Introduction:

The analysis of the relationship between personal income distribution and growth has experienced a sharp rise in academic attention since the early nineties, when endogenous growth models allowed integrating new determinants of growth in macroeconomic models. Although there still is some disagreement, several main channels have been identified through which income distribution affects growth. The central ones are credit-market imperfections, fiscal policy and socio-political instability. The main theoretical result is that high inequality negatively influences growth, although some inequality may be needed at early stages of development or from a behaviouristic point of view (as suggested by statusmodels like Corneo/Jeanne 2001).

Another important link between income distribution and growth, which has received less attention, is the role of demand structure. When households differ in their incomes and wealth endowments, they develop different patterns of demand. So the aggregate structure of demand may change from one country to another, or between stages of development of one country, due to a different distribution of income. The composition of aggregate demand heavily influences the supply side by creating or enlarging markets for desired goods and shrinking those for (relatively) undesired ones. This in turn affects factor incomes because of the different factor input requirements for the production of more or less sophisticated goods. These shifts in functional income distribution lead in subsequent periods to changes in personal income distribution, so that previous interpersonal differences may be perpetuated, enlarged or reduced.

Studies of demand factors are traditionally underrepresented in neoclassical economics. In the case of the inequality-growth relationship, numerous historical experiences and actual economic developments suggest that the problem of aggregate demand and of demand composition may be crucial. The different paths adopted by the US-American and the British economy during the 19th century is an often cited historical example. While in the U.S. the internal demand was mainly constituted by a large middle class of relative prosperous farmers, in Great Britain it was still relatively sharp divided into a working class and an aristocratic upper class. The American farmers demanded simple industrial goods in large amounts, whereas the British upper class preferred primarily expensive handcrafted products. This different demand conditions are regarded as one of the main causes of the astonishing take-off of U.S. mass-production industries during the 19th century and of their overtaking of the British ones. Different ?destinies? due to diverging demand conditions were even experienced by several important developing countries (Brazil, Mexico, South Corea, Taiwan, Singapore, Hong Kong, Nigeria, to some extent the Philippines) in the last decades of the 20th century, especially during the Sixties and Seventies. This period was characterized by high growth rates for all of them, but in some countries it was accompanied by growing, in others by decreasing inequality. Only in those countries, where the latter was the case (South Corea, Taiwan, Hong Kong, Singapore), this high-growth period led to a general and stable development process. In all the others, growth not only increased social differences, but also slowed down in the subsequent years, without providing the desired ?big push? that could lead to a stable developing path. In recent times, the lack of internal demand is regarded as one of the main causes of low growth in Germany during the last years, when the country has experienced a period of high unemployment and (like many other industrialized countries) of growing inequality. It?s been argued that this lack is not ascribable to some forms of ?precautionary saving?, but that it is caused by the level and the distribution of income (Rees 2004, Rebeggiani 2005a).

These historical examples are taken as the starting point in this paper. Its aim is to develop a theoretical framework to study the inequality-growthrelationship through the demand composition channel, after providing an overview over the existing literature (neoclassical and Keynesian). In the second part, the theoretical findings shall be verified using a large data set including several waves of the German household survey ?Einkommens- und Verbrauchsstichprobe? (EVS) together with some aggregate data.