Will the Real Engineers Please Stand Up*

August 27, 2013

Two months ago, I had the rare opportunity to speak to an audience that is foreign to (most of) us: a room full of natural scientists. The conference that I addressed, BioBricks Foundation 6.0, met at Imperial College, London, to discuss the most recent developments in synthetic biology. The panel in which I participated provided a space to introduce ‘recent’ developments in science and technology studies to synthetic biologists. I read it as an occasion to talk about the canonical bread-and-butter of social studies of finance—that is, how economics performs the economy. (I wholeheartedly thank Pablo Schyfter and Jane Calvert for their invitation).

While the contents of my talk are nothing new for the readers of this blog, the experience was particularly stirring in other ways, not the least because of the reactions of the audience to the concept of performativity (in their questions, those who remained in the room to listen to the social scientist sought clarification on how to model and predict what they saw as no more than noisy feedback loops between the abstract descriptions of economics and a detached, yet largely objective real world). Indeed, the greatest reward from participating of the conference came from having been exposed to, even if for a few hours, the awesome metaphors of synthetic biology.

For those who have not followed synthetic biology in recent years, an introduction is warranted. Synthetic biology as we know it today is a bold quest to create an “engineering technology based on living systems” through which it would be possible to construct biological setups designed to solve concrete real-world problems. Synthetic biology, in this sense, amounts to hacking the biological parts found in nature into new forms and configurations: sifting and reassembling genes, proteins, and other macromolecular components as if they were pieces in a minute meccano set. Synthetic biology is the production of designer nature at a sub-cellular level. As one of the key figures of the field argues, whereas most of the developments in biotechnology over the twentieth century resulted from an “expensive, unreliable and ad hoc research process” (Endy 2005), synthetic biology seeks to make the production of new biological forms commonplace, to give

undergraduates and high school students, without prior training in biology or biological engineering, [the ability to] to design synthetic biological systems of their own invention comprised of several dozen pre-existing standard biological parts, order and receive the DNA encoding the system, and show it to work (Endy 2005).

A bold vision, indeed.

The quotations from the previous paragraph come from an essay that has become a kind of manifesto for synthetic biologists: Drew Endy’s ‘Foundations for Engineering Biology’, published in Nature in 2005. And, bringing the discussion closer to the sociology of economic life, perhaps the single closest development in economics to Endy’s vision is found in the contributions of Paul Milgrom and Nobel laureate Alvin Roth—credited as founders of what we now call ‘market design’. In a nutshell, market design utilizes “laboratory research, game theory, algorithms, simulations” (Milgrom 2008) and a host of other resources to “bear on practical questions of microeconomic engineering” (Roth 2007). Much like Endy’s vision, market design seeks to build novel objects in the world, by furthering “the design and maintenance of markets and other economic institutions” (Roth 2002), as Roth wrote in his widely known essay ‘The Economist as Engineer’.

Unlike Endy and other synthetic biologists, however, market designers stress the scientific character of their nascent sub-discipline. As Roth wrote, the relationship between economics and market design should be the same as that between physics and the construction of a suspended bridge. “As marketplaces proliferate on the web”, he argued,

a great deal of market design is going to be done by computer programmers, among others, since they possess some of the essential expertise. Economists will have an opportunity to learn a lot from the markets that result, just as we will learn from our own work designing unusual markets. But if we want this knowledge to accumulate, if we want market design to be better informed and more reliable in the future, we need to promote a scientific literature of design economics. Today this literature is in its infancy. My guess is that if we nurture it to maturity, its relation with current economics will be something like the relationship of engineering and physics, or of medicine and biology (Roth 2002).

And it is here, I argue, that the engineering metaphors of market design simply fall short of what they ought to do, particularly if contrasted to the bolder agendas of synthetic biology. Whereas synthetic biologists focus on making, on creating tools, standards and components that will allow twelfth-graders to design entirely new biological systems in their garages, market designers seeks to maintain the status of markets and their relation to economists and economics. True: synthetic biologists need science—indeed, much of the conversation at BioBricks Foundation 6.0 had to do with identifying the limits of the state of the art and how these translated into limits on biological design. Yet the ultimate objective of synthetic biologists is to empower individuals to remake nature themselves, not the production of knowledge itself. The difference is one between nineteenth century engineering—which was often presented as a combination of craft and science—and engineering as the constant manufacture of a built reality. Market designers have perhaps not been ambitious enough. They should, perhaps, rephrase their engineering metaphors, to imagine a world where high school students without prior knowledge will be able to design markets and show them to work.

