1. Introduction
I am very grateful to Rod O'Donnell for his stimulating account of the contribution of John Maynard Keynes and for his charitable and useful comments on my chapter in my How Economics Forgot History (Hodgson, Reference Hodgson2001: chapter 15). I also note with appreciation that O'Donnell's article is intended to strengthen my overall argument rather than to weaken it.Footnote 1
I wish to stress some of our many areas of agreement. For example, O'Donnell (Reference O'Donnell2018: 1) wrote in his abstract that ‘deep compatibilities exist between the General Theory and institutionalist analysis … Keynes’ conception of a general theory is very different from that underpinning neoclassical economics’, and ‘Keynes is an ally, not an enemy, of institutionalism’. My agreement with this is affirmed here and now, just in case the reader might possibly get the impression that I have expressed disagreement with these sentiments. I have frequently stressed the links and affinities between Keynes and parts of the original institutional economics (Hodgson, Reference Hodgson and Pheby1989; Reference Hodgson and Setterfield1999; Reference Hodgson2004: 207, 285, 309–315, 385–386). O'Donnell (Reference O'Donnell2018: 2) also wrote of Keynes ‘it is erroneous to lump his thought in with neoclassical economics’. Again, I agree, and I make no such conflation. Personally, I regard myself as non-neoclassical and a Keynesian.
In the above examples and elsewhere, O'Donnell uses a powerful rhetorical device. He makes many statements in the form ‘X is untrue, but Y is valid’. In a more specific case: ‘Keynes does not say X, but he says Y’. This is all fair and above board, unless the reader is misled into thinking, in a critique aimed at me, that I affirm every case of the proposition X. I do not. There are so many of these cases that it would be tedious to go through all of them and make my position clear. I ask the reader to accept that I have not affirmed X in any case, unless textual evidence is provided to the contrary. In most cases, as in the examples in the preceding paragraph, O'Donnell does not provide textual evidence that I affirm X.
Sometimes O'Donnell skips the X and goes straight to the Y. Consider, for example, section 6 of his paper on ‘Keynes and institutions’, which begins as follows: ‘Although Keynes did not theorise in the general manner of institutionalist economists, he was nevertheless fully cognisant of, and engaged with, institutions throughout his entire career.’ In the General Theory, ‘institutions play explicit and essential roles, theoretically and practically’ (O'Donnell, Reference O'Donnell2018: 14).
As if we did not know this. Who would claim otherwise? Again, the reader may falsely conclude that someone had made the big mistake that Keynes wrote little about, and engaged minimally, with institutions. But the reader should not assume that I deny any Y, unless there is clear textual evidence to the contrary.
O'Donnell (Reference O'Donnell2018: 1) stated that his paper has a wider purpose than criticism of my position: ‘Although the primary focus concerns Hodgson and Keynes, the paper also touches on important broader issues’. This conflation of two projects may confuse some readers. I shall not comment here on the ‘broader issues’ he raises, other than to note that I am largely in agreement with what he writes on these.Footnote 2
O'Donnell's substantial criticisms are important, and I tackle them below. But first I explain that O'Donnell omits some key aspects of Keynes’ intellectual evolution. Keynes was detached from both the Marxist and the historical school traditions. These were largely responsible for the development of the critique of ahistorical economics and for efforts to establish historically sensitive modes of theorising. Keynes also failed to acknowledge some of the key contributions made by American institutionalism in preparing the ground for the Keynesian Revolution.
2. Keynes and the historical theorists
From the 1840s, the limits of general theorising in economics, the problem of dealing with the historical (and geographical) specificity of economic institutions, and the need for historically sensitive theory were explored by Karl Marx and members of the historical school in Germany, the British Isles and elsewhere (Hodgson, Reference Hodgson2001: chapters 3–9). These arguments had a major effect on economics in both the UK and the US, including on the original American institutionalists. But they were later forgotten.
Like many aspiring young economists in the 19th century, Alfred Marshall went to Germany to study under members of the historical school. He was in Dresden in 1868 and in Berlin in 1870–1871. Contrary to a widespread myth, Marshall was a supporter rather than an opponent of that school. He repeatedly praised and endorsed Gustav Schmoller (among others) in his writings (Hodgson, Reference Hodgson2001: chapter 8). In his 1885 inaugural lecture in Cambridge, Marshall (Reference Marshall and Pigou1885) clearly recognised the need for historically sensitive economic theory.
But after Marshall's retirement in 1908, economists at the University of Cambridge paid little attention to the German historical school. Arthur Pigou took over the sole chair of economics in a small department. Although he was a brilliant and well-published economist, he was then only 30 years old. Unlike Marshall and the former generation at Cambridge, Pigou was largely unfamiliar with the historical school.
