Academy of Management and SSF: Back to School Reflections

September 11, 2010

We are happy and honoured to have Bruce Kogut as a guest blogger on SocFinance. Bruce is  the Sanford C. Bernstein & Co. Professor of Leadership and Ethics at Columbia Business School.

Welcome aboard, Bruce!

I’m not a regular AOM attendee, for reasons both trivial (I don’t like conference hotel locations in hot cities in the summer) and more substantive.  However, Montreal clearly redressed the problem of location and the opportunity to share a panel with the likes of Daniel Beunza, Fabrizio Ferraro, Yuval Millo, and Teppo Felin was substantively too enticing to decline.  At AOM Montreal, the curiosity and energy around the SSF research program was clearly evident.   But we also know from studies on innovations that most of them die.  Will this one do better?

My money is on the bet it will for a number of reasons, the primary one being that SSF has had the most success in mustering both the micro behavioral details and the big picture implications to discuss a major industry that has been insufficiently researched. And a financial disaster also helps the urgency.  Of course, there are wonderful intellectual antecedents such as those contributed by Wayne Baker, Mitchel Abolafia, Viviana Zelizer, and many more.  But SSF has a powerful though contradictory weapon on its side: it gives a positive analysis of whether financial markets perform.

Whereas the past work excelled at explaining the embeddedness of financial markets in networks or the cultural re-interpretation required to legitimate new financial instruments (e.g. life insurance), SSF puts performativity into the core of its agenda –for this we have MacKenzie and Millo to thank for their AJS article that so clearly and brilliantly solidified this ‘performative turn’.   We have shed our hermeneutics in preference for action, hallelujah.

This question, do these markets work, often strays into old and over-trodden fields, such as social construction and ‘reality’, and when it does, the results are not at all satisfying.   Some people get caught up in the self-fulfilling claims. Here too, this line of investigation –however meritorious its argument- is ancillary. Moreover, it is not the same matter as endogeneity, namely that the adoption of an innovation (e.g. Black-Scholes pricing algorithm) changes subsequently the market.

These kinds of phenomena were central, for example, to the debate on the adoption of the Milton Friedman admonition to index the growth of the money supply to the economy –once such a rule was adopted, it would of course be found to be performative. But performative does not mean ‘best’ or ‘optimal’ and it can also mean ‘bad’, that is counter-performative.  Now this is a challenging agenda –filled with non-linear consequences per the MM AJS article, and that’s what is so appealing to the SSF research program –it is ambitious.

In my view, this focus on performance has three great merits.  One, it establishes a clear empirical agenda and an emergent tool kit. By tool kit, I mean a set of ideas and analytical devices that enable a community of scholars to seize a hammer and get to work.  Now some will rightly complain that the tool kit is better defined for cultural sociologists doing the micro field work than for a sophisticated statistical treatment that is at least up to the challenges of endogeneity, if not identifying the predicted non-linearities (i.e. the market may crash).

For the big picture implications, simulating the macro consequences of micro behavior would be a good place to start, if data and statistical methods prove to hard to acquire anytime soon.    Still, however much I would like to see more effort allocated to a firmer analysis of the macro implications, I remained enthralled to the historical narratives that MacKenzie and others have woven to trace the performative consequences of financial innovations. I just don’t think these narratives will persist to enthrall and a more tedious empirical program will be needed to satisfy the larger community (and me).

The second merit is that it is creative destruction. The creative part is clear by my comments above and the destructive part is implied.  Rather than give you my views here, I have translated some passages from an interview with Michel Callon conducted by the intrepid and incisive Herve Dumez published in a type of feuilleton that one tends to find in France more than elsewhere.  The translations are to be found below; Francophones can go for the full interview to: .

The third merit is an aspiration.  By looking at performativity, the SSF program has the opportunity to say why an innovation worked or did not work.  Does it have the courage (I think this is the right word) to then state what we might be able to say a priori about good policies?  Does it add more insight into why markets may fail massively and what can be done?  I am somewhat pessimistic that there are simply enough SSF sociologists (or economic sociologists in general) to tackle this ambition.  I personally would look towards the new science of networks to come forward to state principles of good system design.  But there are other doors of entry –if sociology has the desire to enter and to engage in the tool development and commitment.

Here is a good place to conclude in reference to what I said earlier that performativity was a contradictory weapon. What did I mean? I meant two things. The first, which I sought to illustrate at the AOM panel by a discussion of excessive risk and compensation, is that an analysis of performativity will require the use of the exact models under study.  Here is the Karin Knorr Cetina observation on physicists: they have to build their instruments to measure their experiments in accordance with the theory.  It is this problem that haunts the efficient markets program: it is not enough to state the CAPM equation as a regression and then test for agreement –we also have to test the model itself (sometimes called ‘model error).  MM AJS got away with showing a chart of the ‘smile’, but the empirical test is whether the BS model agrees with a smile –is it empirically consistent with the model.   It is not, hence the deduction of counter-performativity.

The second contradiction is what sometimes is called a ‘mean’ observation.  For SSF, as has been noted, praises economists and financial engineers ever so ingratiatingly, while standing apart from the action.  If sociologists would wipe that occasional smug smile (not the option pricing smile discussed above) off their collective face and say, what would a world of policy look like if sociology sat proportionally at the desk of policy making and market construction, I would be ever so curious to know if they would fall prey to the uncomfortable occupancy of a golden cage from which economics sings its song.  May sociology come to know such a fate!


From “La performativité de l’économie”, Lilbellio, N. 3, 2006, Michel Callon interviewed by Herve Dumez

Callon does not like self-fulfilling

  • Since a statement is always performative, the important variable is the degree of dissemination and extension of arrangements to which it is associated and which it is true.  Either these arrangements are rare and the theory is true, but in few places. Or they are numerous, widely scattered, and the theory is true but this time in many places. For a self-fulfilling prophecy, there must be arrangements suitably shaped: it is what Donald MacKenzie shows perfectly in the case of Black and Scholes. With the notion of performativity, it is not located in the relativistic framework of the self-fulfilling prophecy and more generally of such conventions as coordination. What is sometimes interesting research is being able to get rid of some entire aisles of libraries:  here you can get rid of everything that has been written on the role of conventions in the functioning of the markets.

Callon does not like embeddedness

  • If the other sciences want to have the social impact of the confrontation (with policy), they will have to start, too, to experiment [to do empirical research]. The experiment is the crucible where the theory is elaborated and a framework which allows this theory to have an effect, has become a major issue in my thinking.  The stakes are both theoretical (for the practice of social science) and social (we need an open experiment, not an experiment monopolized by economics.) If we do not turn to experiment, we, sociologists and anthropologists, are condemned to repeat: embeddedness! embeddedness “, hoping to disqualify, with this magic word, economic theory. But as this repetition  puts us position to play the of embeddedness in our favor, by embedding as much as possible the economic activities in economic theory, this leads to nothing as before! We are fooled by embeddedness!

Callon dismisses the dichotomy of selfish and altruistic: people are compounds of both.

  • I take this opportunity to emphasize a point, which is the complexity or heterogeneity of practical arrangements are always arrangements of compositions “pure”.  Is the thesis that Gates qua agent is calculative since he managed to amass the largest fortune of world hypocritical when he then creates a foundation and makes huge donations? That is not the question. What is needed is to analyze Bill Gates as a compound, a combination of arrangements altruistic and calculative. We have begun to show, I think, that there are no calculative settings (agencement) without altruism, and vice versa.

4 Responses to “Academy of Management and SSF: Back to School Reflections”

  1. danielbeunza Says:

    Hi Bruce — thank you so much for this extremely lucid analysis of the social studies of finance. I particularly like that you support the intellectual innovation but also challenge us to do a better job and, as you say, avoid, ahem, “death.”

