Daniele Tavani and Luke Petach have a working paper out that has become one of my favorite readings for the past year on what they call the Capitalist coordination problem. A straightforward summary of the problem they point out is that, following Michal Kalecki, there is a contradiction between the interest of an individual capitalist and capitalists as a whole. The former wants to minimize labor costs, while the latter would benefit if workers, who are the majority of consumers, have more cash to spend on their products. Thus, in many circumstances, an advanced capitalist economy will wind up working below potential since it is more rational for individual capitalists to have a larger piece of a smaller pie than a smaller piece of a larger pie.
Daniele and Tavani sum up the central theme of this blog so far: the fact that a capitalist economy is incredibly dynamic and can also move toward stagnation if it is not somehow regulated toward an optimal growth path. This contradiction is a very timely problem because we are going through a freakout amongst large segments of the economy in the transition toward a more dynamic, rapidly expanding economy due to the return of consumer demand that has been low since the start of the "great recession." The consequence of this boost in consumption, especially of durable goods, is that we are observing a structural shift where systems of production aren't able to catch up to demand because their capacity to do so has been eroded for more than a decade.
Economics has had a tough time wrapping its mind around this problem since both the ne0classical and various heterodox traditions have paid very little attention to the issues of capital stock. The capital stock is a huge problem because it is very difficult to understand exactly what capital is. The whole Cambridge Capital Controversy was, at its simplest, a question of whether it is even possible to measure capital as some fungible quantity without first measuring the rate of profit. And if the latter shifts due to changes in a factor cost, then how do we understand its impact on capital stock?
The best answer to this problem that I have encountered is in the work of Adolph Lowe. Lowe was a German-American economist who once led at the Keil Institute for the World Economy. The Institute was a center for the inter-war study of the business cycle and was strongly influenced by the German Historical School and its focus on economics as the study of a changing structure of a "national economy" in historical time. The Keil School under Lowe was home to some future luminaries in such a structural approach including Jacob Marshak and Wassily Leontief. As a consequence of the Nazi seizure of power, Lowe and his colleagues left for the United States where Lowe found himself at the New School where he is best remembered for training Robert Heilbroner.
What interests us most is Lowe's 1971 work, The Path of Economic Growth. Lowe's approach to economics is a bit different from how we typically approach growth. Lowe believed that the problem of growth economics was not understanding the process of growth as an emergent system but rather studying growth as a prescriptive story about the most efficient path from one steady-state to another: what he called "force analysis." In other words, the economist's job isn't to theorize growth in abstract but rather to compare and implement the most efficient path from one collection of capital stock and labor to another. This process is what he called "the traverse." Lowe distinguishes between efficient and inefficient traverses where the former preserves some socially determined optimal mix of capital and labor and the latter is one where capacity and employment are destroyed as part of the resetting of fixed capital toward a new steady state. What Lowe is saying is that if we have a dedication to "full employment" we need some schema to figure out how to move from one steady-state to another without shouldering the burden onto workers. As he wrote, within "classical" capitalism resources could be put idle for long periods of time to facilitate such a shift in capital stock at the cost of vast unemployment:
To understand this we must recall one of the main characteristics of this era of industrial capitalism, namely, the large fluctuations in economic activity which, up to the Second World War, placed the study of business cycles at the center of analytical interest.
From the viewpoint of growth the recurrent, long-lasting stretches of
underutilization of resources had a paradoxical effect. A periodic drag on
output and income and, hence, on welfare generally, it was, nevertheless,
the very presence of large pools of idle resources that greatly facilitated
the system's adjustment to changes in aggregate demand and technology.
Expansion of individual sectors of production remained unaffected by the
state of capacity utilization and employment in other sectors, thus avoiding
the frictions and bottlenecks that impede intersectoral shifts of resources -
the unavoidable mode of adjustment under conditions of full utilization.
Thus, growth became a by-product of the cycle and hardly distinguishable
from the latter's phase of recovery.
However, if we are to try to avoid that misery, we have a problem. How do we create the conditions for the transition in the capital stock – of real machinery and resources – without incurring mass unemployment? Again he writes:
These conditions have drastically changed during the last generation.
Full employment has become the universally adopted aim of public policy
in the mature countries, so much so that, until quite recently, even rising
rates of inflation were rarely considered as too high a price for its achievement.
Although the policy of the "new economics" was nowhere completely
successful, where properly applied, it eliminated at least the large fluctuations of the past, and, thus, greatly reduced the degree of underutilization of resources and the duration of their idleness.
Here we encounter a reverse paradox. The greater the success of this
policy of stabilization, the smaller the flexibility of the system, and the
greater the difficulties in achieving a smooth expansion path. Growth can
then no longer feed on pools of idle capital and labor and, more than
ever before, becomes conditional on the shift of employed resources and,
especially, on capital formation. Furthermore, in the developing countries
the situation is not basically different. Though endowed with large reserves
of manpower, they sorely lack the real capital necessary for the productive
employment of their labor force.
If we go one step further we realize that the root of all these difficulties
is technological. Obstruction of resource shifts, bottlenecks in production,
inelasticity of supply owing to the longue duree of capital formation and
even more to the large costs of sunk capital, these and most other impediments to smooth expansion are the effect of the large size and the technical specificity of inputs. Consequently "fixed coefficients of production" dominate the adjustment processes and, in particular, the adjustment to the major growth stimuli - changes in population, resource supply, and technology.
Lowe is thus trying to solve the problem that confronts the world right now. How to move from one economic regime to another without suffering either excess unemployment or inflation.
Lowe's system of analysis rests on a tri-partite understanding of fixed capital. Starting like Marx, he divides the economy into two sectors: Sector I which produces fixed capital used by producers, and Sector II which produces the goods consumers use. However, he adds something new to the classical schema of production: the "machine tools" sector produces the tools that make other tools that in turn make the output of sector I and sector II. Lowe also adds a few other conditions for his model. First, he simplifies the schema to say that, in a steady-state, sector I and sector II are full-on value chains where each input from raw material, to the final product, is one process without any interference to another sector. He is then able to divide sector I into two sectors: Sector IA, which is the chain for producer goods, and sector Ib, which makes the inputs into sector II.
The flywheel in this model is the machine tools industry, which can be used in either sector Ia or Ib. Again, in a simplification, and machine tool can work as an input into producer goods or consumer goods. In equilibrium, this all works very nicely. However, if there is a change in the composition of the labor force, technology, or final demand in sector II, then there has to be some adjustment in the distribution of machine goods and the labor needed to operate them. So if there is more demand in sector II, like we are experiencing right now, there has to be first some movement of capital into sector Ib, which in turn might mean some shift of labor from sector Ia. But in turn, that can create bottlenecks in producer goods that are needed to make the output of sector Ib.
As we can see, the simplification of independent value chains in each sector in a steady-state break down and we are now on the traverse. Now, we have to deal either with a restriction of consumption, inflation in price, or unemployment in some subsectors until we get back to that ideal. Lowe wants his book to be a schema for understanding the costs of such transitions to society and how to choose the most efficient path with the smallest damage to workers.
Lowe is a bit of a "classical" economist in Keynesian terms. The adjustment mechanism he describes from the point of view of workers and consumers is not consumption but rather savings. This actually makes sense. While it is fair to criticize the schema for being economically unrealistic, in so far as saving is a residual of two variables – investment and consumption – it is simpler variable to target than consumption and, given that S = I, it helps us get to the steady state level of investment.
However, let's not be coy here. What increased savings means is decreased consumption. In Lowe's schema if there is a demand shock to the system in sector II – i.e. there is more demand for consumer goods – the supply will have to adjust by paradoxically restricting consumption through either voluntary or forced savings to allow for investment and thus capital formation in sector Ib.
Lowe's model is thus somewhat analogous to the classical Feld'man-Mahalanobis growth model with two sectors where maximum consumption can be satisfied by expanding the inputs of production goods into consumer goods industries. While good for explosive growth, the Feld'man-Mhalanobis model has a mixed track record since it does not do a good job handling foreign trade and taxation. Even worse, it does not have any mechanism for moving fixed capital from producing capital goods to consumer goods. Like Lowe, Feld'man-Mhalanobis require some burden to be placed on the consumption of the workers in the traverse. Not specifying the traverse is why the Feld'man-Mhalanobis has a mixed track record. The model doesn't do a good job telling us about how savings are formed via trade and taxation, nor does it tell us how and when to switch investment from capital goods to the production of consumer goods. That switch, in historical practice, has always been political.
