This text is translated using UiO GPT-4.
Econometrics
Many would argue that The Probability Approach in Econometrics revolutionized econometrics as a discipline. In the publication, Haavelmo tries to convince economists to use probability models as the basis for testing economic theories.
Indeed, though rarely recognized as such, the Probability Approach is one of the masterpieces of twentieth century methodological writing in economics.
Aldrich J. (1989): Autonomy. Oxford Economic Papers, 41 (1989), 15 – 34.
Theory-building has little value if it is not possible to confront theories with observations. In many natural sciences, testing occurs through experiments where one factor can be varied at a time. An economist, on the other hand, must test theories based on observations in the economy, where a number of different factors influence the outcome. Therefore, the economist must try to extract the information he is looking for from composite observations.
Haavelmo showed that many of the methods that had been used previously gave very misleading results and derived criteria for how to reveal a particular relationship from observations that had emerged through an interplay of relationships.
The Concept of Autonomy
A central concept for Haavelmo was the concept of autonomy, which is about the degree of permanence between economic relationships. A variable can be expressed by other variables as a function. But this function is not worth much unless we specify the conditions under which the function is presumed to hold. Haavelmo thus believed that one should always search for relationships with as high a degree of autonomy as possible.
An autonomous relationship in a simultaneous model should not change content even if we remove other relationships in the model. Autonomous relationships do not break down even if there are major or minor changes in the structure surrounding the relationship. The now so well-known Lucas critique, that the Phillips curve estimated under one political regime would collapse when the political rules of the game changed, is a special case of Haavelmo's critique of basing policy decisions on relationships with a low degree of autonomy. Haavelmo believed that an econometrician's main task is to investigate whether it is possible, based on data, to quantify these autonomous relationships.
The Identification Problem
In The Probability Approach in Econometrics, Haavelmo clarifies the identification problem - one of the basic problems in quantifying economic relationships. Assume that a certain set of economic variables satisfies a system of equations, each with a specific degree of autonomy. We are interested in quantifying some of the parameters that are part of the system. If, based on the system of equations, we can derive an infinity of new systems of equations that have exactly the same form as the original system but with different values for the coefficients, we will never find unique estimates no matter how much data we have. Haavelmo showed that in a simultaneous equation system, it was only possible to identify the structural parameters if one imposed theoretical restrictions on the equations.
The Simultaneity Problem It is not satisfactory to try to determine each individual equation in a system separately, without considering the restrictions the other equations impose on the same variables. This can result in observations that are not expectation-consistent.
Sources
Biørn Erik, "Om Trygve Haavelmos bidrag til økonometri" Sosialøkonomen Nr 11, 1989 (6-7)
Development Economics
The causes behind the economic disparities between regions is the theme of Haavelmo's book A study in the Theory of Economic Evolution.
Although A study in the Theory of Economic Evolution was published as early as 1954, it is still described as surprisingly modern in its approach. In the book, Haavelmo anticipates a range of contemporary research areas, including the accumulation of human capital, analysis of strategic behavior between countries, and "rent-seeking". The simplest explanations for the differences between countries, which to some extent assume what is to be explained, are rejected by Haavelmo. For example, he believes that the differences cannot be explained by natural resources, skillfulness, or some being further ahead on the "development track". He wanted to create a model where the behavior relationships are approximately the same for all countries, but which would still explain the economic disparities.
His models consist of three parts:
- A macro production function where the total production of a country depends on labor, capital, and the general level of education and knowledge
- Accumulation equations where the growth of the capital stock and the growth in the level of knowledge are explained by economic variables
- A population "law" that explains how births and deaths depend on economic variables.
A series of model experiments are conducted in the book. The first is inspired by the classics (Smith, Ricardo, Malthus, and Marx). Along with other simple growth models, this model anticipates Solow's well-known textbook model. Haavelmo assumes diminishing returns to scale in the production function, which gives the model economic growth with a falling rate of profit. If population growth reacts strongly enough to incomes above the subsistence level, the working population remains at the subsistence level. If population growth is kept in check, real wages increase in step with production growth.
He also shows how small differences that mean little for welfare today can accumulate into significant disparities in the long term. The initial situation is given more attention than has become common in growth theory since then. He argues, among other things, that economies with poor starting points will always be on a lower growth path even if all else were equal.
Within the economic picture of the world that we can piece together from ongoing international statistics, perhaps economic inequalities are the most prominent feature. There are regions where millions of people starve, others where such experiences are long forgotten; there are regions where productivity is held back by the lack of the most obvious aids and technical improvements; there are regions where people wade in books and magazines, others where the illiterates are in the majority; there are parts of the world where animals receive better medical care than humans do in other areas.
Introduction to the book A study in the Theory of Economic Evolution.
Sources
Karl Ove Moene, Utvikling og Miljø - To eksempler på Haavelmos samfunnsøkonomiske engasjement, Sosialøkonomen 11 1989 side18 -19
Trygve m. Haavelmo, A study in the Theory of Economic Evolution. Amsterdam: North-Holland, 1954.
Investment Theory
As early as 1940, Haavelmo published several short articles within investment theory, but it wasn't until 1960 that his main work in the field, A Study in the Theory of Investment, was released. The book is an attack on the standard Keynesian demand for investments and an attempt to explain why net investments are so volatile and almost always positive.
There are no later studies of investment behavior that surpass Haavelmo's study in terms of philosophical depth or logical rigor. The empirical literature has mainly been based on the addition of various lags and the introduction of adjustment costs. Both of these developments were anticipated and analyzed in Haavelmo's book.
Førsund, 1989
Haavelmo envisioned a model with a profit-maximizing firm. Given a sequence of product prices, factor prices, and interest rates, there would exist an optimal capital level. This level is either equal to or different from the capital level the firm has from previous periods. Haavelmo envisioned that investments are the rate of change in the capital level per unit of time. In the initial period, the capital level will either remain unchanged or make a "jump" to a new level. Then, the rate of change per unit of time would be infinite. The investment demand at this point would therefore not be defined, and one cannot express investment demand as a continuous function of the interest rate.
A main point in Haavelmo’s investment theory was that from profit-maximizing behavior by firms, you can find out what volume of real capital the firms desire. If the desired volume of real capital is greater than the existing one, the firms want more real capital. Real investments are therefore defined as the desired real capital growth divided by how quickly the capital goods can be delivered. This delivery time is determined by factors such as capacity utilization in the capital goods-producing part of the economy, domestically or abroad. The desired real capital can be a function of the interest rate and thus investments can be as well. But, the delivery time is an amount determined by conditions that may not have anything to do with interest rates.
Critics (including Jorgenson 1967, Sandmo 1971) claimed that Haavelmo was wrong and that it was possible to derive a continuous demand function. A special case would be, for example, when the capital stock is in equilibrium and prices and interest rates change continuously. Instead of asking what happens to investments when the interest rate is raised ceteris paribus, Jorgenson asked what happens with investments when the interest rate increases along with future capital prices so that today's price on capital remains unchanged.
A Study in the Theory of Investment is a very comprehensive book and some have claimed that critics might have misunderstood some of the main arguments in the book. In the 1950s, Haavelmo held several lectures to clarify the subject, and these lecture notes could shed new light on the work.
Haavelmo's investment theory shares many similarities with Tobin's q-theory (1967), which is still used in textbooks today. The difference lies in that Tobin used a single-sector model where, among other things, the adjustment costs occur internally in the firm, while Haavelmo created a two-sector model where the adjustment costs are instead at a macro level outside the firm. Haavelmo himself believed that the model he created was closer to what Keynes actually meant in "The General Theory" than the standard Keynesian demand function. Le Roy (1983) argued that a two-sector model with sector-specific capital would have been even closer.