Keynes was born over five years after Pigou and was similarly uninformed about the historical school. Apart from positive citations of Georg Knapp's (Reference Knapp1924) theory of money, there are few other references to the German historical school or to American institutionalism in Keynes’ works. In Keynes’ Collected Works as a whole, there are only two minor footnote references to works of Gustav Schmoller and no reference to Werner Sombart (who had one of the most developed theoretical frameworks for accommodating historical specificity).
There is also no evidence to suggest that Keynes was generally familiar with the works of leading American institutionalists. Keynes never referred in his writings to Thorstein Veblen. In 1925, Keynes (Reference Keynes1931: 303–304; Reference Keynes1981: 438) acknowledged John R. Commons as an ‘eminent American economist’ and expressed agreement with Commons’ (loose and questionable) idea that history had passed through ‘the era of scarcity’ followed by ‘the period of abundance’ in the 19th century and the ‘period of stabilisation’ (ironic in the light of the subsequent 1929 crash) in the 20th. But there is no indication that Keynes turned a single page of Commons’ key institutionalist works such as The Legal Foundations of Capitalism (Reference Commons1924) or Institutional Economics (Reference Commons1934). In particular, there is no evidence that Keynes was aware of Commons’ more sophisticated discussions of the need for historically sensitive theory.
Keynes (Reference Keynes1930) cited Wesley Mitchell's work on business cycles several times in his Treatise on Money. In 1934 he met this leading institutionalist in New York City and received an honorary doctorate at Columbia University, where Mitchell was a professor (Keynes, Reference Keynes1973: 456; Reference Keynes1982: 320). But Keynes failed to mention Mitchell in his General Theory. This omission was all the more glaring because Mitchell had helped to develop the statistical methodology of economic aggregates that played a crucial part in the development and application of Keynesian economics.
In the General Theory (Keynes, Reference Keynes1936) there are no references in the index to leading institutionalists or historicists such as Commons, Knapp, List, Mitchell, Schmoller, Sombart, Veblen or Weber. Schumpeter also receives no mention. As George Garvy (Reference Garvy1975: 393) pointed out, Keynes was a product of a relatively insular academic environment:
A simple answer to the question why General Theory was written in an intellectual vacuum can be given by referring to the well-known isolation of Cambridge economists within the proverbial insularity of Great Britain. … Keynes read little of what contemporary economists had to say … Nor did he dig deeply into the wealth of ideas and analyses of the preceding generations of economists. He largely ignored the contributions of his contemporaries (and the generations preceding them) who published in languages other than English.
A number of other authors have also noted ‘the Cambridge habit of focusing mainly on work by other Cambridge economists’ (Backhouse, Reference Backhouse1996: 35).
It is well known that Keynes was also unfamiliar with the economic writings of Marx. Here the language barrier was no excuse, as Capital and other major works had been translated into English. The insularity of his university was no good reason either, because alongside Keynes at Cambridge were Maurice Dobb and Joan Robinson, who both had studied Marx's economics. Consequently, Keynes seemed largely unaware of the ways in which in which both Marxist and historical school authors had periodised the development of economic systems, to enable an analytical focus on historically specific institutions.
The positive aspects of Keynes’ legacy need no elaboration here: Keynes was one of the greatest economists of the 20th century. But, following Pigou, Keynes’ economics bore little trace of the historical school, which on the contrary was important for Marshall. Similarly, Keynes did not acquire from Marx any schema of historical periodisation. The influence of the American institutionalists on Keynes was slight at best. In this way, Keynes, Pigou and other contemporaries did participate in what I call the Great Forgetting. O'Donnell denies this, but in doing so he overlooks the historical evidence from my How Economics Forgot History that I have repeated (in a highly condensed form) in this section.
3. Keynes and general theorising
O'Donnell (Reference O'Donnell2018: 5) proposes that Keynes intended his ‘general theory’ to be ‘a non-universal, historically specific theory of capitalism, and not a universal theory fundamental to all economies and agent behaviour regardless of time or space’. I agree that this implicitly resembles what Keynes ends up doing. But unfortunately, because of his ignorance of the relevant conceptual tools in Marxism, institutionalism or the historical school, Keynes is unable to formulate his objectives in these terms. He fails to make it sufficiently clear that his general theory applies to a historically specific socio-economic system such as capitalism, and not to all modes of production.
Many of the economists who read Keynes’ General Theory were also unfamiliar with Marxism, the historical school or the American institutionalists. Others might be influenced by the highly general, ahistorical theorising found among the neoclassical or Austrian schools. Keynes gave this significant mainstream readership inadequate warning that his theory, by contrast, was historically specific. Hence Keynes’ liberal use of the term ‘general theory’ played to the neoclassical audience, instead of forging links with the original institutionalism or the historical school. His rich theoretical exposition was the poorer as a result.
