WAGE-LED REGIME, PROFIT-LED REGIME AND CYCLES: A MODEL

We propose a dynamic model which deals with the impact of income distribution variations on growth. In that goal, we use two models: the classical Goodwin model (1967) and the Bhaduri-Marglin model (1990), which also focuses on the links between income distribution and growth, but in a Keynesian frame. We introduce Keynesian demand constraints within the Goodwin model and modify its investment function, which becomes non-linear. With these new hypotheses, we show that Goodwin cy-cles may either be maintained or disappear. If most trajectories oscillate around a classical equilibrium, the economy may also fall during a cycle into a Keynesian un-employment state. In that case, cycle dynamic is broken because wages are squeezed whereas the economy is in a wage-led regime. This model allows to capture some speci c characteristics of the French economic situation that took place in the 1980s-1990s. The impact of income distribution on growth has been studied by two di¤er-ent strands of economic literature, belonging respectively to the classical and post-Keynesian traditions. On the classical side, Goodwin (1967) proposes a formaliza-tion of endogenous cycles where investment is directly related to saving. Because the propensity to save of entrepreneurs is higher than that of workers, an increase (respectively a decrease) in the pro\u85t share always boosts (respectively hinders) ac-cumulation and growth. The latter assumption has however been questioned by some Keynesian economists. According to Bhaduri and Marglin (1990), the e¤ect on growth of a rise in pro\u85ts remains uncertain: it may either be positive (pro\u85t-led regime) or negative (wage-led regime). Nevertheless, the Bhaduri-Marglin model remains incomplete as it does not tackle dynamic issues. CES Matisse, Université Paris 1. MSE 106-112, boulevard de lHôpital 75013 Paris. France