Sources
Moene, Karl Ove, Rødseth Asbjørn, Nobel Laureate Trygve Haavelmo, Journal of Economic Perspectives, 1991, 5, 3, 175-192
Førsund, Finn, “Om Haavelmos investeringsteori” Sosialøkonomen 1989, 43, 11.
Tobin, James, "A General Equilibrium Approach to Monetary Theory," Journal of Money, Credit and Banking 1969, 29, 40-47
Le Roy, Stephen S., "Keynes' Theory of Investment" History of Political Economy, 1983, 15, 397-421.
Haavelmo's Theorem
If authorities increase taxes and public purchases of goods and services by the same amount, balancing the public budget, the aggregate demand in the economy will increase. This result is commonly referred to as Haavelmo's theorem [1].
In today’s textbooks, the result is often explained by showing that public consumption has a full impact on aggregate demand, while a tax increase results in a smaller reduction in private consumption demand because part of the tax increase goes towards reduced savings [2].
In his article "Multiplier Effects of a Balanced Budget" from 1945, Haavelmo thoroughly goes through the reasons behind the theorem. Among other things, he argues against the idea that a balanced public consumption increase just creates increased demand because it limits private savings. Using a simple example, he demonstrates how increased public consumption, covered by an equivalent tax increase, boosts demand independently of the marginal propensity to consume.
Haavelmo, therefore, proposed a different explanation for the theorem; While private demand for goods and services depends on the consumers' income after taxes (net income), employment will depend on consumers' income before taxes (gross income). A balanced increase in public consumption will raise consumers' average gross income so that they have exactly the same net income as before the tax increase. Since the net income is unchanged, private demand will also be unchanged compared to the situation before the tax increase was implemented.
The authorities, therefore, collect tax money without any direct compensation to each taxpayer. At the same time, the authorities demand goods and services as payment for their nominal outlays. If labor and other production resources in the economy are idle, the government's production and employment will add to the demand from the private sector in the economy. The aggregate gross income (the nominal value of all goods and services in both the private and public sector) will increase even though the aggregate net income is unchanged [3].
Sources
T. Haavelmo, "Multiplier Effects of a Balanced Budget" Econometrca Vol 13. No. 4 (1945) pp. 311-318.
Footnotes
[1] Although several economists had already pointed out this result, Haavelmo was the first to formalize it, and therefore the result was named after him. See, among others:
- P.A. Samuelson, ”Full Employment After the War” in Postwar Economic Problems, editet by S. E. Harris, New
- York, 1944, pp 44
- A. H. Hansen and H. S. Perloff, State and Local Finance in the National Economy, New York, 1944, pp. 245-246
[2] S. Holden, ”Stabilisering i en enkel Keynes-modell del 1” Lectrue notes no. 4 2006
[3] For the result to hold, a closed economy must be considered as Haavelmo assumes.
Environment
Haavelmo was a pioneer in the field of environment and development. Although his works on the environment have only been published in Norwegian, they have inspired a number of recognized international publications (Hoel 1978, Førsund and Strøm 1988).
His main idea is that consumers' utility functions include both produced goods and variables representing the quality of the environment. While produced goods are represented as a flow quantity, the environment is affected by accumulated amounts of pollution from production and consumption (such as carbon emissions). Every consumer, therefore, faces a trade-off between the immediate pleasure of consuming more and the eternal distress that increased pollution entails. Here the free-rider problem is significantly important, since consumption is a private good, while pollution is a public bad. Unregulated development can, therefore, have very adverse consequences.
Sources
Moene, Karl Ove, Rødseth Asbjørn, Nobel Laureate Trygve Haavelmo, Journal of Economic Perspectives, 1991, 5, 3, 175-192.
Hoel, Michael "Resource Extraction and Recycling with Environmental Costs," Journal of Environmental Economics and Management, 1978, 5, 220-35.
Førsund ,Finn og Steinar Strøm, Environmental Economics and Management: Pollution and Natural Resources. London: Croom Helm, 1988.