As I pointed out in my book (Hodgson, Reference Hodgson2001: 225) the prominent Keynesian Victoria Chick has argued forcefully that the General Theory was specific to a stage of capitalism where finance was well developed, with extensive credit markets (Chick, Reference Chick1986). Chick argued that the institutional realities of market economies without a developed financial system meant that saving necessarily preceded investment. But when banks were able to create large amounts of credit this was no longer the case – massive investment could occur by borrowing rather than saving. By the 20th century, banking institutions and the credit system had evolved to the point that investment could and often would precede saving. This was the quite specific historic period and type of system to which the General Theory applied.
In his article O'Donnell promises to provide evidence that Keynes intended his theory to be a general theory of capitalism rather than a universal theory apply to all possible types of economic system. This passage is part of his argument:
The overall institution and its component institutions are accepted from the start and form part of the ‘givens’ of Keynes’ theoretical system. They are not explicitly listed, but they are certainly present at the start, waiting off-stage until called upon to enter the analysis at the appropriate point in the determination of aggregate output. (O'Donnell, Reference O'Donnell2018: 5)
But surely if Keynes was fully aware of and versed in the problem of historical specificity then he would have declared at the outset that at least he was writing about a developed market economy with extensive use of prices and money. But while Keynes mentioned markets many times, he made no such historic delineation in the crucial opening pages, or even later in the volume. The abundant mentions by Keynes of markets are not evidence that he is carefully guiding his reader toward a theory that is understood as being specific to capitalism. Keynes mixes up universal and historically specific categories, without clear delineation of their contrasting historical standings.
Many important economists have made comparable omissions. A similar pattern is found, for example, in Adam Smith's (Reference Smith1776) Wealth of Nations. Smith started off with a discussion of the division of labour, which is a phenomenon found in all social species and in all human societies. But a few pages later he introduces markets, prices and money. These instead are historically specific, human phenomena, although they have existed for thousands of years (Hodgson, Reference Hodgson2015). Then even at this early stage in the book Smith introduced his famous example of the pin factory. Factory production is even more recent than markets, prices and money. It eventually becomes apparent that Smith was not developing a universal theory but writing about factory production and commerce in rising industrial capitalism. Without flag or fanfare, the narrative of the Wealth of Nations jumps from the sphere of human universals, then to market economies, and then to the Industrial Revolution that was beginning as he wrote.
For such reasons, Marx criticised previous economists. He wrote in 1846 of:
the error of bourgeois economists who regard … economic categories as eternal laws and not as historical laws which are laws only for a given historical development, a specific development of the productive forces. … politico-economic categories [are] abstractions of actual social relations that are transitory and historical … (Marx and Engels, Reference Marx and Engels1982: 100)
In his Poverty of Philosophy of 1847, Marx noted that previous economists wrote of ‘the relations of bourgeois production, the division of labour, credit, money, etc.’, but routinely deployed them ‘as fixed immutable, eternal categories’. Consequently, mention of de facto historically specific categories does not itself signal an understanding of the need for historically sensitive analysis. Previous economists ‘do not explain is how these relations themselves are produced, that is, the historical movement which gave them birth’ (Marx and Engels, Reference Marx and Engels1976: 162). Hence Marx criticised the classical economists for confusing and conflating historically specific forms with more general ones: they neglected their historical genesis and specificity.
Accordingly, the mere mention of de facto historically specific categories does itself not make an analysis sufficiently sensitive to historically specific conditions. As Marx, the historical school and original institutionalists such as Commons all complained, many classical and neoclassical economists were in the habit using de facto historically specific categories while falsely suggesting that they were true for all time.
I am not lumping Keynes together with the economists criticised by Marx, the historical school or the original institutionalists. Instead I am using Marx's argument to show that prominent mention of historically specific categories is not itself evidence of understanding of the need for a historically sensitive theoretical framework, which can tackle historically specific socio-economic formations.
The word capitalism appears only five times in the General Theory. Keynes never made it clear that the book was specifically and exclusively about capitalism. The word market is more plentiful. But markets are a necessary but not sufficient condition for capitalism, at least when the latter involves the extensive and powerful financial markets that Keynes discussed at length. If we follow Joseph Schumpeter (Reference Schumpeter1954: 78, 318–320) and define capitalism as involving dominant financial markets – including markets for debt, as well as private property and other markets – then capitalism has been prominent in Britain and elsewhere for little more than three hundred years. By contrast, markets and money became established in Greece and the Middle East over two thousand years ago, and they emerged even earlier in China (Hodgson, Reference Hodgson2015). Markets are much, much older than capitalism.
Piecing together statements from Keynes’ works from 1932 and 1933, O'Donnell (Reference O'Donnell2018: 14) informs us that Keynes:
distinguished three types of economies: (1) a barter, real exchange or cooperative economy (2) a neutral money economy, and (3) a money wage or entrepreneur economy (later called a monetary production economy). The first two institutional forms underpinned orthodox theory and the last his own.