Lowe's three sector model gives us more clarity on how to navigate the travers but still suffers some of the same problems as the Feld'man-Mhalanobis. It still depends on some kind of coordination mechanism that is either market driven, or political but the fact that there is an explicit "final demand" that drives it lets it plug into Keynesian macro-models to give us some insight into making these decisions. However, like Feld'man-Mhalanobis, there is a politics here that Lowe explicitly does not want to get into. The advantage of Lowe over Feld'man-Mhalanobis is that it is at least a starting point to reveal the coordination problem at the heart of modern economic growth.
This is where I want to come back to Daniele and Luke's paper. Because of the coordination problem at the heart of capitalism, we have a bias toward putting those costs on workers via unemployment and keeping the overall rate of growth below potential because of the low aggregate demand that implies. What is the alternative? Lowe believed these choices are less difficult in what he called "collectivist economists" – in other words, the planned economies of the Soviet world. However, Lowe leaves that aside by explaining that their decision system is completely political thus force analysis applied to them isn't the domain of the economist. This is, in my opinion, wrong and I will come back to that in my next blog post. However, like Lowe, let's put that aside.
Lowe's schema is calling for planning and economics associated with it that works in between the large macro-aggregates of Keynes and the micro-systems of the input-out economics advocated by Lowe's former colleague Wassily Leontief. Leontief and Lowe were thus on the same page for some need to develop state capacity to understand the state and evolution of the capital stock. That capacity, Leontief argued, would allow a preemptive strike on the bottlenecks in capital stock that prevented an efficient move across the economic traverse. In a 1974 op-ed in the New York Times, Leontief advocated for an "economic planning board" that could perform this function in future crises:
Steps taken now in this direction could not possibly lead to the solution of the present crisis. They might, however, keep the country from stumbling into the next crisis, and the next, each contributing one more turn to the inflationary spiral in which we have been caught.
Reading Leonteif with Lowe, however, brings up some important issues. Even if we can map our capital stock through input-output methods, what does that tell us about the intermediate level that Lowe is focusing on? We might know what the capital stock looks like but what is the efficient traverse from one steady-state in capital stock to another? More importantly, when we set up these traverses as alternatives, who gets to make the decision about the "efficient traverse?" The rub of Lowe's analysis and his general mission for economics as a discipline is that all economic theory is prescriptive. There is no "neutral" growth that is emergent out of thin air.
This means we need to rethink quite a few things about coordination under capitalism. The Hayek version of coordination assumes that independent actors will work like a giant computer coming to an optimal solution through the exchange of signals in a market. This is increasingly untenable as a position. With the climate crisis, the Covid pandemic, and the decade of stagnation in developed and developing economies, market coordination is increasingly hard to justify. However, then what is the answer to the coordination problem that Kaleckians like Daniele and Luke have identified? Lowe might give us some technical directions but he stays out of the political problem of who makes the choice? Once upon a time, we used to talk about "industrial democracy" as the worker's control of the workplace. This makes sense as a vision on the micro-level. In an analogy to economic tools, it is something like an input-output table. We also might have some inklings about democracy on the aggregate level through elections to legislative bodies.
However, the theory we are still a bit iffy on is that messo-level where Lowe operates. How do we work through the dynamic problems of economic change in some way that legitimizes and guides us across the traverse? The political success of neoliberalism is that it solved this problem by creating a constituency for its own policies. However, if neoliberalism is increasingly becoming abandoned, any order that comes after has to do the same work of creating and responding to constituencies in a democratic manner. This is actually a very timely issue since many advocates, including myself, are attempting to create an institutional solution to the crisis of supply chains by fostering investment in resiliency before they come under pressure. Balancing democratic accountability, administrative autonomy, and expert decision-making for such bodies will be a hard problem to solve. To be honest, I'm not sure what the solution to all those tradeoffs looks like, or even if there is one, but I am very happy people I work with in my day job are trying to get their hands around this issue.