What could be gained by a stronger vision of market design, one that, like synthetic biology, sought to identify, standardize and build components of markets which users could refashion for specific ends? Indeed, embracing a stronger, less scientized metaphor of engineering would perhaps permit imagining a more complex and elegant uses of markets in modern societies. And to illustrate this point, I offer the following example: how expanding the repertoire of market design to emulate the vision of synthetic biology may permit the construction of self-imploding markets.

In nature, cells can die through what is known as apoptosis—a form of endogenously controlled biochemical events whereby cells cease to work, degrade and die. A recent development in synthetic biology has been to control and program apoptosis, in order to deactivate certain cells exposed to specific external stimuli. The utility of this process is clear: understanding how to trigger apoptosis, for instance, may lead to better cancer treatments; but it may, too, allow engineering cells that are capable of being deployed to control environmental contamination only to then be disabled from a distance. Self-imploding life, designed for single use.

If we were to make the engineering metaphor of market design closer to that of synthetic biology, could we not also move beyond the creation of ad hoc auctions, to envisage the programming of market apoptosis, the creation of markets designed to self-implode? How might these markets bear upon real-world problems? Consider the illegal global trade in rhinoceros horns.  In the 1970s, there were roughly 60,000 white rhinos in the southern horn of Africa. In 2012, and despite an important growth in the expenditure on preservation efforts, the population had reduced to around 20,000. A decisive factor in the reduction has been poaching: in 2011 alone, conservative figures estimate that around 300 rhinos were poached in South Africa, with some 560 poached globally.

The key driver in rhinos hunting is the procurement of their horns. In today’s black market, a kilogram of rhino horn can sell for about $40,000 USD, with demand driven primarily by China, Yemen and Vietnam. In practical terms, this means that rhino poaching can be a rather sophisticated operation involving well-organized midnight helicopter flights, automatic weapons, and night-vision devices whilst remaining profitable. Given current international restrictions, the trade in rhino horns is entirely illegal, and all major actors in the market are known: most of the rhino horn is distributed through middle-sized pharmaceutical companies that sell horn powder as an alternative medicinal remedy. The high demand of rhino horns, in addition to the high costs of wildlife conservation, are coalescing in an unsustainable situation which, if continued, may lead to a depletion of the stock and the eventual extinction of wild rhinos.

In response to this, the Endangered Wildlife Trust of South Africa suggested introducing a regulated market of rhino horns. This is a solution that has worked elsewhere with much success, but generally in cases where preserved species were very local and hunting permits were very scarce. The market for rhinos, in particular, would imply the creation of a central selling organization—much like De Beers in diamonds—who would distribute horns to a handful of recognized partners. Profits would then be used to fund conservation efforts which would increase the risks for illegal poachers and disincentive illegal trade.

As a colleague from the International Union for Conservation of Nature pointed out, though, the problem with this solution is that, first, it assumes that horns are well-behaved goods (that higher prices will translate into lower demand); second, it underestimates the regulatory costs associated to maintaining a complex international market; and third, by creating a legal channel of distribution, the regulated market provides an mechanism for laundering illegally poached horns. Given the prohibitive costs of protecting rhinos, the possibility for poaching remains and the incentive for laundering increases. Indeed, the market does little to reduce demand and, if failed and ill-designed, may provide a system that will only increase the rate at which rhinos are depleted.

The question is thus the following: could we not imagine a market that, along with the design of an auction mechanism and clearinghouse, contemplated a system for the abatement of demand? That is, can we imagine a market designed as a stopgap of poaching in the short term (one that makes use of all the ‘traditional’ advantages of a market solution) but that, by construction, will also erode demand in the long run? Can we build a market designed to fail?

The questions are clearly pertinent beyond rhinos and their horns (think, for instance, on the recent discussions on the legalization of some classes of drugs and their introduction to the market). Rather than thinking of market design as auction design, we should perhaps follow Endy and think of biologically-designed solutions that are coupled to broader ecological systems. Self-imploding markets do just that, and offer clear benefits. And all this poses the question: where is the Endy of market design?