Clearly, type (3) is the most specific of the three.
But the widespread existence of money wages and entrepreneurs is insufficient to constitute capitalism. Money (in various forms) has been in existence for several thousand years (Graeber, Reference Graeber2011). In England, rural labour for day wages was common after the fourteenth century (Postan, Reference Postan1972, Epstein, Reference Epstein1991). Hence money wages became prominent in medieval times. There are examples of entrepreneurship going back hundreds of years, the explorer Marco Polo, the artist-inventor Leonardo da Vinci and the printer Johannes Gutenberg being familiar examples.
Consequently, O'Donnell's depiction of Keynes’ taxonomy of different types of economic system is not sharp enough to identify capitalism, at least in the Schumpeterian sense of that word. On this point Keynes might have benefited from reading Schumpeter, who saw the extensive development of financial institutions as a crucial defining feature of that system.Footnote 3 Schumpeter (Reference Schumpeter1934: 126) saw the money markets as the ‘headquarters’ of capitalism: ‘capitalism is that form of private property economy in which innovations are carried out by means of borrowed money, which in general … implies credit creation’ (Schumpeter, Reference Schumpeter1939: 223).
Of course, Keynes wrote extensively about financial markets. That is important, and it is not an issue under dispute. Instead, the point here is that Keynes had no sufficiently developed taxonomy of types of economic system where some notion of (financial) capitalism clearly appears. The reason for this is simple, While Keynes was highly astute and far from ignorant of financial institutions, he was largely ignorant of previous attempts to taxonomise different types of economic system, such as within Marxism, institutionalism or the German historical school.
4. Conclusion
While O'Donnell is generous in his praise of my 2001 book, he gets several things wrong about my argument and my position. For example, contrary to O'Donnell (Reference O'Donnell2018: 2), I do not argue that ‘economic and social science theorising should avoid general theories claimed as true for all times and all places’. Instead, I wrote the following:
While the contemporary fixation on general theoretical systems is criticised, it is also argued here that some general principles are indispensable. (Hodgson, Reference Hodgson2001: xiv)
Emphatically, the argument in this book is not against the power or value of generalisation. While general theorising can never be enough, general statements are not only necessary but also unavoidable. … A more sophisticated position has to be found, recognising a significant role for general theories but also their limitations. (Hodgson, Reference Hodgson2001: 20)
The social sciences must thus combine general principles with theorising that is aimed at specific domains. (Hodgson, Reference Hodgson2001: 40)
I repeat these and similar points several times over in the book. Chapters 18 and 21 in the final part of the volume adopt some universal principles and consider multiple levels of abstraction, from complex systems in general down to human societies and capitalism as very particular economic forms.
My criticism of Keynes is not that not that he saw universal (as well as culturally specific) features of human nature. I have endorsed the possibility of long-evolved universal features myself (Hodgson, Reference Hodgson2013; Reference Hodgson2015: chapter 2). Neither is my criticism of Keynes that he adopted general concepts. My work on generalised Darwinism with Thorbjørn Knudsen show that we are also involved in the search for general frameworks of analysis (Hodgson and Knudsen, Reference Hodgson and Knudsen2010).
Instead, my criticism of Keynes is that he lacked the conceptual apparatus to clearly identify his theory as a general theory of capitalism. While it may be regarded by the reader as such, is nowhere stated clearly as an objective. While Marx (Reference Marx1976: 92) made clear that the aim of Capital was ‘to reveal the economic law of motion of modern society’, Keynes made no such historically targeted claim for his General Theory. Instead, the narrative of the work wanders up and down through different levels of historical abstraction, with inadequate markers of its altitude.
Partly, because all historically sensitive theorising must also rely on generalities, it is vital not to confuse different levels of theoretical abstraction. My ultimate criticism of Keynes is that he fails to be clear about this. If the General Theory was intended as an analysis of financial capitalism (which is the most reasonable and viable option) and not a universal and ahistorical theory, then at least its contrast with the growing fashion for purely ahistorical theorising – as pursued by Lionel Robbins (Reference Robbins1932) and many others – should have been clearly announced. Keynes failed to do this.
I agree entirely with O'Donnell that the General Theory is consistent with an institutionalist perspective. Indeed, its publication was greeted with enthusiasm by contemporary institutionalists, who found parallels in Veblen's work (Hodgson, Reference Hodgson2004: 207, 385). But, even today, much more remains to be done to develop Keynesianism along institutionalist lines. Of course, it is necessary to retain the most important and enduring parts of Keynes’ analysis of capitalism. But parts have to be rebuilt and the conceptual structure of a historically specific analysis has to be developed and added. Repeating the obvious – that Keynes frequently mentioned institutions and his theory was really about financial capitalism – is important. But it does not complete the unaccomplished task.
1. Introduction
I am very grateful to Rod O'Donnell for his stimulating account of the contribution of John Maynard Keynes and for his charitable and useful comments on my chapter in my How Economics Forgot History (Hodgson, Reference Hodgson2001: chapter 15). I also note with appreciation that O'Donnell's article is intended to strengthen my overall argument rather than to weaken it.Footnote 1
I wish to stress some of our many areas of agreement. For example, O'Donnell (Reference O'Donnell2018: 1) wrote in his abstract that ‘deep compatibilities exist between the General Theory and institutionalist analysis … Keynes’ conception of a general theory is very different from that underpinning neoclassical economics’, and ‘Keynes is an ally, not an enemy, of institutionalism’. My agreement with this is affirmed here and now, just in case the reader might possibly get the impression that I have expressed disagreement with these sentiments. I have frequently stressed the links and affinities between Keynes and parts of the original institutional economics (Hodgson, Reference Hodgson and Pheby1989; Reference Hodgson and Setterfield1999; Reference Hodgson2004: 207, 285, 309–315, 385–386). O'Donnell (Reference O'Donnell2018: 2) also wrote of Keynes ‘it is erroneous to lump his thought in with neoclassical economics’. Again, I agree, and I make no such conflation. Personally, I regard myself as non-neoclassical and a Keynesian.
In the above examples and elsewhere, O'Donnell uses a powerful rhetorical device. He makes many statements in the form ‘X is untrue, but Y is valid’. In a more specific case: ‘Keynes does not say X, but he says Y’. This is all fair and above board, unless the reader is misled into thinking, in a critique aimed at me, that I affirm every case of the proposition X. I do not. There are so many of these cases that it would be tedious to go through all of them and make my position clear. I ask the reader to accept that I have not affirmed X in any case, unless textual evidence is provided to the contrary. In most cases, as in the examples in the preceding paragraph, O'Donnell does not provide textual evidence that I affirm X.
Sometimes O'Donnell skips the X and goes straight to the Y. Consider, for example, section 6 of his paper on ‘Keynes and institutions’, which begins as follows: ‘Although Keynes did not theorise in the general manner of institutionalist economists, he was nevertheless fully cognisant of, and engaged with, institutions throughout his entire career.’ In the General Theory, ‘institutions play explicit and essential roles, theoretically and practically’ (O'Donnell, Reference O'Donnell2018: 14).
As if we did not know this. Who would claim otherwise? Again, the reader may falsely conclude that someone had made the big mistake that Keynes wrote little about, and engaged minimally, with institutions. But the reader should not assume that I deny any Y, unless there is clear textual evidence to the contrary.
O'Donnell (Reference O'Donnell2018: 1) stated that his paper has a wider purpose than criticism of my position: ‘Although the primary focus concerns Hodgson and Keynes, the paper also touches on important broader issues’. This conflation of two projects may confuse some readers. I shall not comment here on the ‘broader issues’ he raises, other than to note that I am largely in agreement with what he writes on these.Footnote 2
O'Donnell's substantial criticisms are important, and I tackle them below. But first I explain that O'Donnell omits some key aspects of Keynes’ intellectual evolution. Keynes was detached from both the Marxist and the historical school traditions. These were largely responsible for the development of the critique of ahistorical economics and for efforts to establish historically sensitive modes of theorising. Keynes also failed to acknowledge some of the key contributions made by American institutionalism in preparing the ground for the Keynesian Revolution.
2. Keynes and the historical theorists
From the 1840s, the limits of general theorising in economics, the problem of dealing with the historical (and geographical) specificity of economic institutions, and the need for historically sensitive theory were explored by Karl Marx and members of the historical school in Germany, the British Isles and elsewhere (Hodgson, Reference Hodgson2001: chapters 3–9). These arguments had a major effect on economics in both the UK and the US, including on the original American institutionalists. But they were later forgotten.
Like many aspiring young economists in the 19th century, Alfred Marshall went to Germany to study under members of the historical school. He was in Dresden in 1868 and in Berlin in 1870–1871. Contrary to a widespread myth, Marshall was a supporter rather than an opponent of that school. He repeatedly praised and endorsed Gustav Schmoller (among others) in his writings (Hodgson, Reference Hodgson2001: chapter 8). In his 1885 inaugural lecture in Cambridge, Marshall (Reference Marshall and Pigou1885) clearly recognised the need for historically sensitive economic theory.
But after Marshall's retirement in 1908, economists at the University of Cambridge paid little attention to the German historical school. Arthur Pigou took over the sole chair of economics in a small department. Although he was a brilliant and well-published economist, he was then only 30 years old. Unlike Marshall and the former generation at Cambridge, Pigou was largely unfamiliar with the historical school.
