Robert M. Solow was born on 23 August 1924 in Brooklyn, New York. He was the oldest of three children of parents who were children of immigrants.

A scholarship allowed him to enrol in Harvard College from 1940 and he became, as he formulated it many years later, “curious about what made society tick”. After two years and at only 18 years of age he left the college and became a soldier for a three-year period, which he spent mainly in Italy.

After the end of the war he studied at Harvard University and came in contact with Wassily Leontief who had a lasting influence on him and who hired him as a research assistant. Solow’s thesis was about changes in the size distribution of labour income and won him a prize. At Columbia University he enhanced his statistical and econometrical abilities, and in the early 1950s he became an assistant professor (later full professor) at the Massachusetts Institute of Technology (MIT), where he stayed for his whole academic career. There he developed a close friendship and productive collaboration with Paul A. Samuelson, who had the office next to him. In 1995 he retired but continued to do research, lecture and participate in public discussions - often with handy formulations on complicated issues.

Solow received a number of important awards and prizes. In 1961 he was awarded the John Bates Clark Medal, which goes annually to the best economist under 40 years of age. In 1987 he was awarded the Nobel Memorial Prize in Economics and in 1999 he received the National Medal of Science of the United States. The most important award, the Nobel Memorial Prize of the Sveriges Riksbank was awarded to him “for his contributions to the theory of economic growth” (see especially Solow 1956, 1957, 1970).

In contrast to the founders of modern growth theory, Roy F. Harrod and Evsey Domar, Solow was not interested in the process of how equality between investment (I) and saving (S) is achieved. In his opinion this was a short-run problem not to be dealt with when long-run questions such as economic growth are on the agenda.

He assumed therefore that I = S is always fulfilled, which amounts to adopting Say’s law. However, he went further. Postulating a linear homogeneous macroeconomic production function with substitutive factors of production and perfect factor markets with flexible prices, he excluded problems such as unemployment and under-utilization of capital. Since the price mechanism is assumed to take care that all factors of production are fully used, the growth of real income in this world is completely determined by the growth of the factors of production. The decisive task of growth theory therefore is to investigate the determinants of the growth of the factors of production.

In Solow’s basic model, without technical progress there are only two factors of production, labour and capital. As far as the labour force is concerned, he assumed that it is a constant share of population and that it increases with a constant rate, which is given from the outside. Net saving (identical with net investment or increase of the capital stock) is assumed to be proportional to income. It is quite obvious that three different situations can be distinguished: (1) capital and labour grow with an identical rate - the situation of steady-state growth - (2) capital grows with a higher rate than labour, and (3) labour grows with a higher rate than capital.

The first situation implies that real income grows at the same rate as labour and

capital. We are in the situation of a steady state, in which capital intensity and per capita income are constant.

When capital increases more swiftly than labour, the assumptions of substitutability and of constant returns to scale imply that the growth rate of real income is smaller than capital’s growth rate but larger than labour’s. We are in a situation in which capital intensity as well as the capital coefficient increase. However, the growth rate of capital is

where K is the capital stock, s is the propensity to save and Y is the social product.

Since in the situation under consideration the denominator on the right-hand side increases, it is obvious that the rate of accumulation (UK) will fall. In other words: as long as the rate of accumulation exceeds the (constant) growth rate of the labour force, it is declining. This decline takes place until the growth rate of capital is equal to the growth rate of the labour force.

When labour grows at a faster rate than capital, we have the reverse situation: capital intensity and the capital coefficient decrease until the growth rate of capital has become equal to that of labour.

A remarkable feature of the model is the following: an increase (decrease) in the saving or investment rate can only temporarily but not permanently influence society’s rate of growth. This can again be clarified with the equation above. A higher s will increase the accumulation rate UK and at the same time the growth rate of income. But the latter will be lower than the former as the elasticities of production of substitutive production functions have values between zero and one. As a consequence the capital coefficient increases and that will decrease UK, which finally will again coincide with the exogenously give rate of population growth. Only during the process of transition does a higher investment rate have a positive effect on the rate of growth.

