Composition of public expenditure, effective demand, distribution and growth

We introduce public expenditure (PE) in a general post Keynesian framework characterized by a non-linear investment function. Our aims are:1) to provide a systematic analysis of the effects of PE (‘productive' or ‘non productive') on economic growth within a post Keynesian framework in which effective demand plays a crucial role. Our work fills a lacuna in the post Keynesian literature given that scant attention has been devoted to this topic (an exception is Pressman 1995). Concerning Kaleckian models, the few formalisations involve the effects of PE on growth only via government debt (You and Dutt 1996) or deficits (Lavoie, 2000). As it is well known, the analysis of the effects of PE within Endogenous Growth models started with the seminal work of Barro (1990) and it has been successively developed in detail within that approach. Following Barro (1990) ‘Productive' PE is such since it enters in the production function affecting average factors productivity. More recently, Devarajan et al (1996) question Barro's approach. Following these authors PE is ‘productive' whenever it affects (positively) the rate of growth. In our paper, ‘Productive' Public Expenditure affects the (fixed) coefficients of production/average factors productivity similarly to Barro. We should note that in our analysis, due to the so-called ‘paradox of costs' (defined later), not necessarily an increase in the average factor productivity (following an increase in ‘Productive' PE) implies an increase in the rate of growth;2) to compare and contrast two post Keynesian approaches to endogenous growth, which stress the role of demand differently. The standard Keleckian approach is characterised by an autonomous (linear) investment function according to which investment is driven by i) (exogenous) long run expectations about the growth rate of demand; ii) expected profitability (proxied by the current rate of profit); iii) the expected level of demand (proxied by the current degree of capacity utilisation). The basic result of this approach is that in equilibrium effective demand drives growth. An important corollary is the so-called ‘paradox of thrift' according to which a reduction (an increase) in the average propensity to save of the economy, rising (reducing) the level of demand and the degree of capacity utilisation affects positively (negatively) capital accumulation and the equilibrium rate of growth of the economy. Introducing PE financed by income taxation and a balanced budged constraint, this effect could be induced by increasing (lowering) the tax rate. Another important corollary is the so-called ‘paradox of costs' according to which an increase (reduction) in the mark up and the redistribution in favour of profits (wages) reducing (increasing) the level of demand and the degree of capacity utilisation could dampen capital accumulation. With PE this effect could be induced by changes in productivity. In the Classical or Harrodian approach (as elaborated, among others, by Duménil and Lévy, 1999), long-run expectations are linked to saving decisions. The equilibrium rate of growth corresponds to the Harrodian ‘warranted rate of growth'.3) to reproduce a variety of complex phenomena (multiple equilibria, hysteresis, low growth traps, regular and irregular growth cycles), not uncommon in the real world, by introducing a simple nonlinearity in the investment function in the spirit of Kalecki's (1937) investment theory – based on the principle of increasing risk –, Kaldor's (1940) trade cycle model and Goodwin's (1951) non-linear accelerator. The specific shape of the investment function, often employed in the literature (see for example Chang and Smith, 1971; Varian, 1979; and more recently Gardini et al., 2006) is such that capital accumulation increases with the current degree of capacity along to a ‘S' shaped curve.