Economic Growth and the Balance-of-Payments Constraint: An Introduction John Mc. Combie Director Cambridge Centre for Economic and Public Policy University of Cambridge 1
Introduction This lecture provides an introduction to the concept of the demand-oriented approach to economic growth and the concept of balance-of-payments equilibrium growth. First put forward in a short paper in 1979 by Tony Thirlwall in BNL Quarterly Review entitled “The Balance of Payments Constraint as an Explanation of International Growth Rate Differences”. Led to a huge literature testing, and theoretically, extending the model. What is it, how successful is it and what are the criticisms of model? 2
Introduction Approaches to economic growth Two distinct paradigms (1) The neoclassical supply-oriented approach, based on the production function and full employment. Often perfect competition is assumed. (i) The Solow growth model (ii) Endogenous growth models Generally closed economy models, no role for monetary factors or demand. Theoretical and empirical problems with the aggregate production function. 3
Introduction (2) Demand oriented and balance-of-payments constrained growth Why does demand matter for growth? (i) (iii) Harrod-Domar model The endogeneity of the natural rate of growth Kaldorian approach and cumulative causation But trade needs to be balanced in the long-run. . . (iv) Balance-of-payments constrained growth 4
Introduction (2) The Demand Oriented Approach (including the balance of payments growth model) (i) Harrod – Domar model: y. A = s/v where y. A = actual growth rate; s = savings ratio and v = the incremental output ratio. Generalisation of the Keynesian model to the long-run and used until recently by the World Bank. Focus on the difference between y. A and yn. yw and y. A Solow model or Cambridge growth models 5
The Endogeneity of the Rate of Growth • Endogeneity of the natural rate of economic growth or the growth of productive potential. • In the neoclassical approach both the Solow and the endogenous growth model) this is taken as exogenous, independent of the growth of demand. y = a - b U% Yn = a when U%=0 But what happens if growth is above the natural rate? 6
The Endogeneity of the Natural Rate Including a dummy variable y = a - b %U + c. D It can be seen that there is a shift of the natural rate of growth from a (a 2) to a + c (a 2 + b 2) 7
Empirical results for the advanced Countries. “The implication for growth Theory and policy is that it Makes little economic Sense to think of growth as supply constrained, if Demand, within limits, Creates its own supply” (Thirlwall, 2002, p. 95) Supported by other more recent evidence. 8
The Endogeneity of the Natural Rate Lanzafame (2014) 9
Growth and Cumulative Causation (ii) The Kaldorian view of growth (cumulative causation) The rate of capital accumulation and the savings ratio are not exogenous determinants of economic growth but endogenous. Because of increasing returns to scale, growth proceeds in a cumulative and circular manner (Myrdal, 1957). “In the case of industrial activities (‘manufactures’) the impact effect of exogenous changes in demand will be on production rather than on prices. “Supply’, at any rate long run supply, is normally in excess of demand – in the sense that producers would be willing to produce more, and to sell more, at the prevailing price (or even a lower price) 10 in response to an increased flow of orders.
Growth and Cumulative Causation Capital accumulation is the result of output growth (derived from profits from increased production not an exogenous determined savings ratio). The most important exogenous component of demand is that from exports. 11
Export growth output growth productivity growth greater price and non-price competitiveness faster export growth . . . • Demand orientated: What determines the growth of exports? • Crucial role of increasing returns to scale (cf the Solow model). • Supply responds to the growth of demand. . . Capital accumulation is a function of the expected growth of demand; as is labour (disguised unemployment; greater efficiency, in migration). 12
13
Structure of the Model Complete overview yt = xt (1) xt = - t-1 + w + z (2) t = w – pt (3) pt = re + yt (4) [output growth is determined by export growth] [export growth is determined by the growth of prices and world income] [growth of regional prices is determined by growth of money wages and productivity] [Productivity growth is determined by the growth of output] growth is a “circular and cumulative” process 14
The model in diagrammatic form Convergent growth Growth of relative prices =f(productivity Growth) Verdoorn’s law Export growth = f (growth of relative prices) Foreign trade multiplier 15
Growth and Cumulative Causation (i) The solution to the model is that countries/regions converge to (different) equilibrium paths (Dixon & Thirlwall) or (ii) There is explosive growth which is limited by bringing in supply side constraints (labour shortages or land rents) (Kaldor) But there is a problem that in equilibrium the growth of exports and imports can differ. There has to be a mechanism for ensuring that this cannot happen indefinitely. 16
If there is no feedback through the Verdoorn effect and the growth of relative prices, then the model reduces to y = ( z) which brings us to a consideration of Thirlwall’s law and balanceof-payments constrained growth. 17
Balance-of-Payments Constrained Growth (2) Balance-of-payments constrained growth For Keynesians, demand drives the economic system, but there is a constraint imposed by the balance of payments. If demand is increased, imports also increase but exports are remain the same (determined by the growth of its world markets), so the economy moves into a balance-of-payments deficit. But this cannot last indefinitely. Two hypotheses • Deficit cannot be financed indefinitely by international borrowing (capital inflows). • Changes in (international) relative prices are ineffective. 18