This result challenged the conventional wisdom that an increase in the rate of investment is decisive for economic growth, which had instructed development policy.

The model in its basic form has an irritating feature because growth of per capita income is only a transitional phenomenon. However, statistics tell us that highly industrialized countries have long been experiencing growing per capita incomes. Therefore, an important ingredient explaining per capita growth is missing. Some economists spoke of this missing ingredient of a “third factor”, some of technical change or technical progress. In an empirical investigation (Solow 1957), which supplemented his theoretical contribution (Solow 1956), Solow showed that capital intensification had contributed only a small part to the growth of per capita income, whereas the predominant part had to be attributed to “technical progress”, which was not taken into account in the basic model.

He stressed that this is a catch-all for the totality of influences we are not able to identify exactly; Moses Abramovitz (1956: 11) spoke of a “measure of our ignorance”.

Technical change, as introduced by Solow, raises several questions. The first concerns the special form of it. As can be shown, the only form of technical progress compatible with steady-state growth is so-called labour-augmenting or Harrod-neutral progress which in a macro-economic production function has the form (β stands for the rate of efficiency growth of labour):

Solow does not discuss the question of why this form of technical change should dominate and only a few later contributions were concerned with this question.

A further feature of Solow’s technical change is that it is disembodied, that is, it occurs without the necessity of introducing new capital goods. In a later paper (Solow et al. 1966) he discussed the consequences of embodied technical change in which gross investment is necessary to introduce new capital goods (so-called vintage models).

Finally it has to be mentioned that technical change in Solow’s model is completely exogenous, it is simply a function of time. During the revival of growth theory (“neo-neoclassical growth theory”) in the second part of the 1980s, attempts were made to endogenize technical progress, often using Solow’s model as the point of departure.

Solow also made important contributions to other fields of economics. To mention but a few: (1) in Dorfman et al. (1958) the authors offered not only a rigorous introduction to the then new field of linear programming but showed at the same time that important messages of marginal analysis were not restricted to “well-behaved” twice differentiable functions but can also be formulated in the framework of linear equations subject to constraints as they are characteristic of linear programming approaches; (2) together with Samuelson (Samuelson and Solow 1960) he transformed the Phillips curve into a relation between the rate of unemployment and the rate of inflation, which seemed to offer politicians a choice between different options; (3) in McDonald and Solow (1981), taking up an idea of Leontief, the solution of the so-called monopoly union model (the union first chooses the wage and then the firm chooses employment subject to this wage) is shown to be Pareto inefficient and can be improved when the firm and the union negotiate an efficient contract; (4) in Solow (1990) he showed that the conventional market model has to be modified with respect to the labour market and he demonstrated why unemployed workers in general do not even try to undercut wages, so applying a game theoretic approach shows that it may be rational for workers to behave in that way, and (5) Hahn and Solow (1998), dissatisfied with the ruling representative-agent models in macroeconomics, tried to put macroeconomics on another track, with little success, as Solow himself admitted.

In contrast to the latter publication, Solow’s contributions had a tremendous effect on economics and economic policy. This holds true first and foremost for his contributions to the theory of economic growth. They triggered innumerable theoretical and empirical studies and had, beyond that, a significant influence on economic policy because they challenged the previous view that an increase in the rate of investment of physical capital is crucial for economic growth. (However, things have come full circle: in new growth theory much of the attention focuses on the share of investment.) Although in Solow’s model technical change was exogenous, it suggested that it would be beneficial for society to allocate resources to research and development, on the one hand, and to

an enlargement of human capital on the other. With the burgeoning neo-neoclassical growth models these messages became a run-of-the-mill wisdom.

Peter Kalmbach

Robert M. Solow was born on 23 August 1924 in Brooklyn, New York. He was the oldest of three children of parents who were children of immigrants.

More on the topic Robert M. Solow was born on 23 August 1924 in Brooklyn, New York. He was the oldest of three children of parents who were children of immigrants.: