Roger Backhouse On Robbins, Keynes, Samuelson, & The Relationship Between Economics And The Other Social Sciences

Roger Backhouse is Emeritus Professor of Economics at the University of Birmingham.

By Aiden Singh, September 8, 2023

One feature of much post-war interdisciplinary research was that the unifying discipline was widely perceived to be psychology. The common feature in social science problems was the human factor, which placed psychology at the centre. The result was that psychologists were involved in many social science projects. In contrast, from the 1960s onwards, economics came to be seen as the central discipline, with rational choice/optimization theory - the “economic approach” to human behaviour (Gary Becker’s phrase) - being a shared framework. Perhaps the pendulum is swinging back towards psychology with the rise of behavioural economics and other economists finding that psychological ideas have insights to offer.

Question: Let’s turn to the impact of specific economists. In his influential 1932 An Essay on the Nature and Significance of Economic Science, Lionel Robbins articulated what is referred to as the scarcity definition of economics. You've written about how this definition of economics came to be widely accepted, how it influenced the scope of the discipline, and how it influenced the interaction between economics and the other social sciences. Could you discuss the impact of this 1932 essay on economics and its relationship with neighboring disciplines?

Answer: From around the 1930s, economists increasingly used mathematics in both theoretical and empirical work, talking about what they were doing as modeling. One way to think of this is that this was a way of recognizing that what they were doing was abstracting from many of the complications found in the real world. There were many reasons for this development, which pre-dated Robbins's essay. Indeed, Robbins quite correctly claimed that his essay was not original but articulated ideas that had a long history, albeit ideas that were perhaps less universally accepted than he claimed. In particular many American institutionalists - adherents to the movement that began after the First World War and sought to make economics more scientific by grounding it more firmly in evidence - rejected the scarcity definition in favour of definitions such as that economics dealt with the business system. One could also trace the rise in mathematical economics to broader cultural and intellectual changes in relation to science and engineering in the twentieth century. For example, two of the main centres for mathematical economics after the Second World War were the Massachusetts Institute of Technology and Stanford University, institutions where engineering had been transformed by being linked to what one might call pure science. The ethos of these institutions provided for a setting that was congenial to the new approach to economics.

So any causal connection with Robbins's essay is hard to establish. What the essay did was to articulate an approach to economics that rationalized the new approach to economic theory. It justified theorizing based on assumptions that were believe to be true, freeing economists from the need to provide careful empirical support for those assumptions. It was also read as severely limiting the scope for welfare economics through arguing that scientific economics could not rest on value judgments. Of course, although some economists embraced the narrow, anti-empirical conception of science outlined in the essay, many did not. Robbins was dismissive of econometrics - the use of formal statistical methods, such as correlation, to establish economic propositions - yet econometrics thrived, especially once computing had developed sufficiently to make calculations less time-consuming. And there were many economists who justified economic theories on the Popperian ground that they were potentially falsifiable statements that had not yet been falsified.

When it comes to the other social sciences, again causation is hard to establish. As Steven Medema and I have shown, at least in the academic journal literature, economists increasingly embraced Robbins's definition of economics in the 1960s, at the same time as other social sciences were taking up arguments based on rational choice and optimization.

Question: Proceeding chronologically with our list of economists, let's move on to Keynes. Your research has suggested that, during the 1930s, there were two competing conceptions of market failure articulated by economists attempting to explain the Great Depression. What were those two competing conceptions of market failure and how do they relate to Keynes' ideas?  

Answer: The paper in question is specifically about the United States. There, in the 1920s and 1930s, there was a belief that the success of the American economy was because it was dynamic and competitive: the land of opportunity. Belief in competitive markets and in free enterprise has a significant place in American history even if the country also has a long and influential protectionist tradition, and even if there was at times a blindness to the fact that certain groups (such as people of color, Jews, or women) were excluded from many of those opportunities. There was also a long tradition going back at least to the late nineteenth century, the age of the so-called Robber Barons such as John D. Rockefeller and Andrew Carnegie, of scepticism, if not outright hostility, to big business.

When the economy collapsed in the Great Depression these two strands of thought came together to offer an explanation. There was a perception that competition had been reduced (indeed, some of the early New Deal programs had sought to reduce competition in order to maintain prices) and that this meant that the dynamism of the economy had been reduced. This was one conception of market failure: that competition was no longer working as it had in previous decades.

The second conception of market failure, which came closer to Keynes, though it was being developed before his most influential book was published, was the notion that it was the “financial machine” that had broken down. The financial machine is the mechanism that connects savers with those who want to invest. Simplifying the process, savers deposit their money in banks, which then lend it to businesses wanting to invest. The result was a failure in the circular flow of income.

In the late 1930s there was a massive government inquiry into these problems, and in it you can see a shift in emphasis away from the first of these explanations of the Depression to the second. I don’t want to go into detail, but this is very much an American story, and it is distinctive from what happened in, say, Britain, the home of Keynes, where the notion of competition as a feature of a dynamic, expanding society was never so deeply rooted.

Question: Let’s move now to the development of modern economics in the second half of the 20th century. One of the key intellectual figures here is Paul Samuelson whom you've called the 'founder of modern economics' and 'the people's economist.'  Who was Paul Samuelson and why do you consider him the founder of modern economics? How does his work relate to that of Keynes?

First, a caveat to your question. Publishers and their editors play a role in deciding on titles, and both of these titles are ones where an editor played a role. Having said that, I do not want to repudiate either title completely, though I would certainly want to claim merely that Samuelson was one of the founders of modern economics. He was far from the only one, though he was a key figure because he was involved in so many branches of the subject.

After an undergraduate education in one of the main centres of economics in the 1930s, the University of Chicago, he then did graduate work at Harvard, in the course of which, as a Junior Fellow, he turned out a string of journal articles that proved foundational for much of what was done after the war. One part of this story is mathematical economics where, as I tried to explain earlier, changes took place for reasons that go far beyond Samuelson. Samuelson had the knack of using mathematics to simplify arguments, some of which were very confused, providing a place from which others could start. Because he wrote in such a way that it was not necessary to grapple with the previous literature, he can appear as a founder rather than someone who systematized and simplified. Whether anything significant was lost in this process is a different question that I won’t go into. In the book that brought together much of this work, Foundations of Economic Analysis (1947) one of the unifying themes was that apparently diverse problems had a common mathematical structure, involving optimization.

The other part of the story is his involvement in what has come to be known as macroeconomics - the analysis of the economy as a whole - and this is where Keynes comes in. In contrast with his initial work in mathematical economics, which is primarily a pre-Second World War story, this is mainly a wartime story. He began to think about saving and invesment in 1937-8, and in 1939 published two articles in which he translated a model formulated by Alvin Hansen, a key figure in arguing that the Depression was the result of a breakdown of the financial machine, into mathematics. However, under Hansen’s tutelage, he became involved in extensive empirical work on consumption and saving in order to understand what might happen after the war. Early in the war he worked in one government agency and later, after nine months working as a mathematician on engineering projects, he acted as a consultant to another. This work involved data and lots of calculations, but there was no “high theory” involved. It is arguably the background to the other book that cemented his reputation, Economics: An Introductory Analysis (1948).

This book was important because it more or less swept the board in the market for introductory economics textbooks. National income analysis, on which he had worked during the war, and which had become entwined with Keynesian ideas, was central: it was the organizing theme of the book. In the book he displayed a talent for writing about economics in a way that people who were innocent of advanced mathematics could understand. So for many students, Keynesian economics was what they encountered in Samuelson’s textbook; it is likely that few of them read what Keynes himself had written.

In Samuelson’s subsequent career, right up to his death in 2009, we find these two sides to his work. There was as the mathematical theorist whose work provided a starting point for theorizing in numerous applied fields: consumer behaviour, production, international trade, the public sector (taxation and government spending), finance, and welfare economics (by which I mean the analysis of how well off people are, not the provision of support to those in need). In this role he could be called the economist’s economist, writing articles that, to a lay audience, would seem little more than abstruse mathematics, but which were considered insightful to those who understood them. The other was as the political economist (political economy was his term for this side of his work) who discussed current events and explained them to a wide audience. This includes his textbook, which sold millions of copies, shaping the market for such works, and thousands of articles in newspapers and popular magazines, the most famous being those in the column he wrote every third week for Newsweek, between 1966 and 1981. Less well known, at least outside the countries involved, are the hundreds of columns he wrote for outlets in Japan, Korea, and other countries.

Question: How did Samuelson's work affect the relationship between economics and the other social sciences?

This is a question I find hard to answer. I suspect the answer is very little. He did engage with people from other disciplines, but his own work did not draw on psychology, sociology, or other social sciences. In a sense his work made it easier for economists to do their work in ignorance of what was happening in other social science but I would not want to emphasize this because economists’ cautious relationship with other social sciences has deeper roots. I would rather say that his own attitudes reflected widely held attitudes among economists. Indeed, he was arguably more open to approaches different from his own - more open to non-mathematical work - than were many economists.

Question: So we've seen how the work of Robbins, Keynes, and Samuelson influenced economics and you've discussed how their work affected the relationship between economics and its neighboring disciplines. Turning now to your own views on the subject, what do you think the relationship between economics and the other social sciences should be? 

Answer: That is a question on which I am going to pass. As a historian, I see my role as being to understand what is going on, not to pronounce on what is good and what is bad. In any case, I suspect that the answer will depend very much on circumstances: on institutional arrangements, on how the social sciences are structured, on the state of knowledge at any time, and on the problems facing society that need to be solved. Take the biggest issue facing us today (at least the one I see as the biggest), the breakdown of the climate. Clearly, to work out policies we need to pay attention to the economic dimension, and we need to pay attention to psychology, helping people to see the problem in a way that motivates them to take the action that is necessary. Economists and psychologists (and no doubt other social scientists) need to be involved. But I would not wish to pronounce on the nature of the relationship.

Question: For our final question, let's step back and look broadly at the subject of today's interview - the history of economic thought. How much of a role do you believe the history of economic thought should have in economic research and teaching? Do you believe it should be taught/researched in economics departments, or left to historians?

These are more difficult questions. Start with the second. History of economic thought, or history of economics as many prefer to call it, is history, not economics; so logic suggests it should be located in history. However, it requires understanding of economics, which can be acquired only through close engagements with economics. Some historians have done that very successfully (I think of Peter Clarke, a historian who has written outstanding books on Keynes, as an excellent example), but it is easy for people who have never seen economics from the inside to misunderstand what is going on. So there are reasons why economics departments should contain historians of the subject. Note that many economics departments already contain statisticians, and few people question this, so there is no problem in principle with the idea that economics departments should contain people who are not economists. One might argue that, because disciplinary boundaries are to a certain extent artificial, we should have social science departments in which cross-fertilization between disciplines can occur more easily than in departmental silos. But experience suggests that such departments, which in many institutions would be very large indeed, do not always work. There are pragmatic as well as intellectual reasons for the way fields are institutionalized. There needs to be some common interest to hold people together.

In my experience, there are generally students who want to study the history of the subject as part of their training in economics, and for such students I think it is important that they can do so. It broadens their education and, hopefully, makes them better economists. However, I would not make this a universal rule. There is immense pressure on the graduate curriculum, and I can fully understand the reasons why the history of the subject, along with economic history, was progressively squeezed out. If someone wants to focus purely on the technical side of economics, then studying history may be a waste of time for them, even if it gives other students an opportunity to broaden their horizons.

The same goes for research. There are economists who believe that knowing something about the history of the subject is useful. For example, Paul Krugman once wrote a piece titled something like, “How to be a crazy economist.” His advice was that the history of economic thought was a source of ideas. However, not all economists think this, and for others, the history of the subject, even if they find it interesting, is of no practical use. It is an undesirable feature of modern academic life, with its auditing of research, that it can be uncomfortable for someone to be doing research that does not fit within established institutional/disciplinary structures. That is undesirable, whatever disciplines are involved, but it is a fact of life. One can, of course, debate whether modern academic institutions are optimal, but that would take us into a completely different set of issues that I am not prepared to get into because they go way beyond the questions you have asked.