Keynes was born over five years after Pigou and was similarly uninformed about the historical school. Apart from positive citations of Georg Knapp's (Reference Knapp1924) theory of money, there are few other references to the German historical school or to American institutionalism in Keynes’ works. In Keynes’ Collected Works as a whole, there are only two minor footnote references to works of Gustav Schmoller and no reference to Werner Sombart (who had one of the most developed theoretical frameworks for accommodating historical specificity).
There is also no evidence to suggest that Keynes was generally familiar with the works of leading American institutionalists. Keynes never referred in his writings to Thorstein Veblen. In 1925, Keynes (Reference Keynes1931: 303–304; Reference Keynes1981: 438) acknowledged John R. Commons as an ‘eminent American economist’ and expressed agreement with Commons’ (loose and questionable) idea that history had passed through ‘the era of scarcity’ followed by ‘the period of abundance’ in the 19th century and the ‘period of stabilisation’ (ironic in the light of the subsequent 1929 crash) in the 20th. But there is no indication that Keynes turned a single page of Commons’ key institutionalist works such as The Legal Foundations of Capitalism (Reference Commons1924) or Institutional Economics (Reference Commons1934). In particular, there is no evidence that Keynes was aware of Commons’ more sophisticated discussions of the need for historically sensitive theory.
Keynes (Reference Keynes1930) cited Wesley Mitchell's work on business cycles several times in his Treatise on Money. In 1934 he met this leading institutionalist in New York City and received an honorary doctorate at Columbia University, where Mitchell was a professor (Keynes, Reference Keynes1973: 456; Reference Keynes1982: 320). But Keynes failed to mention Mitchell in his General Theory. This omission was all the more glaring because Mitchell had helped to develop the statistical methodology of economic aggregates that played a crucial part in the development and application of Keynesian economics.
In the General Theory (Keynes, Reference Keynes1936) there are no references in the index to leading institutionalists or historicists such as Commons, Knapp, List, Mitchell, Schmoller, Sombart, Veblen or Weber. Schumpeter also receives no mention. As George Garvy (Reference Garvy1975: 393) pointed out, Keynes was a product of a relatively insular academic environment:
A simple answer to the question why General Theory was written in an intellectual vacuum can be given by referring to the well-known isolation of Cambridge economists within the proverbial insularity of Great Britain. … Keynes read little of what contemporary economists had to say … Nor did he dig deeply into the wealth of ideas and analyses of the preceding generations of economists. He largely ignored the contributions of his contemporaries (and the generations preceding them) who published in languages other than English.
A number of other authors have also noted ‘the Cambridge habit of focusing mainly on work by other Cambridge economists’ (Backhouse, Reference Backhouse1996: 35).
It is well known that Keynes was also unfamiliar with the economic writings of Marx. Here the language barrier was no excuse, as Capital and other major works had been translated into English. The insularity of his university was no good reason either, because alongside Keynes at Cambridge were Maurice Dobb and Joan Robinson, who both had studied Marx's economics. Consequently, Keynes seemed largely unaware of the ways in which in which both Marxist and historical school authors had periodised the development of economic systems, to enable an analytical focus on historically specific institutions.
The positive aspects of Keynes’ legacy need no elaboration here: Keynes was one of the greatest economists of the 20th century. But, following Pigou, Keynes’ economics bore little trace of the historical school, which on the contrary was important for Marshall. Similarly, Keynes did not acquire from Marx any schema of historical periodisation. The influence of the American institutionalists on Keynes was slight at best. In this way, Keynes, Pigou and other contemporaries did participate in what I call the Great Forgetting. O'Donnell denies this, but in doing so he overlooks the historical evidence from my How Economics Forgot History that I have repeated (in a highly condensed form) in this section.
3. Keynes and general theorising
O'Donnell (Reference O'Donnell2018: 5) proposes that Keynes intended his ‘general theory’ to be ‘a non-universal, historically specific theory of capitalism, and not a universal theory fundamental to all economies and agent behaviour regardless of time or space’. I agree that this implicitly resembles what Keynes ends up doing. But unfortunately, because of his ignorance of the relevant conceptual tools in Marxism, institutionalism or the historical school, Keynes is unable to formulate his objectives in these terms. He fails to make it sufficiently clear that his general theory applies to a historically specific socio-economic system such as capitalism, and not to all modes of production.
Many of the economists who read Keynes’ General Theory were also unfamiliar with Marxism, the historical school or the American institutionalists. Others might be influenced by the highly general, ahistorical theorising found among the neoclassical or Austrian schools. Keynes gave this significant mainstream readership inadequate warning that his theory, by contrast, was historically specific. Hence Keynes’ liberal use of the term ‘general theory’ played to the neoclassical audience, instead of forging links with the original institutionalism or the historical school. His rich theoretical exposition was the poorer as a result.
As I pointed out in my book (Hodgson, Reference Hodgson2001: 225) the prominent Keynesian Victoria Chick has argued forcefully that the General Theory was specific to a stage of capitalism where finance was well developed, with extensive credit markets (Chick, Reference Chick1986). Chick argued that the institutional realities of market economies without a developed financial system meant that saving necessarily preceded investment. But when banks were able to create large amounts of credit this was no longer the case – massive investment could occur by borrowing rather than saving. By the 20th century, banking institutions and the credit system had evolved to the point that investment could and often would precede saving. This was the quite specific historic period and type of system to which the General Theory applied.
In his article O'Donnell promises to provide evidence that Keynes intended his theory to be a general theory of capitalism rather than a universal theory apply to all possible types of economic system. This passage is part of his argument:
The overall institution and its component institutions are accepted from the start and form part of the ‘givens’ of Keynes’ theoretical system. They are not explicitly listed, but they are certainly present at the start, waiting off-stage until called upon to enter the analysis at the appropriate point in the determination of aggregate output. (O'Donnell, Reference O'Donnell2018: 5)
But surely if Keynes was fully aware of and versed in the problem of historical specificity then he would have declared at the outset that at least he was writing about a developed market economy with extensive use of prices and money. But while Keynes mentioned markets many times, he made no such historic delineation in the crucial opening pages, or even later in the volume. The abundant mentions by Keynes of markets are not evidence that he is carefully guiding his reader toward a theory that is understood as being specific to capitalism. Keynes mixes up universal and historically specific categories, without clear delineation of their contrasting historical standings.
Many important economists have made comparable omissions. A similar pattern is found, for example, in Adam Smith's (Reference Smith1776) Wealth of Nations. Smith started off with a discussion of the division of labour, which is a phenomenon found in all social species and in all human societies. But a few pages later he introduces markets, prices and money. These instead are historically specific, human phenomena, although they have existed for thousands of years (Hodgson, Reference Hodgson2015). Then even at this early stage in the book Smith introduced his famous example of the pin factory. Factory production is even more recent than markets, prices and money. It eventually becomes apparent that Smith was not developing a universal theory but writing about factory production and commerce in rising industrial capitalism. Without flag or fanfare, the narrative of the Wealth of Nations jumps from the sphere of human universals, then to market economies, and then to the Industrial Revolution that was beginning as he wrote.
For such reasons, Marx criticised previous economists. He wrote in 1846 of:
the error of bourgeois economists who regard … economic categories as eternal laws and not as historical laws which are laws only for a given historical development, a specific development of the productive forces. … politico-economic categories [are] abstractions of actual social relations that are transitory and historical … (Marx and Engels, Reference Marx and Engels1982: 100)
In his Poverty of Philosophy of 1847, Marx noted that previous economists wrote of ‘the relations of bourgeois production, the division of labour, credit, money, etc.’, but routinely deployed them ‘as fixed immutable, eternal categories’. Consequently, mention of de facto historically specific categories does not itself signal an understanding of the need for historically sensitive analysis. Previous economists ‘do not explain is how these relations themselves are produced, that is, the historical movement which gave them birth’ (Marx and Engels, Reference Marx and Engels1976: 162). Hence Marx criticised the classical economists for confusing and conflating historically specific forms with more general ones: they neglected their historical genesis and specificity.
Accordingly, the mere mention of de facto historically specific categories does itself not make an analysis sufficiently sensitive to historically specific conditions. As Marx, the historical school and original institutionalists such as Commons all complained, many classical and neoclassical economists were in the habit using de facto historically specific categories while falsely suggesting that they were true for all time.
I am not lumping Keynes together with the economists criticised by Marx, the historical school or the original institutionalists. Instead I am using Marx's argument to show that prominent mention of historically specific categories is not itself evidence of understanding of the need for a historically sensitive theoretical framework, which can tackle historically specific socio-economic formations.
The word capitalism appears only five times in the General Theory. Keynes never made it clear that the book was specifically and exclusively about capitalism. The word market is more plentiful. But markets are a necessary but not sufficient condition for capitalism, at least when the latter involves the extensive and powerful financial markets that Keynes discussed at length. If we follow Joseph Schumpeter (Reference Schumpeter1954: 78, 318–320) and define capitalism as involving dominant financial markets – including markets for debt, as well as private property and other markets – then capitalism has been prominent in Britain and elsewhere for little more than three hundred years. By contrast, markets and money became established in Greece and the Middle East over two thousand years ago, and they emerged even earlier in China (Hodgson, Reference Hodgson2015). Markets are much, much older than capitalism.
Piecing together statements from Keynes’ works from 1932 and 1933, O'Donnell (Reference O'Donnell2018: 14) informs us that Keynes:
distinguished three types of economies: (1) a barter, real exchange or cooperative economy (2) a neutral money economy, and (3) a money wage or entrepreneur economy (later called a monetary production economy). The first two institutional forms underpinned orthodox theory and the last his own.
