The wages of precarity, the costs of globalization
Last October, on the Wall Street Journal’s “Real Time Economics” blog, there was a discussion of a new work of economic analysis which purports to be “a model for predicting a nation’s future growth more accurately than any other techniques out there.” The model they discuss is rather sensible. It ranks countries according to their “productive knowledge,” a concept the blog summarizes as,
the skills, experience and general know-how that a given population acquires in producing certain goods. Countries with a high score in the report’s “economic complexity index” have acquired years of knowledge in making a variety of products and goods and also have lots of room for growth. Essentially, the more collective knowledge a country has in producing goods, the richer it is — or will be.
This focus on actual production knowledge is unique in US culture because it admits (not explicitly, but what do you expect), to channel a sort of Marxist Whitney Houston, “workers are the future.” At a time when we are being told that workers will increasingly lose out to robots in the coming years, it seems instructive to think less about what could happen and more about what should. Or, more accurately, to think what might be missed in the fetishization of replacing living labor with capital intensive technologies and globalization that dominates our culture.
Namely that the deployment of technology may be a short term labor-saving device, but it doesn’t produce new knowledge or practices in the labor process. In the long run, not only will there be less flexibility in terms of generating profits; there will also be fewer new options generated by the labor process itself. In a sense, I’d like to argue, what the Complexity Atlas reveals in its rating of the US is that the long run costs of Post-Fordist casualization and globalization are that it has fewer domestic resources to draw upon for future growth. This sclerotic rentier economy is not some accident or a reasonable consequence of its maturity: it is the result of intentional policies and actions designed to destroy the power of labor, thereby limiting the possibility of increasing returns from it’s most dynamic “factor” of production. The US remains “economically complex,” but the possibility of future growth is cut off by refusing to harness this complexity in its most elemental form. I will lay out this argument below and then discuss what this means for not only future economic planning, but the future of higher education.
Long run economic value is not produced solely through the extraction of surplus from the labor process (i.e. it’s all about profits); instead it is the result of what the authors of the paper call the “social accumulation of productive knowledge” as applied to the productive process.
The essential theory … is that countries grow based on the knowledge of making things,” Mr. Hausmann said in a phone interview. “It’s not years of schooling. It’s what are the products that you know how to make. And what drives growth is the difference between how much knowledge you have and how rich you are.
This gap between productive knowledge and current GDP is presumed to be one that will eventually narrow for countries near the bottom of the income ladder today—but who currently generate much of the know how at the level of production because they, well, produce a lot of stuff. So China, which makes a lot of stuff but has a relatively low per capita GDP, has a lot of room to expand via this intellectual productive capacity; The US, which they rank at 91 out of 128, “is very rich already and has a lot of productive knowledge, but it doesn’t have an excess of productive knowledge relative to its income.” So, in the next decade, we don’t have far to grow.
This sits nicely beside arguments like that of Tyler Cowen, in his latest book The Great Stagnation. The widely covered thesis of this book is, basically, that we’ve picked all the low hanging fruit of technological innovation and the development of our workforce. According to Cowen, barring some transformational breakthrough (equivalent to the internal combustion engine or electricity) countries in the west can look forward to a long period of stagnant growth. Using their concept of economic complexity, Hausmann and Hidalgo claim that they can both explain and quantify why this is the case. In the intro to their paper they say:
During the past two centuries, the amount of productive knowledge we hold expanded dramatically. This was not, however, an individual phenomenon. It was a collective phenomenon. As individuals we are not much more capable than our ancestors, but as societies we have developed the ability to make all that we have mentioned – and much, much more.
The authors give much of the credit for pooling this knowledge in “institutions and markets” and they speak of this productive know-how as if it is some sort of natural attribute. They also speak of the potential economic growth possible in this mode of production as if that is also a completely natural endogenous process—as opposed to, as in the case of China, India and the countries that near the top of their index also benefitting from changes resulting from globalization. [something here on their previous domination].
I’d like to speak to each of these, but I think the most controversial claim is that economically important knowledge is produced not through a separate science or technological process, but through the productive process itself. In mainstream economics, the closest thing to this is the the closest thing I know to this argument is that made by Paul Romer in his pathbreaking articles of the early 1990s. These are chronicled by David Warsh in his book Knowledge and the Wealth of Nations. As Warsh puts it, the main paradox Romer introduces in his work is the idea that a good can be both non-rivalrous and (at least partially excludable)—and in so far as it is both, it represents a separate input into the production process. ”Knowledge” and “technological change” are primary examples of these goods. This paradox hinges on several premises Romer’s pivotal article introduces into the debate about growth. First is the rather conventional assertion that technological change lies at the center of economic growth. If we allow that technology is not just a thing, fashioned to do a job, but also includes social technologies like the division of labor, perhaps this is true (though typically when we talk about tech in this country, we’re talking about machines and gadgets.)
