Consumer Demand

Consumption is the largest component of GDP, and the consumption function is one of the key structural relationships in the Keynesian model. As noted, one of Stone’s earliest publications in 1938, with his first wife Winifred, was estimating the marginal propensity to consume and the multiplier [3]. Their work was a direct econometric estimate of the postulated stable relationship in three chapters of Book III of The General Theory. Keynes had made some back-of-the-envelope calculations, but the Stones sought to estimate the function and the multiplier for more countries over longer time periods. They fitted aggregate consumption functions with a graphical analysis using data on consumers’ expenditure and income with time trends for interwar periods for Germany (two periods), the UK, Sweden (two periods), the Netherlands, and the USA. They concluded that the functions were reasonably stable over time, but varied across countries. They calculated leakages with implied multipliers of between 1.7 and 4.9, depending on the openness of the economy [3: 15].

However, when Stone returned to modelling consumers’ behaviour after the war, it was to explore detailed spending for the UK and it was this work for which he is celebrated by the profession, for its meticulous attention to statistical detail, for its integration of theory and data, and for its use of econometric techniques. Two of his most cited works were in this area, published in 1954. The most cited with 155 citations over the period 1966-1984 (Garfield 1985: 473) is the article in the Economic Journal on the Linear Expenditure System [53], which is path-breaking in that it was derived from utility theory, imposing the critical theoretical constraints of adding-up, homogeneity, and symmetry algebraically and adopting a method of simultaneous equations to estimate the system. Note that the dependent variables are nominal expenditures, and one of the independent variables is total expenditure, so that the adding-up property is imposed by linear regression. The next most cited work (97 citations) was Stone’s book published in two parts in 1954 and 1966 [56, 123] (with Derek Rowe and others) called The Measurement of Consumers’ Expenditure and Behaviour in the United Kingdom, 1920-1938. Angus Deaton states of the first volume: ‘There is a masterly exposition of the theory of demand and of revealed preference, and there is a chapter on econometric methodology that reads like a text until one realizes that this is where the texts originated’ (Deaton 1987: 510).