The Balance-of-Payments Constraint The key proposition “At theoretical level, it can be stated as a fundamental proposition that no country can grow faster than the rate consistent with balance of payments equilibrium on current account unless it can finance evergrowing deficits, which in general it cannot”. (A. P Thirlwall) Evidence from the IMF suggests that as a rule of thumb international financial markets become distinctly nervous if the overseas debt to GDP ratio exceeds around 50% or the current account /GDP ratio is about 10% It does, however, depend on the size and level of development of the country concerned. 19
The Growth Rates of Countries are Interlinked • The growth rates of even the advanced countries are inextricably linked – LINK model • If there is a slowdown in growth of the US (or more recently China), this impacts on the rest of the world. • The beggar-my-neighbour policies of the advanced countries in the 1930 s. “Exporting unemployment” by tariffs and quotas. • Rapid post-war growth led by first the US and then Germany as the engine of growth. Rapid reduction of tariffs under GATT. It was the growth of world demand that became crucial for the Golden Age. 20
The Growth Rates of Countries are Interlinked The Case of the US The impact of the “Great Recession” : 21
The Balance-of-Payments Constrained Growth Model A Quick Overview The approach concentrates on the determinants of export and import growth; such as the relative prices of exports and imports (price competition) and the income elasticities of demand (non-price competition). The causation is from export growth output growth import growth Not the Rodrik approach of undervaluation of REER changes in price of tradables/nontradables greater investment faster output/export growth. 22
But did not the floating exchange rates essentially delink the economies of the world? No. . . • In the current crisis, attention is still focused on the degree to which the US and China can lead the world out of recession. • The “lost decade of growth” in Latin American associated with balance of payments crisis and collapse of the currency. Mexico, Brazil and Russia • The 1997 Asian Crisis; Globalisation has exacerbated the situation. Excessive borrowing; rapid turnaround from inflows to outflows, collapse of the currency; internal liquidity crisis as the repayments of debt in the domestic currency rocketed. Evidence discussed later 23
The Balance-of-Payments Constrained Growth Model A Quick Overview A very parsimonious model. Three equations: (i) A dynamic Keynesian demand for export function (x = z) (ii) A dynamic Keynesian demand for imports function (m = y) (iii) A balance-of-payments constraint that the growth of exports and imports must be equal (although the growth of capital flows can be incorporated into the model) (m = x) Substituting (i) and (ii) into (iii) gives y. BP = z/ 24
The Balance-of-Payments Constrained Growth Model Thirlwall’s law y. BP = x/ = z/ This is the locus of growth rates for which the balance of payments is in equilibrium. It is a function of the ratios of the income elasticity of demand for exports and for imports. These reflect differences in a country’s non-price competitiveness. It also demonstrates the importance of export led growth, 25
The Balance-of-Payments Constrained Growth Model • It is explicitly demand oriented; the supply side adjusts to the long term growth of demand. • The balance of payments constrained growth rate is the maximum a country can grow at without encountering balance of payments problems. • The supply side can be incorporated into the model (later). • The simplest case can be illustrated in the next figure. 26
27
The Balance-of-Payments Equilibrium Growth Model Let us now justify the model in greater detail. Categories of growth y. P = the growth of productive capacity (maximum growth rate). y. BP = the balance-of- payments constrained growth rate. y. A = the actual growth rate. y. P > y. BP y. A Balance-of-payments constrained growth (lower rate of capital accumulation, technical progress, disguised unemployment) 28
The Balance-of-Payments Equilibrium Growth Model • Not all countries can be simultaneously balance-of-payments constrained. (An exception is a deep recession and is due to the dynamics of adjustment). • Some countries will be either resource constrained (or supply-side constrained) or policy constrained (currently Germany? ) • Therefore, failure to find one country is not balance-of-payments constrained does not refute theory as a whole. 29
The Balance-of-Payments Equilibrium Growth Model The Extended Model 30
The Balance-of-Payments Equilibrium Growth Extended Model (1) Export demand equation x = z - (pd - pf - er) The growth of exports (x) is determined by the growth of world income (z) and the rate of change of relative prices. is the price elasticity (2) Import demand equation m = y + (pd - pf - er) The growth of imports (m) is determined by the growth of domestic income (y) and the rate of change of relative prices. is the price elasticity. 31
The Balance-of-Payments Equilibrium Growth Model (3) Balance-of-payments identity x + (1 - )f m + (pf – pd – er) The (weighted) growth of exports and the (weighted) growth of capital flows (f) must equal the growth of imports plus the rate of change of prices in a common currency. 32
The Balance-of-Payments Equilibrium Growth Model (3) Balance-of-payments identity x + (1 - )f m Suppose relative prices do not change, then if the growth of imports is greater than export growth, there must borrowing from abroad; hence the growth of capital flows into the country. 33
Solution to the model We substitute three equations into each other and solve for the growth of domestic income. The growth of output y is determined by the growth of world income z (as this determines the growth of exports x), the rate of change of relative prices (relative price competitiveness) and the growth of capital inflows f. 34
Solution to the model 1. The first term z/ is the effect of the exogenous growth of world income. 2. The second term (1 - - )(pd -pf -er)/ gives the effect of the changes in relative prices. 3. The third term (1 - )/f gives the effect of real capital inflows/outflows. 35
Solution to the model (1 - - )(pd -pf -er)/ = 0 • • Marshall-Lerner condition is only just met Pricing to market under oligopoly J-curve effects Non-price competitiveness more important (1 - )/f = 0 • A country cannot finance a deficit indefinitely. Incorporating the growth of capital flows for plausible values makes little difference to the balance of payments growth rate. 36
This Gives Thirlwall’s Law y. B = z/ = x/ UK = 0. 5 = 1. 0 z = 4. 0% JAPAN = 3. 5 = 1. 0 z = 4. 0% If y = y. B then we can infer that x =2% y. BP = 2% x= 14% y. BP = 14% (i) the growth of capital flows is negligible (ii) price effects are also minimal 37
How do we Test the Law? 38
How do we Test the Law? • Estimate (Keynesian) export and import demand functions (including relative price effects). • Obtain statistical estimates of and . Estimate the export and import demand functions • Calculate the growth of the country’s export markets, z. • Calculate Thirlwall’s law. 39
Testing the Law Estimate • Ln. Xt = c + b 1 ln. Z + b 2 ln(RXP) + error term • ln. Mt =c + b 2 ln. Y + b 4 ln(RMP) + error term Note: No time-trend Originally OLS (Houthakker and Magee); now ECM, ADRL, testing for cointegration, panel data etc. Does not make a great deal of difference. Obain estimates of the long-run income elasticities , Calculate long-run growth rates (average over, say, ten years) of Z (weighted growth of country’s export markets) and Y. Calculate / z = y. BP 40
Testing the Law Qualifications • Appropriate estimation technique(s) • Specification. Should an export supply function be included? • XS = f (RMP), Y or K)? But Y and K are assumed here to be exogenous, so misspecification to include them. They are assumed to be endogenous in the Bo. P approach. • High frequency data, in the short-run, firms have excess capacity to meet extra demand; long-run induced investment increases supply. Nearly all estimations of import and export demand functions make this assumption. • Difficulty in measuring REER, RXP, RMP. Proxies have to be used. 41
A Digression As Why not simply estimate y = c + b 5 z + b 6(RP) + error term? • Loss of information • High frequency (annual) data • Serious misspecification omitting f, not constant. What happens when f switches from > 0 to <0? • Problem using two price terms (RXP, RMP). The term used above convenient for theoretical modelling. • Nothing really to be gained. 42
The Weak and Strong Tests of the Law Compare actual growth rate y and y. BP (1) Weak Test y. BP = x/ (2) Strong Test y. BP= z/ If they are close, then it shows that the growth of relative prices and capital flows are important and the country cannot grow faster than y. BP. If they diverge, then the rate of change of relative prices and/or the growth of capital flows are quantitatively important. 43
Parametric Tests of the Balance-of-Payments Constrained Growth Model • Regress y on y. BP across countries i. e. y = a + b y. B (a = 0; b = 1) But, possibility of bias: 1. Incomplete sample 2. Outlying observations • Mc. Combie Test Calculate ( ) separately for each individual country (y x/ ) and compare with estimated ( *). If not significantly different from *, y. B will be a good predictor of y. 44
Calculations of the Growth Rate Consistent with Balance-of-Payments Equilibrium 1951 -1973 Change in GDP % Change in Exports (x) % Income Elasticity of Demand for Imports ( ) Balance of Payments Equilibrium Growth Rate Austria 5. 1 10. 7 n. a. - Belgium 4. 4 9. 4 1. 94 4. 84 Canada 4. 6 6. 9 1. 20 5. 75 Denmark 4. 2 6. 1 1. 31 4. 65 France 5. 0 8. 1 1. 62 5. 00 Germany 5. 7 10. 8 1. 89 5. 71 Italy 5. 1 11. 7 2. 25 5. 20 Japan 9. 5 15. 4 1. 23 12. 52 Netherlands 5. 0 10. 1 1. 82 5. 55 Norway 4. 2 7. 2 1. 40 5. 14 United Kingdom 2. 7 4. 1 1. 51 2. 71 U. S. A. 3. 7 5. 1 1. 51 3. 38 Country 45
Comparing African with Asian Countries, (in percentage points) Country Actual Rate of Growth Terms of Trade Effect Export Volume Effect Real Capital Flows Effect Growth Rate Given by the Extended Model Algeria 4. 90 10. 15 4. 21 -8. 72 5. 64 Benin 2. 90 1. 44 0. 96 1. 35 3. 75 Burkina Faso 4. 20 -5. 17 3. 03 5. 63 3. 50 Burundi 5. 60 1. 69 3. 21 -1. 26 3. 65 Cameroon 5. 50 -1. 12 7. 08 0. 00 5. 97 Congo 6. 59 0. 42 3. 88 2. 38 6. 67 Cote d’Ivoire 4. 50 0. 39 4. 23 0. 81 5. 43 Egypt 6. 90 -2. 37 4. 36 7. 31 9. 30 Ethiopia 2. 20 -0. 09 0. 74 2. 53 3. 17 Gabon 5. 10 0. 49 6. 81 -0. 04 7. 33 Ghana 1. 40 -3. 81 0. 15 2. 88 -0. 79 Kenya 6. 24 -0. 50 1. 62 5. 59 6. 71 Lesotho 4. 40 -3. 43 6. 62 1. 55 4. 74 Madagascar 0. 48 -0. 10 0. 06 0. 95 0. 91 Mauritania 2. 30 0. 68 1. 58 0. 42 2. 69 Mauritius 5. 80 0. 92 5. 13 0. 19 6. 23 African Countries 46
Country Actual Rate of Growth Terms of Trade Effect Export Volume Effect Real Capital Flows Effect Growth Rate Given by the Extended Model Morocco 4. 62 -1. 34 2. 83 3. 47 4. 96 Niger 0. 81 -5. 07 1. 79 3. 47 0. 20 Nigeria 2. 50 2. 37 1. 28 -1. 17 2. 48 Senegal 2. 67 0. 23 1. 56 1. 05 2. 83 Sierra Leone 1. 58 -0. 23 -0. 67 2. 65 1. 75 Somalia 3. 40 -1. 10 0. 18 5. 00 4. 07 South Africa 2. 42 -1. 03 1. 32 7. 74 8. 03 Sudan 3. 10 0. 14 1. 13 1. 92 3. 20 Tanzania 2. 90 0. 33 -0. 55 5. 01 4. 79 Togo 2. 90 0. 08 2. 31 0. 61 3. 00 Tunisia 5. 69 0. 87 5. 24 1. 48 7. 59 Zambia 1. 40 -0. 31 -1. 29 0. 58 -1. 02 Zimbabwe 3. 23 -2. 40 2. 23 -1. 24 -1. 41 47
Estimates of Balance-of-Payments Constrained Growth for Asian and Latin American Countries GDP growth (y) Export growth (x) Income Elasticity of Demand for Imports ( ) Balance of Payments Constrained Growth Rate (y. B) Indonesia 6. 90 16. 3 2. 98 5. 47 Malaysia 7. 40 14. 5 2. 25 6. 44 Philippines 3. 70 9. 9 1. 92 5. 16 Thailand 7. 60 13. 0 2. 86 4. 55 Costa Rica 4. 7 5. 8 1. 10 5. 26 El Salvador 3. 4 3. 3 1. 75 1. 88 Guatemala 3. 8 4. 4 1. 35 3. 34 Honduras 3. 8 2. 7 3. 70 0. 73 Nicaragua 2. 6 3. 4 2. 04 2. 10 Country Asian countries a Latin American Countries b Sources: a. Ansari, Hashemzadeh and Xi, ‘The Chronicle of Economic Growth in South East Asian Countries: Does Thirlwall’s Law Provide an Adequate Explanation? ’ Journal of Post Keynesian Economics, 2000. 48 b. Moreno-Brid and Perez, ‘Balance of Payments Constrained Growth in Central America’, Journal of Post Keynesian Economics, 1999
Considering Price and Non-price Competition Further 1. The “Kaldor Paradox” 2. Is the growth of relative prices important? If not, why not? 3. Is Non-price competitiveness important? The results of recent research. 4. “What you export Matters” (Hausmann et al. ) 49
1. The “Kaldor Paradox” Countries with deterioration in price competitiveness experienced an increase in export shares. 50