The second premise is more contentious: most theories of technological change consider it to be an exogenous factor—research is produced, innovation occurs, through a separate social, educational, or otherwise public R&D process, then business adopts and adapts to these changes. Knowledge is seen as somehow exogenous to the process of economic growth, i.e. it is not created by the production process. Romer cites Robert Solow’s assumption that knowledge was largely produced through some sort of public entity, (therefore nonexcludable and nonrival). Romer’s argument here, is that knowledge is produced by “private, maximizing behavior.” Romer asserts that technological innovation is both internal (or endogenous) to the production process, and that it, ”arises in large part because of intentional actions taken by people who respond to market incentives.” Here his main targets are economists like Solow who think of technological change as a public or social good, created as an externality or spillover effect rather than an intentional, profit-maximizing initiative within the production process. This, for Romer, sits nicely with his last point.
Romer sums up this position, by now fairly commonplace in thinking about Intellectual Property Rights and the role of technological knowledge in the Post-Industrial, Informational economy:
Once the cost of creating a new set of instructions has been incurred, the instructions can be used over and over again at no additional cost. Developing new and better instructions is equivalent to incurring a fixed cost. This property is taken to be the defining characteristic of technology. (S71)
The third premise, then, is that there is some separate good—called knowledge—that can be unbundled from the production process. One of the key distinctions for Romer is between this design (“a new set of instructions”) and the various forms of human capital outlined by his his University of Chicago ancestor Gary Becker. Human capital (“like the ability to add”) is a very distinctive, material good that it embodied in the individual who possesses these abilities. The abilities must be learned and once they are learned, they can only be engaged in one task at a time. And when the learner dies, her skills die with her. ”Knowledge” lives on. Though the instructions are still instantiated in some material form, their infinite, identical replication is possible, making it a much more non-rivalrous good than human capital. In Romer’s words,
What is unambiguously true about a design is that the cost of replicating it with a drafter, a photocopier, or a disk drive is trivial compared to the cost of creating the design in the first place. This is not true of the ability to add. Training the second person to add is as costly as training the first. For simplicity, the arguments here will treat designs as idealized goods that are not tied to any physical good and can be costlessly replicated, but nothing hinges on whether this is literally true or merely close to being true (S75)
As Romer himself says, this is an “idealized” distinction, for use in analytical, mathematical models. However, this idealized understanding of what has come to be known as Intellectual Property Rights and the role they play in trade has guided public policy on the matter right up to the present day. In a sense, it is a fundamental distinction that is central to political economy: Romer asserts that there is some way to separate the conception of work from its execution. Since at least the middle of the nineteenth century, owners of capital have found this idea deeply seductive: Romer finally worked out the math to illustrate its centrality.
However, in order to do so, he had to alter one of the other fundamental assumptions of neoclassical economics: the competitive market. As Warsh points out, these three premises only operate in the paradoxical environment of “monopoly competition.” If the development of this separate form of knowledge involved significant investment on the part of the firm, then it should be seen as a form of fixed cost. And since these were market oriented firms, they had to deal with fixed costs as fixed costs: they had to be amortized as a business expense.
They have to cover their fixed costs—the costs of going into business in the first place, before a single item can be sold. That means they have to be price makers. They have to act, or try to ac, at least for a time, like monopolists (217)
They are helped in their charade by the rather robust intellectual property rights system, but even Romer seems coy about whether this is absolutely necessary to the process. More on this in a moment.
Though I admit to being completely baffled by the math Romer is probably most known for (what does a comma stand for in an economic equation?), there are several issues to pick apart in Romer’s premises. For the sake of getting to the point, I’ll bracket the questions inspired by the first premise—the first being what exactly qualifies as a technological innovation. As Steven Johnson’s new book articulates very well, truly productive innovations are largely context specific. Likewise, I think he overstates the second premise, though it seems to be made in a specific context as well. I don’t have a sense of how irregular his observation was, but the patent system in general seems to be based on this ideal of the innovative producer. The problem, of course, is that much of the basic science research either remains undone or the product of extensive, publicly funded work at universities. This is the basic argument of Michael Perelman’s book Steal This Idea.
For the sake of argument, I’ll agree that knowledge and technological change are transparently and causally related to economic growth and that the endogenous profit motives in some way drive it. In that case, we arrive at the third premise, which is the crux of Romer’s deviation from the arguments of the authors of this Complexity Atlas. Namely that there is some specific form of knowledge separable from the productive process itself. This raises several questions. What is the nature of this knowledge? How is it produced? How does it contribute to economic growth? and, therefore, what sorts of policies and frameworks help foster it?
I don’t see this as a theoretical question in the least. In many ways, the world we live in today is the spitting image of the one Roemer describes. On the one hand, as insinuated above, the idea that there is a separate, specific knowledge that can be isolated and protected from competition is the foundation for the global push for stronger intellectual property rights. The current frameworks inevitably favor incumbents and arguably do less to create new innovation than other possible models.
On the other hand, the reason for this strict adherence to the current intellectual property regime is that it is necessary to cement and secure the current international division of labor called globalization. The latter is based on the idea that conception and execution can be neatly separated in space and time. In his model, Romer explicitly divides the economy into separate sectors, one where innovation and R&D occur and another where a licensed firm performs the work of actually manufacturing the product. As José Gabriel Palma describes it in an article on the process of deindustrialization,
In [Romer’s] models increasing returns, though generated by research-intensive activities, are explicitly not associated with manufacturing activities as such or with investment in manufacturing; nor do they allow for specific effects from manufacturing on R&D activities
This points to the overall problem with Romer’s model—and, consequently, with the model of the global economy of the last half century. Namely, that it has forgotten the importance of the tacit knowledge the Complexity Atlas discusses. Romer operates with an idealized belief that you can neatly separate the IPR from the tacit knowledge. The rhetoric of the maximalist IPR critics makes it seem as if the costless replication of their products is a reality against which they must be protected. The policy recommendation is then that the state should protect IPR and invest in R&D. These sound very reasonable from this perspective, but they remain trapped in an idealized understanding of how knowledge is produced. As Daniele Archibugi and Andrea Filippetti argued in the Global Policy Journal in May, tacit knowledge is essential to this replication, even when there is full cooperation on both sides of the transfer:
Most knowledge is useless to firms and organisations that do not have the ability to absorb it. In order to exploit knowledge successfully, potential users or competitors have to invest enough time, effort and resources to use it. In principle, IPRs protect the codified part of knowledge but not the tacit component. Many cooking recipes are freely available and everybody can afford a cookbook with recipes from the five continents. But this will not make good cooks of us all. In order to become good cooks, people will need equipment, ingredients, experience and talent. Industrial knowledge is not very different from the expertise required in the kitchen. The Arrow (1962) ‘paradox’ – if the potential buyers do not know the content of the information, they cannot appreciate its value, but if they know it, they do not need to buy it any longer – does not hold true for industrial and innovation-related knowledge. As people involved in technology transfer know very well, within the branches of the same multinational corporation, even when there is much interest and goodwill to transfer best-practice techniques from one plant to another one, it takes a long time and a lot of patience to achieve the same quality and efficiency. If learning costs are so high, more attention should be paid not just to the production of knowledge, but also to what makes this knowledge exploitable for users.
Archibugi and Filippetti use this observation as an opportunity to question the reliance on IPRs to foster innovation: this undermines Romer’s argument that innovation will only happen if there is a competitive monopoly (and leads to the argument, made by Michelle Boldrin and David K. Levine which says that innovation is possible without IPRs, an important argument in the current legislative climate). The implications of this for the future of IPR policies on a global scale should be clear: in effect, less is more.
But more importantly for the present argument, Archibugi and Filippetti effectively disputing both the second and third of Romer’s premises: #3 is undermined by saying that knowledge is created in both tacit and, to use their terms, “codified” forms; #2 by saying its production shouldn’t be left only to “the hands of profit-seeking agents:” ”More resources should be devoted to research carried out in universities and other public institutions and greater attention should be devoted to facilitate users to acquire and apply the knowledge generated.”
In short, there is something to the idea that producers learn in the process of production—developing the tacit knowledge Hausmann and Hidalgo find so instrumental for development. Their conception of the way specialization helps to expand the capacities of production is straight from Adam Smith so it is likely to be widely accepted.
The science of (worker) management
Scientists and their science made no significant visible contribution to new technology before the late nineteenth century. Adams writes that, ‘few if any salient technologies of the Industrial Revolution can be thought of as science based in any direct sense. They can better be described as craft based in important ways.’