Clearly, type (3) is the most specific of the three.
But the widespread existence of money wages and entrepreneurs is insufficient to constitute capitalism. Money (in various forms) has been in existence for several thousand years (Graeber, Reference Graeber2011). In England, rural labour for day wages was common after the fourteenth century (Postan, Reference Postan1972, Epstein, Reference Epstein1991). Hence money wages became prominent in medieval times. There are examples of entrepreneurship going back hundreds of years, the explorer Marco Polo, the artist-inventor Leonardo da Vinci and the printer Johannes Gutenberg being familiar examples.
Consequently, O'Donnell's depiction of Keynes’ taxonomy of different types of economic system is not sharp enough to identify capitalism, at least in the Schumpeterian sense of that word. On this point Keynes might have benefited from reading Schumpeter, who saw the extensive development of financial institutions as a crucial defining feature of that system.Footnote 3 Schumpeter (Reference Schumpeter1934: 126) saw the money markets as the ‘headquarters’ of capitalism: ‘capitalism is that form of private property economy in which innovations are carried out by means of borrowed money, which in general … implies credit creation’ (Schumpeter, Reference Schumpeter1939: 223).
Of course, Keynes wrote extensively about financial markets. That is important, and it is not an issue under dispute. Instead, the point here is that Keynes had no sufficiently developed taxonomy of types of economic system where some notion of (financial) capitalism clearly appears. The reason for this is simple, While Keynes was highly astute and far from ignorant of financial institutions, he was largely ignorant of previous attempts to taxonomise different types of economic system, such as within Marxism, institutionalism or the German historical school.
4. Conclusion
While O'Donnell is generous in his praise of my 2001 book, he gets several things wrong about my argument and my position. For example, contrary to O'Donnell (Reference O'Donnell2018: 2), I do not argue that ‘economic and social science theorising should avoid general theories claimed as true for all times and all places’. Instead, I wrote the following:
While the contemporary fixation on general theoretical systems is criticised, it is also argued here that some general principles are indispensable. (Hodgson, Reference Hodgson2001: xiv)
Emphatically, the argument in this book is not against the power or value of generalisation. While general theorising can never be enough, general statements are not only necessary but also unavoidable. … A more sophisticated position has to be found, recognising a significant role for general theories but also their limitations. (Hodgson, Reference Hodgson2001: 20)
The social sciences must thus combine general principles with theorising that is aimed at specific domains. (Hodgson, Reference Hodgson2001: 40)
I repeat these and similar points several times over in the book. Chapters 18 and 21 in the final part of the volume adopt some universal principles and consider multiple levels of abstraction, from complex systems in general down to human societies and capitalism as very particular economic forms.
My criticism of Keynes is not that not that he saw universal (as well as culturally specific) features of human nature. I have endorsed the possibility of long-evolved universal features myself (Hodgson, Reference Hodgson2013; Reference Hodgson2015: chapter 2). Neither is my criticism of Keynes that he adopted general concepts. My work on generalised Darwinism with Thorbjørn Knudsen show that we are also involved in the search for general frameworks of analysis (Hodgson and Knudsen, Reference Hodgson and Knudsen2010).
Instead, my criticism of Keynes is that he lacked the conceptual apparatus to clearly identify his theory as a general theory of capitalism. While it may be regarded by the reader as such, is nowhere stated clearly as an objective. While Marx (Reference Marx1976: 92) made clear that the aim of Capital was ‘to reveal the economic law of motion of modern society’, Keynes made no such historically targeted claim for his General Theory. Instead, the narrative of the work wanders up and down through different levels of historical abstraction, with inadequate markers of its altitude.
Partly, because all historically sensitive theorising must also rely on generalities, it is vital not to confuse different levels of theoretical abstraction. My ultimate criticism of Keynes is that he fails to be clear about this. If the General Theory was intended as an analysis of financial capitalism (which is the most reasonable and viable option) and not a universal and ahistorical theory, then at least its contrast with the growing fashion for purely ahistorical theorising – as pursued by Lionel Robbins (Reference Robbins1932) and many others – should have been clearly announced. Keynes failed to do this.
I agree entirely with O'Donnell that the General Theory is consistent with an institutionalist perspective. Indeed, its publication was greeted with enthusiasm by contemporary institutionalists, who found parallels in Veblen's work (Hodgson, Reference Hodgson2004: 207, 385). But, even today, much more remains to be done to develop Keynesianism along institutionalist lines. Of course, it is necessary to retain the most important and enduring parts of Keynes’ analysis of capitalism. But parts have to be rebuilt and the conceptual structure of a historically specific analysis has to be developed and added. Repeating the obvious – that Keynes frequently mentioned institutions and his theory was really about financial capitalism – is important. But it does not complete the unaccomplished task.