1. The Kaldor Paradox Still persists. . . 51
2. Is the Growth Rate of Relative Prices Important? Note that to relax the balance-of-payments constrained, there must be a continuous depreciation of the growth of relative prices. This is because of the multiplicative nature of the import and export demand curves. Thus: An increase in the growth of exports requires a continuous real depreciation of the currency. Is the rate of growth of exports determined by the changes in the level of RXP? Implausible. (i) No evidence (ii) Cannot specify the export demand function in log-level terms (iii) A continuous real depreciation would lead to an acceleration in the rate of growth of exports. 52
2. Is the Growth of Relative Prices Important? Estimates of the relative price of exports and imports for the advanced countries UK OBS model UK CBR model Crane et al Canada France Germany Italy Japan UK US Exports -0. 74 (goods) -0. 47 (services) -0. 25 Imports -0. 32 -0. 07 -0. 61 -1. 14 -2. 86 -0. 61 -1. 15 0. 08 -0. 74 -0. 23 -0. 34 0. 05 +1. 17 -0. 60 -0. 61 -0. 63 =53
2. The Study of Chritodoulpoulou and Tkacevs (2014) • ECB publishes Harmonized Competitiveness Indicators of euro member states. • • • Consumer price index (CPI) Domestic sales producer price index (PPI) GDP deflator (GDPD) Unit labour costs in manufacturing (ULCM) Unit labour costs in the total economy (ULTC) Estimated standard export and import demand functions using all of these to see if any performed noticeably better than the others. In all cases the price elasticities were low and the M-L conditions not 54 statistically significant.
2. Is the Growth of Relative Prices Important? Bahmani (et al, 2013). A comprehensive survey studies of the M-L conditions. They find relatively little comprehensive analysis support for the Marshall-Lerner condition. “The results of our analysis are clear: The M-L condition does not hold in a large fraction of the cases in which it is claimed to do so. This has strong implications for further analysis of trade and exchange rate policy” (p. 435). “Those who draw on previous studies to support certain expected benefits from a currency depreciation should think twice before they rely too heavily on such estimates” (p. 439). 55
2. Is the Growth of Relative Prices Important? Why are changes in RXP likely to be unimportant? • Firms do not compete nationally in the long-run with each other in terms of price. Match price and compete in terms of quality etc. (Confirmed by research by the business schools) • Changes in export prices due to exchange rate fluctuations are outside the control of the firm. May be transitory if driven by changes in capital flows. Therefore unlikely to lead to an increase in export capacity. • Shift the demand curve out, not move down the demand curve. For evidence, see Mc. Combie and Thirlwall (1994, Chapter 4) 56
2. What about Primary Commodities? Thursby et al (1986); Wheat “Wheat is one of the most narrowly defined primary products. . . And the raw data are not favourable to the LOP [law of one price]” Ardeni (1989) Wheat, tea, beef, sugar, wool, zinc, tin, “ The law of one price fails not only in the short run, because of slow pass -through, stickiness of prices and various slow adjustments, as many authors have argued, but it also fails in the long run as a law”. 57
3. Is Non-price competitiveness important? The results of recent research. The ECB’s Competiveness Research Network (Benkovskis & Worz) The rediscovery of the “Kaldor Paradox” “How can we reconcile real effective exchange rate (REER) – and hence an apparent deterioration in cost competiveness – with rising world market share – a clear sign of improved competitiveness”. “Catching up in terms of prices [i. e. a deterioration in price competitiveness] while increasingly being able to sell goods and services at the world market has been observed for many emerging economies that are considered as being the most competitive nations in the world. In contrast other countries - and among them often peripheral EU members – have sometimes greatly improved their cost competiveness, yet without a manifestation of increased competitiveness in the form of increased global market shares. ” (B&W) 58
Export market shares and relative export prices of selected Eurozone countries (1999 -2008) 59
Recent Research into the Importance of Non-price Competiveness • The basic approach is as follows (Idea goes back to the 1970 s) • Goods and services are not homogeneous. • Quality changes over time and better quality goods sell more but have higher prices. Thus part of, say, an increase in the market share of a dynamic economy may be due better quality but also a higher relative price. • Unit values: The value of imports divided by physical quantities. Thus at a high level of disaggregation, differences in unit values reflect differences in quality. 60
Recent Research into the Importance of Non-price Competiveness Conventional Relative Prices (RXP) and those adjusted for quality change (RXPAD ) 61
Conventional and Adjusted RXP 2012 62
Cumulative Change in Export Market Shares, Czech Republic and the UK China 63
Non-price competitiveness – econometric evidence The procedure is to include factors determining non-price competitiveness in the export and import demand functions (or net exports) – if significant the estimates of the income elasticities should be reduced. For the UK…. Greenhalgh (1988) “ This lends support to the view that price is not necessarily the most salient characteristic of complex manufactured products, whose typical quality is changing rapidly through time”. 64
Non-price competitiveness – econometric evidence • Athanasogoglu & Bardaka (2008) Greece. • Include the capital stock in the export demand function, following others. • Variables in logs. Coefficient of ln. K is 1. 3 -1. 5 But. . . K/Y = 3; ln. K = a + ln. Y = constant (ln ) + ln. X So the coefficient of ln. K must be statistically significant and 1. 0 Difficulty of finding good proxies. . . R&D, Patents, Investment ratio, . . . ? 65
The Competitive Advantage of Nations The Work of ME Porter 66
The Competitive Advantage of Nations Porter’s detailed studies of countries as to what determines the competiveness of nations and firms. He finds little evidence that it is related to price, low wages, exchange rates or interest rates. It is related to what he terms “innovation” which may be broadly interpreted to mean all aspects of non-price competition. 67
The Competitive Advantage of Nations “Companies achieve competitive advantage through acts of innovation. They approach innovation in its broadest sense, including both new technologies and new ways of doing things. They perceive a new basis for competing or find better means for competing in old ways. Innovation can be manifested in a new product design, a new production process, a new marketing approach, or a new way of conducting training. Much innovation is mundane and incremental, depending more on a cumulation of small insights and advances than on a single, major technological breakthrough. 68
The Competitive Advantage of Nations It often involves ideas that are not even “new”—ideas that have been around, but never vigorously pursued. It always involves investments in skill and knowledge, as well as in physical assets and brand reputations. . “Ultimately, the only way to sustain a competitive advantage is to upgrade it –to move to more sophisticated types [of products]”. Porter, (1990), “The Competitive Advantage of Nations”, Harvard Business Review 69
The Competitive Advantage of Nations Porter’s Diamond 70
The Competitive Advantage of Nations The stages of development 1. Factor driven where a country’s advantage is is driven by basic factors of production. There is little local technology improvement; and domestic firms used imported technology and there is some inward investment. 2. Investment driven when countries can compete in world markets for standardised products but domestic innovation is limited to the adaptation of processes and products. Labour costs still remain low and in spite of a more skilled labour force production is concentrated in labour intensive industries and demand conditions are relatively unsophisticated. 71
The Competitive Advantage of Nations 3. Innovation driven when as its name suggests the firms appropriate technology from abroad but develop it. It is here where spillovers from the spatial clustering become important. The importance of spatial “clusters” 4. Wealth driven. It is here when competition falls because of the rise of large firms. They begin to compete on price and there is vicious circle of decline (UK). Compare the narrow “variety hypothesis” of the “new trade theory” Does it merely show the irrelevance of standard trade theory? 72
3. The Key to Differences in Export and Import Growth The value of ; the world income elasticity of demand for a country’s exports. Differences reflect non-price competitiveness. Why do these values vary between countries? Composition of exports (not the advanced countries. ) Quality, delivery times, characteristics of the goods. The value of , the domestic income elasticity of imports may vary. High for developing countries. 73
4. What You Export Matters; Hausmann et al (2006) For an export i they construct a weighted productivity index of countries who export this good (PRODY) Do the predominantly advanced or less developed countries export this good? Then they construct an index for country j’s export basket based on PRODY or EXPY Jesus Felipe has shown that there is a close relationship between a country’s EXPY and its world income elasticity for its exports. 74
There is a close relationship between EXPY and export growth 75
And between growth and initial EXPY (controlling for other variables) 76
Introducing Capital Flows (Thirlwall and Hussain) y = x/ (i) y = (1/ )( z + (1 - )(f-pd)) (ii) Compares the predictive power of the simple and the extended model. Finds the latter gives a better correspondence. Also divides the ample into countries where the law over-predicts the growth rate and under -predicts it. Other studies have taken up this theme. . 77
Introducing Capital Flows Thirlwall & Hussain (1982) 78
The New Economics of Capital Controls (Korinek 2011) “Fundamental Shock (to exchange rates & asset prices; balance sheets; aggregate demand) fall in aggregate demand exchange rate depreciates & asset prices fall adverse balance sheet effects (declining value of collateral and net worth) Imperfect markets constrain access of economic agents to finance and/or greater credit spreads cut back on spending, shock is amplified. A neoclassical justification for capital inflow controls based on a representative agent model and the failure to meet Pareto optimality conditions. 79
Capital Controls in a Wider Context: The Chinn-Ito Index 80
The “New View” “A key conclusion is that, if the economy is operating near potential, if the level of reserves is adequate, if the exchange rate is not undervalued, and if the flows are likely to be transitory, then use of capital controls—in addition to both prudential and macroeconomic policy—is justified as part of the policy toolkit to manage inflows”. Ostry (2010) What about the evidence as to the effectiveness of capital controls? (i) “Our sense is that the jury is still out on this” (Ostry et al. , 2010). (ii) Klein (2016) “ Gates and walls”. Neither types of capital controls are effective. 81
82
Theoretical Underpinnings of the Law • The Dynamic Harrod Foreign Trade Multiplier • The Dynamic Hicks Super-multiplier (extending the model to two sectors) 83
The Dynamic Harrod Foreign Trade Multiplier Simplest possible model from Harrod (with the usual notation. ) • Y = C + X • Y= C+M • M= m. Y For expositional ease, following Harrod, we ignore investment, savings, government expenditure and taxation and relative prices. 84
The Dynamic Harrod Foreign Trade Multiplier If X = M (equilibrium on the current account) Y = X/m (M = m. Y) Consequently, Y = (1/m) X Y/Y = (1/m)(X/Y)( X/X) But (1/m)(X/Y) = 1/ So Y/Y = ( X/X)/ or y = x/ (Thirlwall’s Law) 85
Harrod Foreign Trade Multiplier: Limitations The income elasticity of demand for imports is unity. Property of a linear import demand function and remedied by have having a conventional log-linear specification. What about the other exogenous components of demand such as investment, government expenditure? 86
The Hicks Super-multiplier When we introduce other components of domestic demand ( I, G, etc), the effect of an increase is the Hick’s super-multiplier which is (i) Harrod trade multiplier plus (ii) The effect of the relaxation of the balance-of-payments constraint that allows other components of domestic demand to increase. 87
The Extended Model The equation y. BP = z/ is a locus of growth rates where the balance of payments is in equilibrium. We need to introduce another element into the model. We still abstract from the supply side and use the traditional Keynesian income expenditure model. Income = (1/k)(Exports + Other Autonomous Expenditure) Y = (1/k)(X + A) where 1/k is the traditional Keynesian multiplier. But exports are a function of world income so X = f(Z). 88
Thus, in terms of growth rates, we have the growth of output: y = f 1(autonomous expenditure growth) + f 1(export growth) y = f 1 (autonomous expenditure growth) + f 2(world income growth or the growth of the other trading bloc). In other words, the growth of income will increase if export growth, investment growth or the growth of government expenditures increase. 89
The Extended Model In terms of growth rates, the growth of income is given by: y = f 1 (autonomous expenditure growth) + f 2 (world income growth or the growth of the other trading bloc). We shall term this the country growth equation. It is shown in Figure 1 as the dotted line AA together with y. BP. We assume that / = 0. 5. 90
Figure 1 Balance-of-Payments Equilibrium Growth A 2% A 0. 5 4% Z 91
Figure 2 An Increase in the Growth of Autonomous Expenditure A 1 Ao f A 1 Ao 4% Z 92
Figure 3 An Increase in Non-price Competitiveness A 1 A 0 4% Z 93
Figure 4 The Effect of a World Recession A 2 A 0 2% b f’ a A 3 1% 2% 4% Z 94
A World Recession We can see how if a major country reduces its growth for policy reasons (e. g. to combat inflation), it will induce a slower rate of growth of the other countries because of the balance-of-payments constraint. 95
Increase in Non-price Competitiveness We can see how an increase in non-price competitiveness will allow the country to increase its rate of growth without encountering the balance-of-payments constraint. 96
The Working of the Model So let us suppose that the country is growing below its potential and the government decides to increase the growth of its expenditures through deficit financing. This shifts the line AA upwards. But this is unsustainable as it needs the rate of growth of output given by f to be covered by the growth of financial inflows. 97
Balance-of-Payments Constrained Growth: An Illustrative Example Initial growth rates y. P Group I 5% Group II 5% y. B - 1. 0 2. 0 1. 0 m x 5% 10% 5% Balance-of-payments equilibrium growth rate y. P Group I 5% Group II 5% y. B 5% 2. 5% 1. 0 2. 0 1. 0 m 5% 5% x 5% The countries comprising Group I which is the more competitive trading bloc may be (i) resource constrained (ii) policy constrained. Its growth rate determines the growth of Group II which is balance-of-payments constrained. 98
The Multi-Sectoral Thirlwall’s Law Next major advance: Araujo and Lima (2007) Weighted separate income elasticises Latin America: Very little difference in the elasticities over time East Asia: Increase in aggregate as structural/sectoral change occurs. 99
Criticisms and Defence of the Balance-of-Payments Constrained Growth Model 100
Criticisms and Defence of the Balance-of-Payments Constrained Growth Model 1. The Law is merely an identity (Mc. Combie, 1980) 101
1. The Law is merely an identity Razami (2015) also repeats this mistake. But the values of the elasticities are estimate from export and import demand functions with relative prices included. If the latter are quantitatively significant, the estimates of and will be small and possibly statistically insignificant. In this case the law would not hold; it is a hypothesis capable of being refuted. 102
Criticisms and Defence of the Balance-of-Payments Constrained Growth Model 2. The assumption of “law of one price” renders the model incoherent • Implies homogeneous goods and service; high (infinite) price elasticities; relative prices therefore only change minimally to ensure balance-of-payments equilibrium. • Income elasticities should be statistically insignificant. • The “monetary approach to the balance of payments” 103
2. The assumption of “law of one price” renders the model incoherent “Law of one price” misnomer. Relative prices do not change greatly 1. Oligopolistic markets; pricing to market. 2. Pass through from increase import prices to domestic inflation 3. Price elasticities are low because of product differentiation and the importance of non-price factors. 104
Criticisms and Defence of the Balance-of-Payments Constrained Growth Model 3. Growth is determined by the supply side (i) Krugman’s variety hypothesis and the 45 -degree model (ii) Palley. Income elasticities adjust to bring the balance of payments back into equilibrium. Palley: Confusion over short-term elasticities, which vary with the trade cycle, and long term-elasticities. 105
3. Growth is Determined by the Supply Side (i) Krugman’s model • Growth of output (by assumption) exogenous increase in number of varieties growth of exports • Number of varieties a function of a country’s total output (rather than the size of the industry) See Barker (1977). • Changes in relative prices unimportant and the number of varieties affect non-price competitiveness. 106
Krugman’s 45 -degree Model 107
Krugman’s 45 -degree Model Tested by Gagnon. Estimated an export demand function Growth of exports = Coefficient 1. rate of change of relative prices, Not statistically sig. 2. growth of foreign expenditure; (fexp) 1. 50 (sig. ) 3. growth of the ratio of GDP of 1. 50 (sig. ) exporter to world GDP (y-z) Variable 3 (y-z) tests the variety hypothesis 108
Krugman’s 45 degree Model Estimating an export demand equation – Panel data 109
Krugman’s 45 -degree Model • But results are compatible with Bo. P constrained growth model • x = f ( rp, fexp, (y-z)) x = 1. 50 y (cross-country regressions) fexp and z are correlated • x = f ( rp, fexp, (2/3 x-z)) Coefficient on (y-z) must be significant and greater than 1. 110
Krugman’s 45 -degree Model Another problem • Two countries A & B. A’s rate of technical progress improves increased number of varieties increased growth of exports ( imports to B). • B’s growth is exogenously determined. • Runs an increasing balance-of-payments deficit. • No adjustment mechanism to bring the balance of payments back into equilibrium. 111
Granger (Predictive) Causality What does it tell us? (i) xt = a +b 1 yt-1 + b 2 yt-2 …. bt-n xt-1 …. (ii) yt = a +b 1 xt-1 + b 2 xt-2 …. bt-n yt-1 …. 1. Based on a short-term cyclical relationship; Thirlwall’s law is a longterm relationship between trend rates of growth. 2. Direction of causality with out-of-phase cycles problematical. 3. Large economy A enters a downturn; A’s imports from B fall leading to a fall in exports from A (B’s Bo. P constraint). ‘Causality’ for A appears as (i) but A could still be balance-of-payments constrained. 112
Granger (Predictive) Causality 4. Another example: Increase in export orders (t) increase in production to fulfil orders (t +1) recorded increased sales in exports (t +2). Appears as (i) but causation the other way around. 5. Case studies: balance-of-payments crisis associated with a rapid prior increase in output. 6. But most studies, do in fact find the direction of causality is from x y. 113
The Fallacy of Composition Can all countries simultaneously pursue an export-led growth strategy? Balance-of-payments equilibrium growth rate y. P y. B m Group I 5% 5% 1. 0 2. 0 5% Group II 5% 2. 0 1. 0 5% x 5% Balance-of-payments equilibrium growth rate y. P y. B m x Group I 5% 5% 2. 0 10% Group II 5% 5% 2. 0 10% 114
The Fallacy of Composition Balance-of-payments equilibrium growth rate y. P Group I ? Group II ? y. B 10% 3. 0 m x 10% 10% Here both countries pursue export-led growth policies and the income elasticity of demand increases to 3 in both countries. Growth rates increase ro 10 per cent per annum. 115
The Fallacy of Composition Blecker and Ramazi (2010) 116
Fallacy of Composition 117
Palley (2003) 118
Examples of the Use of the Balance-of-Payments Constraint Growth Model Pakistan: Can it reach its target rate of growth without encountering a balance of payments problem? 119
An Example: Is Pakistan’s Growth Rate Balance-of. Payments Constrained? • Pakistan’s growth rate was about 6% in the 1960 s. In the 1970 s and 1980 s it became a rapidly industrialising country, but never at the rate of growth of the Asian Tigers or China. • Poverty reduction plan 7 to 7. 5% growth per annum • But problem of recurrent balance-of-payments crises. • Latest was 2007 -2008. Had to go to the IMF for a loan of $11. 3 billion. • As Felipe and Lim (2008, p. 8) put it: “in the case of Pakistan, the current deficit reflects low export growth… and, ultimately export competitiveness problems. ” 120
Pakistan Economic Survey (2007 -08, p. xvi-xvii) The crucial question is. . . Can Pakistan reach its target growth rate of 7% per annum without a balance of payments crisis? Pakistan’s exports suffer from serious structural issues which need to be addressed primarily by the industry itself, with government playing its role of a facilitator. Textile is the backbone of Pakistan’s exports but bears various tribulations. 121
• These include: (i) low value added and poor quality products fetching low international prices; (ii) the machinery installed in recent years has depreciated considerably relative to Pakistan’s competitors; (iii) these machines are power-intensive, less productive, and carry high maintenance cost; (iv) augmented wastage of inputs adding to the cost of production; (v) little or no efforts on the part of industry to improve their workers’ skills; (vi) industry spending less money on research and development and; (vii) export houses lacking capacity to meet bulk orders as well as meeting requirements of consumers in terms of fashion, design and delivery schedule 122
123
124
Policy Implications • Supply characteristics are important. Improve . • It is no use if Pakistan is able to manufacture, say, cheap brass automobile radiators when technology has moved on so that the leading world automobile manufacturers are now using aluminium radiators. (The exception is that Pakistan can sell these radiators in its small protected domestic market. ) 125
126
Policy Implications Recommendations. Short run: Improve transport communications to the ports and reduce red tape with respect to exports. Long run: Tariff protection to reduce , but needs a sunset clause & Subsidies for exports, but again a sunset clause. 127
The Case of Thailand Tharnpanich & Mc. Combie (2013) Balance-of-Payments Constrained Growth, Structural Change and the Thai Economy. 128
The Case of Thailand Tharnpanich & Mc. Combie (2013) • Generally a success story; Growth of GDP 1962 -2009 was 6. 3% per annum. • Collapse in the 1997 -8 Asian crisis (largely precipitated by Thailand) • Rapid recovery but subsequent growth disappointing – why? The key is the Growth of exports. 129
The Case of Thailand Tharnpanich & Mc. Combie (2013) • Estimated import and export demand functions; testing for cointegration in the presence of structural breaks. • Used and error correction model and tested both the weak and strong for of the law. 130
The Case of Thailand Tharnpanich & Mc. Combie (2013) But the picture is a little more complicated. The divergence seems to have occurred in 1997/8. Precrisis years: Fast GDP growth was associated with a fast growth of especially manufactured goods. (The share of manufactured exports in the total 1960 s: 20% 2009 80%) Manufactured exports higher income elasticity of demand than primary products. 131
The Case of Thailand Tharnpanich & Mc. Combie (2013) 132
The export demand function during the boom years (1987 -1996) versus preboom (1962 -1986) 133
Rolling regression, 20 -year window 134
The Case of Thailand Tharnpanich & Mc. Combie (2013) What caused the decline in the income elasticity of Thailand’s exports? 135
The RER and Economic Growth An Alternative View Alternative approach Rodrik (2008) RER = E*PPPi/PPPj Using Indices RER is a measure of exchange rate missalignment. Note PPP refer to GDP price levels. But need to correct for the Balassa-Samuelson effect; in developing countries the PPP is relatively low in the nontradable sector Corrected by ln. RER = a + b(ln. Q/L) + error term ln. RERAdjusted = ln. RER – ln. RERPredicted = error term 136
Baseline model: (y-n) = a + bln(Y/N) + cln. RERAdjusted Panel data; 5 year periods; picking up cross-sectional effects only 137
The RER and Economic Growth An Alternative View The effect is significant for developing countries but not developed. Criticisms& Comments • Econometric (Spurious correlation -between (y-n) and ln. RERAdjusted in the regression caused by the method of estimating the latter. Other specification problems. (Woodford). • Not operating via import and exports where relative export and import prices are relevant. • Specification not compatible with export and import demand functions • x = f(ln. RER) x = f( rer, ln. RER) ln. X = f (ln. Z, ln(RPX)) • Lack of any specific model (tradbles versus non tradable sectors) 138
The RER and Economic Growth An Alternative View Nouria and Sekkat (2012) • More extensive equation to estimate fundamental misalignment • Other control variables in estimating equation “We did not find any support to the claim that a depreciated real exchange rate promotes economic growth”. “At a theoretical level the precise channels through which the effect might operate is unclear”. . . “At the empirical level, the evidence on which the claim is based raises a number of questions that cast doubt on the validity of the positive impact of undervaluation on economic growth”. 139
• Aguirre and Calderon (2005) Both large undervaluations and overvaluations hamper growth, • Haddad and Pancaro (2010) Limited effect confined to developing countries with per capita income below $2, 500. But in the long run it becomes negative and on exports insignificant. Real exchange rate volatility hampers growth. 1. 2. 3. 4. If an undervalued exchange rate is kept too long, excessive accumulation of low-yielding reserves. May lead to high inflation If it is the result of policy, it can constrain monetary policy from domestic targets Reduce incentives to develop a financial sector 140
5. Analogous to a subsidy to the tradable sector; tax on consumers 6. Maybe difficult to exit from this strategy. 141
Introducing the Supply Side 142
Introducing the Supply Side 143
Regional Problems are “Balance-of-Payments Problems” 144
An Example The UK North-South Divide Rowthorn (2010) • The North & the South of the UK (population 25 -30 million each) Structural Change since the Mid-1960 s • Deindustrialisation in both North and South but much greater in the North. • Financial and business services have grown much faster in the South. • North better in “public services”, but this was merely a catching up. • London a special case. Rapid growth in services but a large pool of unqualified workers from its industrial decline. 145
North-South Divide • The export base (tradables); constantly changing due to technical change, etc. • Prosperity depends upon export. Collapse of steel works immediate fall in income & employment short-run (negative) multiplier effect offset by greater welfare payments Long Term: Depends upon migration. Younger leave. Long-run multiplier; shrinking population and employment. The smaller export base can support the smaller population. But will the degree of migration be sufficient? 146
The Ratio of Employment in the North to the South Relative to the South 1. Rapid fall in tradables (red) Fallen by 30% in relative terms 2. Also non-tradables (green) 3. Offset by increase in public services (but only catching up) (blue) 4. Fall in relative population 147
Interregional Transfers (Public Expenditure per Capita) 148
Lifting the Balance-of-Payments Constraint: Policy Implications 1. Currency Devaluation 2. More Capital Inflows 3. Import Restrictions 4. Structural Change The Balance of Payments Does Not Look After Itself! 150
Where to Next and What is being Done? • The income elasticities are not immutable. What causes them to change? Estimation of export and import demand functions by technological sector. • Income elasticities are a crude proxy for non-price competitiveness; need a more explicit approach, other variables. ECB type approach • Need to move on from traditional econometric tests; case-studies (Akerloff). • Is there any role for changes in relative prices, the exchange rate and the distribution of income? Both theoretical and empirical models. • Gross rather than net capital flows. The role of capital flows in initiating a balance-of-payments crisis (IMF and the role of capital controls) • Reassessment of Rodrik (2008). 151
To Conclude. . On Paradigms 152
What is the most widely cited book in the whole of the Social Sciences? Clue is below 153
A little bit of methodology It is Kuhn’s Structure of Scientific Revolutions SSR (1962) • Why should we pay any attention to the philosophers of science? • SSR has more impact outside of philosophy that in it. Taken up by the sociologists (Bloor). • The concept of the paradigm or “disciplinary matrix” • How the concept changed in Kuhn’s later writings 154
The Paradigm • Theories are not rejected after a single or several refutations Duhem. Quine Thesis • “Normal Science” and the role of “puzzle solving”. • Puzzle determined by the paradigm – role of textbooks, demonstration not methodological rules – failure will be the fault of the scientist -. The role of research funding. • In Natural Science anomalies, previously shelved, become more and more difficult to ignore (role of the controlled experiment) • A new competing paradigm emerges; scientific revolution; make time but supplants the previous paradigm.
A Little bit of Methodology • Paradigmatic assumptions or heuristic – untestable by fiat – an article of faith. This is important for what I shall say in the rest of the lecture. • The switch from one paradigm to another involves a whole new way of thinking about a problem. • Kuhn at one time likened it to a “gestalt switch” taken from psychology. 156
A Gestalt Switch 157
The Duck-Rabbit Gestalt Switch 158
Economic Paradigms Classical economics ▼ Keynes’s Economics the GT ► Post Keynesian Economics ►? ? ▼ IS/LM Keynesian Economics ▼ AS/AD model; The Neoclassical Synthesis ► Neo-Keynesian Economics ▼ Monetarism ▼ ▼ ▼ New Classical Economics ▼ ▼ ▼ ………… The New Neoclassical Synthesis …………………. . ►? ? 159
160
161
162
Скачать презентацию Economic Growth and the Balance-of-Payments Constraint An
2c11fdf09f77acec45a550c753524a6a.ppt