Robert Adam’s Paths of Fire
By reducing as far as possible institutional links between the firm and its workers, by encouraging casual work, there is an danger of an even greater gap between workers and the jobs they do - workers are even less consciously involved in the fight for productivity and quality.
Separation of the conception and execution “even easier to establish between large firms and subcontractors, on an inter-regional or international level. Large firms set up in urban areas with highly qualified, highly paid staff, monopolize the design of machines and the sale of end products, and subcontract actual manufacturing to satellite firms operating in poorer areas with widespread unemployment (37).
Lepietz goes on to point out that, while this may have short run gains in revitalizing profitablity, it turns the production chain into something like space exploration, where “modules [are] sent to Jupiter by engineers sitting behind their desks on earth.” But this alienated form of management is only half the problem. Because the workshop is mostly automated and the laborers more casualized, it significantly reduces both the ability and the incentive for workers who experience the workflow in real time to make any improvements to that process. You have to rely on an outside organization (which created the expensive machine) to monitor and repair its performance while the people using it every day have no way of creating improvements at the point of production. In Lipietz’s words, “there are lower ‘on-the-spot’ productivity gains because there are no people actually working on, adjusting and improving the machines on a permanent basis” (38).
Lipietz argument is rather sensible in terms of the downside of automation, and he provides a good example of this in relation to one of the more famous industrial struggles in the last fifty years: that of the Fiat factory in Turin. The factory gives a lot of insight into the reasons behind automation, which often have little to do with economic efficiency or productivity gains per se.
Turin was the flash points of the Italian workerist movement. Central to the antagonism was the Autonomist Marxists who encouraged workers to see their political power as the primary producers in the system. Instead of being afraid of capital and factory owners, they should challenge them head on as they were the ones whose labor produced the value for the system. Were they to simply refuse to work, they would deprive the system of its lifeblood, lead to its demise, and then create a more just world on its ashes. These ideas, combined with increasingly antagonistic conditions inside the factories, led to a massive strike at the Fiat plant in Turin in 1969.
These strikes then spread beyond Turin, to encompass the whole of Northern Italy in a series of strikes and industrial disputes now referred to as the “Hot Autumn.” Current Autonomists like Bifo Berardi claim (and other Autonomist oriented writers like Hardt and Negri concur) that these disputes were the catalyst for pushing capitalist industry to globalize and transform itself into the two tiered post-industrial economy described by Lipietz. This certainly played a part—more on this in a moment—but for our purposes, the other interesting result was that the owners of the factory, according to Lipietz, “set up the almost totally automated workshops of Robotgate, Digitron, and LAM in order to have no workers present in the most troublesome sectors.” This total automation was extremely expensive and, from the perspective of economics, not all that much more efficient.
But it was politically very useful: it destroyed the power those workers had in the production process, i.e. the power they, as holders of creative, tacit knowledge, held against the owners. As Harry Braverman points out in his book, this is the original purpose behind the Taylorist impulse towards “scientific management.” The separation of conception from execution and the continued deskilling of workers was supposed to give more power to the owners and managers: workers were slowly de-skilled therefore reducing their leverage. If they had no union protection, were they to protest, they could be easily replaced by unemployed workers. In the Turin situation, not only were the workers encouraged by their understanding of their “productive knowledge,” but the slowing of immigration into that part of Italy had reduced significantly the “reserve army of the unemployed.” The choice to introduce expensive machinery was an end run strategy of management unable or unwilling to placate workers politically. In retrospect, however, this proved to be a poor decision economically—in ways this new study echoes. As Lipietz says,
The aim was to reestablish managerial authority; the cost of automation far exceeded the investment needed for optimal technical efficiency. At the beginning of the 1980s, after Fiat’s ‘Fordist’ workers had been defeated, the management had to admit that ‘the LAM system, designed at the time when industrial relations had broken down, is an interesting but one-off development. It is very expensive, it takes up a huge amount of space, and it breaks down more frequently than a less sophisticated installation.’ The alternative, which I and many employers, including Italian ones, advocate, is obviously to go for systems which are ‘less sophisticated’ but which mobilize the skills of line workers in real time, at the actual point of production. Out goes ‘paradoxical’ involvement; in comes dialogue between machine design, machine maintenance and line production [… .] This would mean groups of workers, highly skilled in more than one area, being able to adjust and repair machines, and to advise and even cooperate with designers.
The economic advantages of this are perfectly obvious. The skills and know-how developed in the process of working on the factory floor would be continually infused into that process itself. On the other hand, it shifts the balance of power back to workers so that, for their more central involvement, they are able to garner more control over their place in the work process - and how they are remunerated.