Peter Flaschel is Professor Emeritus at Bielefeld University, Germany. He holds a PhD degree in Mathematics and a Habilitation degree in Economics. He has extensively published on classical economics and heterodox macrodynamical model building. He was on numerous occasions Visiting Professor at the University of Technology, Sydney, and was invited in 2006 as Theodor Heuss Professor to the New School for Social Research, New York. He received an Opus Magnum Grant from the Fritz Thyssen/Volkswagen Foundations in 2007–2008.
Your first major work was on ‘Riemannian Hilbert Manifolds’. Riemann and Hilbert are not economists but mathematicians. So therefore our first question would be: How did you become an economist?
Well, the – may I say so – leading guy in the residential community I was living in proposed to read Marx's Capital, Volume I. We started reading Marx in 1972, when I still was assistant at the University of Bonn in the department of mathematics. But Marx was so fascinating that I decided on the next holiday trip to Corsica to read also Volumes II and III of Capital. And Volume III was of course a bit of a surprise to me, because everybody was so focused on the labour theory of value as presented in Volume I. In Volume III I had to realize that labour values are not immediately regulating the exchange ratios of commodities.
To a certain degree I lost interest in mathematics at that time, though in particular Riemannian Geometry and its application to Hilbert Manifolds was a very beautiful theory for me. It also happened then that I told Werner Hildenbrand, Professor of Economics at the economics department in Bonn, about my interests in Marx. He responded, ‘Well, we have a much better theory of value’, and so we decided to have a joint seminar, reading Bródy on Marxian labour theory of value and Debreu's theory of value. Carl Christian von Weizsäcker participated in this seminar, as well as important members of the department of economics at Bonn University. It was a very interesting seminar, but at the end Egbert Dierker concluded that its aim was not really reached, because many of the participating mathematicians became interested in Marx and not so much in Debreu.
How did you move on in your professional career? Did you get involved more and more in applying your mathematical tools to economics?
First I had to learn more matrix algebra, which I had of course dealt with in my undergraduate studies, but not followed up. And I started writing on Marx, because the problem of joint production within the labour theory of value was becoming interesting to me. And funnily enough I formulated a solution where I disentangled joint production into equal value proportions. But Carl Christian von Weizsäcker in a private conversation on a trip from Bielefeld to Bonn suggested to me that I should not take equal proportions, but use relative sales values. I realized that this was a very good proposal and started writing a longer paper on the transformation problem in joint production economies.
What was next in your professional career?
First of all I tried to contact people in economics – I was still in Riemannian Geometry – people who could offer me an assistant position in economic theory. I went to Elmar Wolfstetter and later on to Malte Faber, and Faber brought me into contact with Klaus Jaeger, who was just moving to the Free University of Berlin. He was very open-minded and helped me to get the position of an assistant researcher at his institute in Berlin in 1975. My mathematical doctoral thesis was written in little more than a year, but the Habilitation thesis needed 5 years in Berlin to be completed. So the adjustment to economics was not an easy one.
And who were the economists who impressed you most, making this transition. You talked about Marx already and about the three volumes of Capital, but when we read your books or the titles of your books, it seems it was not only Marx you became interested in, but also Keynes and Schumpeter. Was this already in this period?
I think Klaus Jaeger basically hoped that I would skip Marx and do other more Keynesian type of research. But in fact my Habilitation thesis was on Marx, Sraffa and Leontief, and he was nevertheless very supportive in this respect. In addition, I had to teach macroeconomics in Berlin. Since I had not learned macro at all, this was a very hard time for me. I had a colleague, Michael Ambrosi, who was a definite Keynesian and we did many things together, and so Keynes became more and more of interest to me.
So then you got interested in synthesizing or bringing together Marxian and Keynesian elements to macroeconomics?
There were further colleagues in Berlin like Michael Krüger and Jörg Glombowski – and I think Michael Krüger brought the work of Goodwin to my attention, and therefore Marx and macro became the next objective on my agenda. The marriage between Keynes and Goodwin (Marx), in the form of demand-driven distributive cycles, was done much later on – but first of all I started from the supply side and the Goodwinian conflict about income distribution.
Then you finished your Berlin period with the Habilitation?
Yes, but I had the luck to get a 4-year temporary position in 1980 as professor there, where Klaus Jaeger was again very supportive. During that time I of course applied at other universities. Willi Semmler invited me for two terms, 1984 and 1985, to the New School for Social Research, New York, but then came a tenure offer from Bielefeld, which I accepted and where I have stayed until now.
Maybe you could explain your idea of a Marx–Keynes–Schumpeter (MKS) synthesis?
The concept of an MKS-system may sound very unattractive at first. So I would prefer to put it simply under the post-Keynesian umbrella and use only in brackets the term MKS as a proposal where post-Keynesian economics could go to. Very briefly, Marx is about labour productivity and the distribution of the product of labour. Keynes is about effective demand, not potential output, and about financial markets and interest. And Schumpeter is about product innovation, potential output–capital ratios and banking, because banking is the big push behind this type of innovation. So bringing the three researchers together just gives you six variables, so to speak, as a basis that you can make a theory of.
Only for clarification: Does the labour theory of value still play a role in that? There has been a huge discussion on the value of the labour theory of value.
Yes, and it can in principle be very easily solved, but I think the last 30 years were lost years in this respect. The generation of Duncan Foley, John Roemer and others really contributed to Marxian economics but what happened thereafter was very disappointing from a scientific point of view. There is another prominent contributor however. I already mentioned the von Weizsäcker proposal on the sales value method of firms, and when I read Richard Stone's 1968 System of National Accounts I discovered that he had labour values involved, when he measured labour productivity within this system. He even solved the problem of joint production in the same way as von Weizsäcker proposed it to me, calling it the industry technology assumption. So labour values are part of conventional input–output accounting – they are nothing fancy. We have prices, we have quantities, what we want to have from a Marxian perspective is a scientific language behind prices and quantities. And total labour costs or, in reciprocal form, labour productivity as in Stone, provides this language. You thereby look at the behaviour of the agents of the economy in terms of labour time embodied in the various commodities. The Law of Falling Labour Content is the most basic proven assertion of this approach.
What would you say that you – as an educated mathematician – have contributed to an MKS synthesis in terms of methods?
The macroeconomy is a high frequency economy. It changes every day, for example the price level changes every day, since each day a few prices will change. So basically you have to use continuous time dynamics in macro, and not the oversynchronized period models that are the fashion today. And if you use a continuous time approach, or a high frequency discrete time model, you also have much more powerful tools from mathematics at your disposal. So my claim first of all would be that applicable macroeconomics is continuous time macro, and then we should slowly expand this modelling framework into the area of delayed differential equation systems, as for example Kalecki did it.
To have the full picture it is also helpful to use the feedback structures that have been introduced by Keynes, Tobin, Fisher and others, and get from their perspective, so to speak, an ensemble of feedback channels, where we can work on partial aspects as well as integrated ones. This is the way I would proceed in macrodynamics.
Some heterodox economists hold a more or less antiformalist, antimathematical view of economics. So would you say that, if you use the right models, the right formalization, the right tools, mathematics is very valuable for economics and should be used, or what is the relationship between mathematics and economics or empirical facts?
Economics is about quantities and interdependence, and I think our brain is fairly limited in studying such issues in purely verbal ways. For example, if you have three types of financial assets and the labour market, and the goods market and maybe something else, you cannot think about this in detail without mathematics. Economics is a quantitative science, so it is already about numbers. Using numbers, but not mathematics, is not very plausible. Also, I would point out that already Marx studied mathematics. He wrote mathematical manuscripts and there you can see that he should have studied linear algebra first. But he studied analysis, and the miracles of the division of zero by zero, where you can see that he took such things very seriously.
Would you see any conflict between the use of the models you apply and the term ‘uncertainty’, which is a main topic among many post-Keynesians? As you mentioned in the beginning, the model you are using is from your point of view a post-Keynesian model, so is uncertainty covered in this model?
No. Not even stochastic elements are really covered yet. But stochastic elements are of course easy to introduce; you just add a stochastic process to the given deterministic set-up. Currently I would say, from the formal point of view, what can be done in post-Keynesian model building is ‘dynamic stochastic general disequilibrium’ (DSGD) modelling, because in continuous time, you cannot assume market clearing. No market clears every second. So you have everywhere gradual adjustment, it may be very fast, but it is always gradual. So DSGD would be the thing from the formal perspective I would pursue next, but uncertainty is still a very big issue that is left aside, as is product- but not so much process-innovation. Formalizing animal spirits as the human response to uncertainty has been done by Reiner Franke for example.
A couple of years ago you started with the idea of flexicurity as an attempt at taming capitalism, in the sense that we can get rid of unemployment as a disciplining device regarding wage demands or wage claims. Could you please explain this concept a little bit and how it has developed over the last couple of years?
Flexicurity as a concept is the combination of flexibility and security. In fact I read Schumpeter's Capitalism, Socialism and Democracy, and there he is saying in 1942 that socialism will not come from the East, in fact it will fail in the East, but it will come from the West and will be built on what the Vanderbilts, the Carnegies and the Rockefellers have established. This was very plausible to me. Well, I read it again – and it was no longer so plausible in the (basically Walrasian) setup he had chosen to model his competitive type of socialism.
But from thereon I thought, one has to use Western methods of production also under socialism, because they are the most advanced ones, to control the production of an enterprise and to lead an enterprise into the future. But one has of course to avoid the Marxian consequences for the labour market. So flexibility in the way enterprises are conducted is very important for evolution, but security for households and safe life course perspectives, in particular normal working days, no segmented labour markets and so on, are equally important.
I think, as the situation is today, that we indeed should, on the one hand, say yes to the productive forces of capitalism. But, on the other hand, the relations of production which govern them can be formed and can be given a shape so that the forces of production can remain capitalistic in nature, while we nevertheless arrive at much more than just the welfare state, which means safe life course perspectives, a very well-balanced education system, citizenship-education, and finally, the conduct of elites that is democratically-oriented and not, as in Bourdieu, habitus-based. I would call such a democracy-based social structure of capital accumulation ‘Social Capitalism’, which in fact only adds a few more letters to the word ‘Socialism’, but changes its essence significantly.
Would this change the structure of the firm, or is this only a new institutional framework for the firm?
The objective of the firm would still be profit-seeking, because I think the iPad would not have been invented otherwise. But I also think capitalism, to a certain degree, has always been criminal in nature and indeed has become criminal again after the prosperity phase after World War II. So you have to regulate capitalism, otherwise the criminals will dominate it.
You can produce spoiled meat and sell it on the market and securities in the financial markets are just spoiled meat. If you sell spoiled meat in the real markets, you go to prison. However, if you sell spoiled meat in the financial markets, you do not really face imprisonment, but indeed you should be sentenced to go to prison for that. Because these agents know that you are providing toxic assets to customers.
Consequently, capitalism always needs to be regulated in strong and well-reflected ways. I indeed believe that the profit-seeking motive can be regulated successfully in fairly strict ways, so that you can not only avoid criminal activities, but also the exploitation of labour power within the enterprise.
A final question: You are now retired as a professor in Germany. What is your view after most of your academic life in economics in Germany on the development of non-neoclassical economics, what is your recommendation to the younger generation, those who are still going for their PhD, and are still looking for jobs?
First of all, I would always state that my work is embedded in a relatively large group of old, intermediate and young economists. I never had assistants, because my position has not been one with such an endowment, but it happened in the last decade that I came into contact with a lot of young people.
I could name probably ten young researchers who are not just doing what we did, but who are going their own way by applying similar methods. So I see there is indeed a group of young researchers that will do very good work in the future in a post-Keynesian way. And I think there are also many others, whom I do not know yet, who will go into the same direction. Therefore, I expect very much from this generation which moreover is well ‘adapted’ to the ridiculous publish-or-perish mainstream straitjacket we are living in.
These people are not committed to mainstream economics – quite the opposite – but they are subdivided into doing just things that can be published in a mainstream journal and doing the work they find important. This combination of being forced to ‘howl with the wolves’ and doing post-Keynesian or Marxian work in addition is what I see is developing, and I find this very promising. And that is why I also think that Keynes in my view is a bit superior over Kalecki. Marx knew the political economy of his time in great detail and he wrote a critique of it. Keynes knew the mainstream of his time very well and his General Theory was an important step forward. Kalecki wrote a lot of insightful essays, but they are not really a critique of the mainstream, because he was not too much acquainted with it. Schumpeter, finally, started from Walras to formulate his theory of economic development. I therefore think that it is important to know the mainstream in detail, because then you may see clearer what you can do against it.
The interview was conducted by Eckhard Hein and Torsten Niechoj in October 2011.