    I wanna take up just one of your points. What would it mean for the social studies of finance to influence actual practice. I’ll respond in connection with several blogposts here.

    In line with the social studies of science, the central point of the social studies of finance is to inquire about the nature of capital markets in an uncertain world and a technological, knowledge-intensive market.

    Hence the focus on the social and material aspects — the tools, models, technology, etc. If the thinking behind this perspective were put into action, there would be better tools that would improve how knowledge is incorporated into prices. The outcome would be prices that better reflect what is known out there.

    Here’s some examples in terms of new trading strategies, design of hedge funds, compensation of analysts or public policy regarding exchanges.

    1. Design. My early work with David suggests that hedge funds and prop trading desks in investment banks would reduce their risks by incorporating greater diversity of perspectives. They can do this by, for example, redesigning the trading room and moving towards a multi-strategy model.

    2. Strategy. Traders, according to my more recent work with David, should switch from convergence strategies in which they “go along” with the rest to “divergence” trades in which they play the contrarian. This avoids the problem of resonance and arbitrage disasters.

    3. Compensation. Work in SSF underscores the value of securities analysts — the real knowledge workers of the stock market, and unfairly lambasted by agency theorists. A framework developed by myself and Raghu gives banks a way to appraise their contribution, both in terms of horizon, metrics and evaluation.

    4. Supervision. My recent work with Yuval offers a new way to interpret the current debate on high-frequency trading (no small fish… a likely cause of the next crash). Instead of viewing electronic trading as the inevitable outcome of “progress”, it reveals the political nature of the recent reform and hints at the potential problems posed by the algorithms.

  2. […] leave a comment » Bruce Kogut has a post at Socializing Finance that might interest many readers. […]

  3. yuvalmillo Says:

    Bruce, many thanks for this!
    I want to stress a few of the points that appear in you ‘call to arms’, but should be made even more prominent, in my mind.
    SSF is still an emerging field, but we see a gradual crystallization around three focal points that gradually turn into the field’s ‘holy trinity’: materiality, performativity and long-term institutionalisation. In a nutshell, SSF argues that markets cannot be understood sufficiently without taking into account these three dimensions. First, markets are populated by economic agents who use devices, not by naked tool-less ‘utility maximizers’. That is, if we want to understand homos economics, we have to study and conceptualize the tools that **make him** a fully-fledged economic agent. Furthermore, different tools take part in the construction and performance of different economic agents. Second, bodies of expert knowledge (economics, accounting, law) affect the way markets behave as they gradually penetrate and become integrated into the infrastructure of these technosocial institutions. Third, to understand markets’ behaviour, we have to look at their ‘formative periods’: understand the cultural, political and technological sets of circumstances that led to their emergence. This notion implies, for example, that prices cannot be the starting point in studying markets, but, in a sense, the end point: we have to focus on the tools, conventions and practices that shape prices rather than to study only prices themselves.
    This last point in the SSF triad leads to another area that Bruce touches upon, but should be fleshed out more. The distinction between quantitative and qualitative methods, which divides, needlessly and frustratingly, social scientific disciplines, also affects badly the environment where SSF operates. The roots of SSF are in the ‘internal approach’ in the sociology of science, the one that says that you cannot produce a social analysis of, say, Mad Cow dieses without understanding sufficiently the biology behind it. This is even more crucial when we come to study markets. Like or not, financial economics affected the shape and the dominant discourse surrounding markets in the last five decades. We should not accept it as an objective description of markets’ behaviour, but we do need to understand it and, when appropriate, use its tools when developing SSF accounts. In doing so, we should also strive to enrich the set of quantitative tools finance uses. Again, I agree with Bruce that the new SNA tools, building on methods such as exponential random graph modelling is the way forward.

Leave a Reply

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

You are commenting using your 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: