1. Wage-led Growth: Concept, Theories and Policies Marc Lavoie In collaboration with Engelbert Stockhammer

2. Preliminary remarks The wage share has been falling in several countries over the last decades. There has been a polarization of incomes, even within wage and salary income. Average wages and average labour compensation have not kept up with productivity increases. Growth processes seem to have become more unbalanced. Export-led growth and finance-led growth regimes do not seem to be sustainable or stable, leading to global imbalances. A new wage-led growth strategy is needed. Jornada de Economia Politica, UNQ, Argentina, 2012

3. Wage-led economic strategy Tied to underconsumption theories (Malthus, Sismondi, Hobson) Tied to the Marxist profit realization problem (Kalecki, Baran and Sweezy, Steindl, Bhaduri) Linked with neo-Kaleckian models (Rowthorn, Taylor, Dutt) that emphasize demand and paradoxes Linked with the Kaldorian tradition, which emphasizes the fact that, within limits, demand can create its own supply Jornada de Economia Politica, UNQ, Argentina, 2012

4. ILO project: New perspectives on wages and economic growth: potentials of wage-led growth 1. Conceptual clarification (Lavoie, Stockhammer) 2. Why has the wage share been falling? (Stockhammer) 3. A mapping of wage-led and profit-led demand (Özlem Onaran, Giorgos Galanis) 4. A mapping of wage-led and profit-led supply (Storm, Naastepad) 5. The impact of income polarization (Van Treeck, Sturn) 6. The impact of financialization (Hein, Mundt) Jornada de Economia Politica, UNQ, Argentina, 2012

5. Outline A conceptual framework: policies versus regimes Demand regimes Productivity regimes Jornada de Economia Politica, UNQ, Argentina, 2012

6. A CONCEPTUAL FRAMEWORK Jornada de Economia Politica, UNQ, Argentina, 2012

7. The crucial distinction One has to distinguish between (1) the policies that are being pursued in a country to promote a certain kind of growth regime; (2) and the economic growth regime that this country is actually into, and hence how the economy will react to the policies being put forward. (1) Distributional policies are either pro-labour or pro- capital; (2) Actual growth regimes are wage-led or profit-led. Jornada de Economia Politica, UNQ, Argentina, 2012

8. Pro-labour and pro-capital distributional policies Jornada de Economia Politica, UNQ, Argentina, 2012 Distributional policiesOther factors Pro-capitalPro-labour Policies ‘Labour market flexibility’ Abolish minimum wages Weaken collective bargaining, rights to strike Impose wage moderation ‘Welfare state’ Increase minimum wages Strengthen collective bargaining Changes in technology Globalisation Financialization Fiscal policies Monetary policies Results Weak wage growth Wage share ↓ Increased wage dispersion Rising real wages Stable (or ↑) wage share Decreased wage dispersion

9. Definition of profit-led and wage-led economic regimes Jornada de Economia Politica, UNQ, Argentina, 2012 Overall impact on the economy FavourableUnfavourable Income distribution change imposed on society An increase in the profit share Profit-led regime Wage-led regime An increase in the wage share Wage-led regime Profit-led regime

10. Viability of growth regimes Jornada de Economia Politica, UNQ, Argentina, 2012 Distributional policies Pro-capitalPro-labour Economic regime Profit-led Profit-led growth process Stagnation or unstable growth Wage-ledStagnation or unstable growth Wage-led growth process

11. Actual growth strategies Jornada de Economia Politica, UNQ, Argentina, 2012 Distributional policies and strategies Pro-capitalPro-labour Economic regime Profit-led ‘Trickle-down Neoliberalism’ – Supply-side policies will generate aggregate demand ‘Doomed social reforms’ TINA Wage-led‘Neoliberalism in practice’ – Unstable and has to rely on exogenous growth drivers (credit- led growth, export-led growth) Postwar social Keynesianism Golden age

12. DEMAND REGIMES Jornada de Economia Politica, UNQ, Argentina, 2012

13. q q full capacity UC MC p q normal p Profit per unit Preliminary micro issue: no profit-maximizing constraint; sales constraint instead Jornada de Economia Politica, UNQ, Argentina, 2012

14. Demand regimes, Y=C+I+NX+G An increase in the Wage Share leads toWS ↑ Domestic demand effects –Effect on consumption (s p > s w )C ↑ –Effect on investment I (↑) ↓ Effect on net exportsNX ↓ Total demand effectY ↓↑ Various studies indicate that the propensity to save out of profits is about 0.40 above the propensity to save out of wages Jornada de Economia Politica, UNQ, Argentina, 2012

