRIER working paper no. 150; Kahn, Benjamin Hav; 2005; Economic Growth: Measuring the wrong variable; incorrectly; School of Oriental and African Studies; (journal submission, mid 2006) http://beard.dialnsa.edu/~kahn/KahnMscDiss.doc Kahn, Benjamin Hav: 2006; CGE models: Not General, Not Equilibrium and fixed by assumption, not economics; New School University; (forthcoming) Keynes, John Maynard; 1940; How to pay for the war; London; Macmillan King, Gregory; 1696; Natural and Political Observations and Conclusions upon the State and Condition of England; (re-published in 1932) Kuznets, Simon; 1941; National Income and its composition, 1919-1938; 2 volumes; New York; National Bureau of Economic Research Lall, Deepak; 1983; The poverty of development economics; Hobart Paper no. 16; London; Institute of Economic Affairs; 10 Lan, Yihui; 2002; The Explosion of Purchasing Power Parity; Unpublished Ph.D; University of Western Australia, under supervision of Prof. Ken Clements; http://www.ecom.uwa.edu.au/__data/page/36742/ExplosionPPP.pdf (April 2005) McGilliwray; 1991; The human development index: Yet another redundant composite development indicator?; World Development; Vol. 19 (10); pages 1461-68 Mill, John Stuart; 1844; Essays on Some Unsettled Questions of Political Economy; published in 2000 by Batoche Books, Kitchener; 2nd edition Morris, D.; 1979; Measuring the Conditions of the World Poor, the Physical Quality of Life Index; New York; Pergaman Press Nordhaus, William D. and James Tobin; 1972; Is Growth Obsolete; Economic Growth; Fiftieth Anniversary Colloquium; Vol. 5; New York; National Bureau of Economic Research Ogle, Greg; 2000; Between Statistical Imperatives and Theoretical Obsessions: An Inquiry into the Definition and Measurement of the Economy; Unpublished Ph.D. Thesis; supervised by Prof. Nancy Folbre; University Of Massachusetts, Amherst; http://users.senet.com.au/~gregogle/PhD%20home.html (April 2005) Petty, Sir William; 1676; (published 1691) Political Arithmetick, or a discourse concerning the ... value of lands, people, etc; Mortlock at the Phoenix; St. Pauls Church Yard; London Quesnay, Francis; 1758; Tableau etc.; facsimile 1894; Macmillan, London; all quotes and references are to: Oncken, Auguste; 1888; Oevreus Economiques et Philosophiques de F. Quesnay; Paris. (Translated by Studenski, 1958). Ravallion, M; 1997; Good and bad growth: The human development reports; World development; Vol. 25 (5); pages 631-38 Ruggles, Richard; 1949; National income Accounting and its Relation to Economic Policy; printed in Ruggles, Nancy D & Richard Ruggles; (1999) Ruggles, Richard; 1991; National Income Accounting Concepts and Measurement: Economic Theory and Practice; Presented at conference in Siena; printed in Ruggles, Nancy D & Richard Ruggles; (1999) Ruggles, Richard; 1994; Issues relating to the UN System of National Accounts and Developing Countries; in Ruggles, Nancy D & Richard Ruggles (1999) Ruggles, Richard & Patricia Ruggles; 1995; The Value Added of National Accounting; in Ruggles, Nancy D & Richard Ruggles; (1999) Ruggles, Nancy D & Richard Ruggles; 1999, National Income Accounting and Economic Policy: The united States and UN Systems; pp 144-159; Edward Elgar; Cheltenham UK Seers, Dudley; 1972; What are we trying to measure?; in D. Lehman (ed.); 1979; development Theory: Four critical studies; London; Frank Cass 11 Sen, Amartya; 1983; Development: Which way now?; The economic journal; Vol. 93, no. 372; pages 745-62 Sender, John; 1999; Africa’s Economic Performance: Limitations of the current consensus; Journal of Economic perspectives; Vol. 13 (3), summer; pages 89-114 Stewart, Kenneth; 1974; National Income Accounting and Economic Welfare: The Concepts of GNP and MEW; St. Louis, April; Federal Reserve Bank of St. Louis Stone, Richard and D. G. Champernowne; 1940; The Precision of national income estimates; The review of economic studies; vol. IX, no. 2; pages 111-25 Stone, Richard & J.E. Meade; 1941; The construction of tables of National Income, Expenditure, Savings and Investment; The Economic Journal; Vol. 51; No. 202/203; Jun-Sep; pages 216233 Stone, Richard; 1986; Nobel memorial lecture 1984: The accounts of society; Journal of Applied Econometrics; 1 (1), January, pages 5-28 Studenski, Paul; 1958; The Income of Nations: Theory, Measurement and Analysis Past and Present; New York; New York University Press Summers, Robert, and Alan Heston; 1991; The Penn World Table (Mark 5): An expanded set of international comparisons, 1950-1988; Quarterly Journal of Economics 106, no. 2 (May): 327-68. Summers, Robert and Alan Heston; 1999 ; The World Distribution of Well-being Dissected, Chapter 16 in Alan Heston and Robert E. Lipsey, (editors) 'International and Interarea Comparisons of Income, Output, and Prices ' (1999) Summers, Robert and Alan Heston; 2003; Regional Output Differences in International Perspective; Center for International Comparisson; http://pwt.econ.upenn.edu/papers/RevisionLSE0103.pdf UNDP; United Nations Development Programme (UNDP); 1990 and 1993; Human Development Report; New York; Oxford University Press Zolotas, Xenophon; 1981; Economic Growth and Declining Social Welfare; Bank of Greece; Athens 12