The challenge here is not only one of shifting metaphors but, more fundamentally, of thinking about the type of knowledge and expertises needed to engineer novel market solutions. Demand, like any other form of consumption, is a product of cultural frameworks mediated and maintained by specific institutional arrangements. Studied alike by economists and sociologists, market failure is, similarly, a complex situation in which institutions, cognition and culture play a fundamental role. And so, there is, perhaps, an interesting opportunity for new disciplinary conversations—for exploring a form of market design v2.0 which seeks to better understand the frictions, path-dependencies and constraints of markets as socio-political realizations (cue social studies of finance?) . It is, thus, a novel opportunity to design.

Talking with the language of design provides, as Martha Poon rightly pointed out, a more productive approach to the study of markets. But it also makes possible imagining a bolder version of market design than that currently advocated within economics. While the markets created by Roth and Milgrom are truly feats, much more can be done. Indeed, market design need not be a type of ‘consultancy economics’. Rather, it can follow an alternative metaphor that is pragmatic, perhaps even civic, an image of the future closer to that of the (biological) engineers who today work away in their labs redesigning the fundamental building blocks of nature.

References

*This is the summary of a work in progress on the intersections of STS and market design. A first version of this brief essay was posted on estudiosdelaeconomia. I thank Jose Ossandon wholeheartedly for his original invitation, and Daniel and Yuval for posting this version here.

Endy, Drew. “Foundations for Engineering Biology,” Nature 438 (2005): 449-453

Milgrom, Paul. “The Promise and Problems of (Auction) Market Design,” Nemmers Prize Lecture (2009)

Roth, Alvin. “The economist as engineer: game theory, experimentation and computation as tools for design economics,” Econometrica, 70 (2002): 1341–1378

Roth, Alvin. “What Have We Learned from Market Design?,” NBER (2007)

10 Responses to “Will the Real Engineers Please Stand Up*”

  1. danielbeunza Says:

    Very intriguing. I particularly like your call for standardization. In the inaugural meeting of the Center on Systemic Risk at the LSE, Kevin Houstoun (chairman of Rapid Addition, and with experience in the FIX protocol for communication between exchanges) made a similar point… if the problem is ensuring safety amidst complexity, standardization should be key. However, standards naturally limit differentiation and thus returns. Perhaps the NGOs and political activists that favor a safer financial system should be pushing for standards rather than simply demonizing bankers.


  2. Thank you, Daniel! There are three very interesting aspects to your comment.

    The first has to do with standardization, which is an intriguing challenge for market design. For synthetic biologists like Endy, the key objective is to standardize nature in order to empower hackers to build new designs of life. This entails finding arrangements that can work as fundamental building blocks and which perform specific functions (like the transistors, condensers and resistances in electronics). In markets, however, standardization is not entirely straightforward perhaps because the market is itself, as Phil Mirowski has argued elsewhere for years, one of the most understudied objects in economics. In effect, Mirowski’s Markomata may show the type of work that needs to be performed in standardizing markets, studying and better characterizing individual functions in markets such as clearing, matching, information dissemination, signalling, et cetera.

    Second, your mention of Kevin Houstoun and FIX is very appropriate. As you well know, my pet peeve is looking at market infrastructures. I think there is much to learn by not only studying market infrastructures as historical entities which bear upon the operation of the market, but also as concrete systems that crystallise specific market standards. In particular, in trying to understand better what standards do in market, I think we would benefit greatly from talking with the type of infrastructural workers that, like Houstoun and the others involved with FIX, are responsible for making the market’s infrastructures. Indeed, I feel that a stronger version of market design goes in hand with an infrastructural understanding of markets.

    Thirdly, I completely agree: activists should be less concerned with bemoaning the encroachment of markets in specific areas of social life and more able in influencing the way markets are designed. I am quite amazed, for instance, that the equivalent of ELSI (Ethical Legal and Social Impacts of) has not been taken up by politically active organizations. Perhaps more traction would be gained in the long run by influencing the standards on which specific markets rely than, as you say, demonizing bankers and other market participants.

    • danielbeunza Says:

      Right! What I thought of, for example, is something like a set of standard practices in securitization that can be certified as safe. This would be different from being rated as safe in that the emphasis would be in similarity of the practice rather than the expected riskiness of the returns. Focus on the process, rather than the output.