15. q q ma I S0S0 S 00 = I 0 I, S q0q0 IaIa qmqm S1S1 E Effect of an increase in the wage share in the canonical Kaleckian model (multiplier and accelerator) Jornada de Economia Politica, UNQ, Argentina, 2012 S 01 F G

16. q q ma I0I0 S0S0 I0I0 I, S q0q0 IaIa qmqm S1S1 I2I2 I1I1 Effects of an increase in the wage share in the post-Kaleckian model (profitability effect)

17. Effects of an increase in the wage share and demand regimes Jornada de Economia Politica, UNQ, Argentina, 2012 Effect on total demand (or the rate of capacity utilization) PositiveNegative Effect on investment (or the rate of accumulation) Positive Wage-led demand and wage-led investment NegativeWage-led demand and profit-led investment Profit-led demand and profit-led investment

18. A note on the empirical work assessing demand regimes It should be pointed out that usually in empirical studies of the demand regimes, the accelerator effect is not taken into account. The researcher only looks at the short-run multiplier and profitability effects. This means that, by construction, as long as we suppose that the profitability effect is negative or nil, an increase in the wage share will have a negative effect on the investment component of aggregate demand. Thus, in those studies, by construction, we cannot have a wage- led investment demand. As to the effect on net exports, these depend on unit labor costs and hence will normally react negatively to a higher wage share. Jornada de Economia Politica, UNQ, Argentina, 2012

19. Effects of an increase in the wage share and domestic and total demand regimes Jornada de Economia Politica, UNQ, Argentina, 2012 Effect on total aggregate demand, including net exports PositiveNegative Effect on domestic aggregate demand (investment and consumption only) Positive Wage-led domestic demand regime and wage-led total demand regime Wage-led domestic demand regime and profit-led total demand regime NegativeProfit-led domestic demand regime and profit-led total demand regime

20. Economic structure: wage-led and profit-led demand regimes Jornada de Economia Politica, UNQ, Argentina, 2012 Demand regime Profit-ledWage-led Economic structure Small differentials in propensities to consume Propensity out of wages is much higher than the propensity out of profits Investment is highly sensitive to profitability and accelerator parameter is low Investment is not sensitive to profitability and accelerator parameter is high Very open economy with high net export price elasticity Relatively closed economy with low net export price elasticity

21. Past empirical results on demand regimes Jornada de Economia Politica, UNQ, Argentina, 2012 Domestic DemandTotal Demand wage-ledProfit-ledwage-ledProfit-led Euro areaSOE09, OG12 Germany BB95, NS07, HV08, SHG11, SS11, OG12 NS07, HV08, SHG11, OG12 BB95 France BB95, NS07, ES07, HV08, SS11, OG12 (SO04), NS07, HV08, OG12 BB95, SE07 NLNS07, SS11HV08NS07HV08 AustriaSE08, HV08, SS11 SE08, HV08 UK BB95, NS07, HV08 OG12 SS11 BB95, NS07, HV08, OG12 JapanBB95, OG12NS07OG12BB95, NS07 USABB95, HV08, OSG12, (SS11), OG12 NS07BB95, HV08, OSG12, OG12 (SO04), NS07, BFT06

22. A summary of these empirical results Most countries seem to be on a wage-led domestic demand regime. Some of these countries turn to a profit-led total demand regime when international trade is taken into account. Most studies conclude that the USA are in a wage- led domestic demand regime; – and the latest studies even conclude that the USA are in a wage-led total demand regime. Jornada de Economia Politica, UNQ, Argentina, 2012

23. New empirical results; Onaran (2012) on most G20 countries (80% of world GDP) All countries have wage-led domestic demand. Most developed countries have wage-led total demand, including the overall eurozone, except staple-led countries Canada and Australia. Large developing countries (Argentina, Mexico, China, India, South Africa) have profit-led total demand, except Turkey and Korea. In Argentina, a 1% point increase in the wage share leads to an increase of 0.20 % increase in domestic demand but a 0.07% decrease in total demand (GDP). Jornada de Economia Politica, UNQ, Argentina, 2012

24. An error of composition While a country may be under a profit-led demand regime when looking at the total effect of an increase in the wage share, a simultaneous increase in the wage share of all countries may still have a positive effect on the aggregate demand of a profit-led country if its domestic demand is wage-led. Onaran (2012) performs this experiment. She finds that most countries or regions, including Argentina, benefit from a simultaneous increase in the wage share, with the exception of Australia and China. She also finds that a 1% point increase in the wage share of all countries leads to a 0.36% increase in world GDP. This means that some individual countries can successfully pursue beggar-thy-neighbour policies via wage moderation, but this does not constitute a viable strategy for demand on a global scale. Jornada de Economia Politica, UNQ, Argentina, 2012

25. SUPPLY EFFECTS: PRODUCTIVITY REGIMES Jornada de Economia Politica, UNQ, Argentina, 2012

26. Economic structure: wage-led and profit-led productivity regimes Jornada de Economia Politica, UNQ, Argentina, 2012 Direct or Partial Productivity regime Economic structure Profit led Wage restraint leads to productivity-enhancing investment Higher real wage growth leads to slower productivity growth Wage led Wage growth has strong positive effects on labour effort and on productivity–enhancing investments, in an attempt to to recover profitability Higher real wage growth leads to faster productivity growth (efficiency wage hypothesis, Webb effect)

27. Interaction between productivity and demand There is a lot of empirical evidence showing that faster overall growth, and faster growth in manufacturing, leads to faster productivity growth. This is the so-called Kaldor-Verdoorn effect. Thus, an increase in real wages or the wage share, besides its direct effect on productivity, will have an additional indirect effect on productivity, through the impact of the wage share on aggregate demand, and hence, through the Kaldor-Verdoorn effect, on productivity growth. The strenght of the demand regime will also be affected by the feedback effects of the productivity regime Jornada de Economia Politica, UNQ, Argentina, 2012

28. Productivity growth pwpw Profit-led Productivity x0x0 Output growth p0p0p Kaldor-Verdoorn relationship E Direct impact on productivity growth of an increase in wage share or in growth rate of real wages Wage-led productivity

29. Combining the productivity and the demand regimes x = a 0 + a 1 (w – p) the demand equation p = b 0 + b 1.x + b 2.w the productivity equation dx/dw = a 1 (1 - b 2 ) /(1 + a 1 b 1 ) dp/dw = (b 2 + a 1 b 1 ) /(1 + a 1 b 1 ) With the combination of a wage-led demand regime and a wage-led productivity regime (all coefficients positive), the positive effects of increases in real wages will be enhanced for productivity, but diminished for demand Jornada de Economia Politica, UNQ, Argentina, 2012

30. Productivity growth p x0x0 Output growth p0p0 ptpt Kaldor-Verdoorn relationship Direct and indirect impact of growth rate of real wages with wage-led regimes Wage-led demand xdxd xtxt D P DP

31. The impact of an increase in real wages combining the productivity and the demand regimes x = a 0 + a 1 (w – p) the demand equation p = b 0 + b 1.x + b 2.w the productivity equation dx/dw = a 1 (1 - b 2 ) /(1 + a 1 b 1 ) =.19 dp/dw = (b 2 + a 1 b 1 ) /(1 + a 1 b 1 ) =.35 With a 1 = b 1 = b 2 = 0.30 And hence the change in the growth rate of employment is: dx/dw - dp/dw =.19 -.35 = -.16 !!!!! The impact on employment growth is negative! Jornada de Economia Politica, UNQ, Argentina, 2012

32. A case of ommitted variable? Perhaps there is an omitted variable. In the demand equation, instead of: x = a 0 + a 1 (w – p) perhaps we should have: x = a 0 + a 1 (w – p) + a 2.p meaning that high productivity growth (technical progress) induces faster capital accumulation and hence faster output growth. In this case, the negative impact on employment would be dampened. Jornada de Economia Politica, UNQ, Argentina, 2012

33. Conclusion: The real world is not so simple Several countries wish to pursue an export-led policy, restraining wages to gain a competitive advantage. But planet earth is a closed economy. All countries cannot be net exporters. Thus what really counts are the effects of an increase in the wage share on domestic aggregate demand. Empirical studies show that most countries are in a wage-led domestic demand regime. In such countries, rising real wages will generate high rates of productivity growth. Thus, it is possible that a wage-led growth strategy might slow down employment growth (technological unemployment). This strategy requires other expansionary policies. Jornada de Economia Politica, UNQ, Argentina, 2012