      • Juan Pablo Pardo-Guerra Says:

        Great idea! If you think of it, rating an instrument as safe is a bit absurd in and of itself (it sort of naturalizes the object as if it had an intrinsic behaviour independent from an outside world; then again, that is what financial economists do with markets when they obsess about price behaviour). To stretch the idea to an admittedly absurd limit, certifying the safety of a security is a bit like certifying the safety of a semi-automatic weapon. The weapon may not blow up in your face (like the first generations of guns did, many years ago), but that doesn’t mean it is safe in a broader sense.

  3. yuvalmillo Says:

    JP, this is a really nice and thought provoking piece. Thanks a bunch for this!
    A couple of interrelated points.
    First, I think that the topic of ‘power to design’ that you raise is valid and fascinating. However, I think that politics enters the story in a more fundamental way than you imply in your piece. Namely, the different nature of scientific rule at play. Whilst in molecular biology as we know it today, for example, the structure of proteins is going to be decided by the order of DNA bases and not the other way around, there is no limitation on the number of ways in which an auction could be designed. That is, rules in economics are far less binding than rules in natural sciences. Innovation in economics, thus, such as market design, I would argue, is less a process of engineering (that is, creating technological applications of scientific discoveries), but more a political task – accumulating enough influence so that a specific market design would materialise.
    Second, the main reason mentioned in the literature for economics (and other social sciences) not having ‘Strong Scientific Rules’ is the reflexivity of actors. This is also the main obstacle, I would say, to the rhino horn self-imploding market. What would prevent economic actors from acting in such a way that would prevent the market from imploding? There is nothing in the market itself that can prevent them from sabotaging the initial intent of the design. There is, however, a force that can prevent them from doing so. That is, of course, the brute power of the state. A market designed by a regulator would also come with the clout of the state behind it. So, I would say that market engineering is always (as long as economics stays the way it is now) a political exercise and much less a scientific one.

  4. Juan Pablo Pardo-Guerra Says:

    Thanks Yuval!

    On the first, I fully agree that politics is an integral part of the story of design in general, and market design in particular. For instance, to think of auctions as a-political would be far too simplistic. Indeed, at a very basic level, auctions are sorting devices (they determine on the basis of rules how a distribution will take place) and as we know too well, any form of sorting and classification is political at its core. In this sense, I do agree that innovation in economics is a political task of, as you write, accumulating influence so that specific market designs materialise, but I would argue that that is the case with most forms of heterogeneous engineering and innovation. Designing and making things requires creating the institutions and environments in which they will flourish, which involves an overtly political task.

    [A minor technical note: I am not entirely sure that there is in practice no limitation on the number of ways in which one can design auctions. For instance, there are a handful ways in which you can organize an auction (e.g. sealed bid auctions, English auctions, Dutch auctions, etc). And I guess that these different auction formats have some type of formal equivalence from a game-theoretic perspective. Further, if we were to look at real-world auctions (there are discrepancies between the outcomes of auctions in experimental settings), I would also imagine it would be possible to come up with some sort of empirical taxonomy, which may be useful or not for market design]

    I also agree with your second point, though I think there are workarounds. Atoms are not reflexive; humans are (though I know some modellers and econophysicists who would argue otherwise). The key thing is that, as I think we agree, market design is not only about building the transactional space in which objects are valuated and distributed but also the institutions that allow for these processes to happen. The state is an example, and a strong argument for market design is also, in one sense, a strong argument for the state. In effect, the presence of the state has given weight to other forms of self-imploding (or, rather, temporally bound) market entities. Patents are, at the end of the day, limited monopolies guaranteed by the state. And so are options which, as contracts, have a form of juridical backing which gives meaning to their expiry dates. Of course, you may always observe forms of innovation that seek to exploit opportunities from a particular market (I could imagine, for instance, a secondary market in some form of exotic insurance for rhino horn futures). And there is always a risk of sabotage from within. But then again, that is a problem with any other complex form of economic organization.

  5. Juan Pablo Pardo-Guerra Says:

    Reblogged this on Real-World Economics Review Blog and commented:

    I’d like to share this post that I wrote for estudiosdelaeconomia/socializing finance. An invitation to, perhaps, rethink markets in society?

  6. yuvalmillo Says:

    JP, these are really good points. Let’s talk more – I ran by some colleagues here the idea of doing a ‘market design’ workshop and people are quite interested.


  7. […] episode complements JP’s post ‘Will the real engineers please stand up?’ where he raised the possible relationships between synthetic biology and market design.  What I […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: