Does Management Research Need to Become More Empirical?

23 August 2007 at 2:25 pm 8 comments

| Nicolai Foss |

Or, to put it more precisely, does management research (i.e., the journals) need to become more empirical in the specific sense of allowing for research that is pre-theoretic, but addresses an issue of relevance or detects a pattern to organizational stakeholders, that is, identifies a potentially important stylized fact?

In two SO!APBOX Editorial Essays in the May issue of Strategic Organization, Danny Miller and Constance Helfat both argue, in Miller’s words, that “the current institutional setting within which administrative studies develop has evolved to de-legitimize [this] type of research” (p. 177). As examples of the benefits of pre-paradigmatic, atheoretic research Miller points to Fleming’s discovery of penicillin, and, in management, to Woodward’s discoveries of the impact of technology on organization structure, the Hawthorne experiments, Tushman’s work on firm trajectories, etc. Helfat points to the Philips curve and the learning curve.

Miller and Helfat may be quite right that descriptions of stylized facts that are not somehow informed by theorizing are seldom, if ever, seen in the leading management journals (of course, to a hardcore economist — all management “theorizing” is essentially the kind of work that Miller and Helfat want to be done ;-)). However,  

  • Are there any known cases of this kind of research being killed by the journals to the detriment of the management field?
  • Isn’t it — contrary to what is implied by Miller and Helfat — the case that sometimes harm may have been done by atheoretical work on empirical regularities? Think of the PIMS
  • What is necessarily so bad about requesting that authors try to come up with a theoretical rationalization of an observed regularity? Is this such a barrier that it will stop potential authors from publishing their result? Theorizing comes in many forms, and some kinds of theorizing is much harder to do than other kinds of theorizing. What is wrong with requesting that, as a minimum, authors provide some verbal account of the possible underlying mechanisms that may produce an observed regularity?

Entry filed under: - Foss -, Methods/Methodology/Theory of Science, Recommended Reading.

Me in Chinese Investor Protection and Firm Governance: Substitutes or Complements?

8 Comments Add your own

  • 1. Steve Phelan  |  23 August 2007 at 7:53 pm

    Astronomy is a discipine with a long history of gathering empirical facts without theory – e.g. Tycho Brahe’s measurements of planetary motion.

    I am tempted to quote Feyerabend and claim that “anything goes” in science but clearly there are fads and fashions and cliques in research as in anything else.

    As I have mentioned before, there is a belief that good research requires large datasets with sophisticated models. Often theories are data driven rather than the reverse.

    Personally, I think management, as a discipline, would be better served aligining itself with a professional model (such as medicine or law) rather than pursue its science envy even though I (somewhat) understand the desire to gain legitimacy with traditional academics in the natural and social sciences.

    Since I received tenure I have been making a concerted effort to re-engage in the world of practice – to refresh my sense of what the problems are in the “real” world of management and feed that back into my writings. Time will tell.

    As you might tell I have a lot of sympathy for the “strategy as practice” crowd but I hope that they, as a group, don’t become a synonym for critical management studies. This would be akin to teaching medical doctors that medical practice is the same as critical health studies – which it most certainly is not. (Note that I am not saying that critical studies have no value to practitioners.)

  • 2. JC  |  23 August 2007 at 10:00 pm

    I think someone (who – ahem?) is being deliberately provocative here.

    Well, anyway, I flew back from London yesterday and happened to read:von Mises, L. (2007). The Historical Setting of the Austrian School of Economics. Auburn AL: Ludwig von Mises Institute.

    Blessedly short – since I am unsure of how much of that stuff I can take. On page 23 he writes: “Today, all over the world, but first of all in the US, hosts of statisticians are busy in institutes devoted to what people believe is ‘economic research’. They collect figures … They surmise they are thereby measuring mankind’s ‘behavior’. .. They look with pity and contempt upon those economists who, as they say, like the botanists of antiquity rely on much speculative thinking instead of ‘experiments’. And they are fully convinced that out of their restless exertion there will one day emerge final and complete knowledge …’. You get the general message, though I have tinkered with the way von Mises uses quotation marks – in a curiously American fashion, to add emphasis.

    Von Mises is arguing, especially on page 43 for ‘speculation’ and his own assumption that “there is a body of economic theorems that are valid for all human action irrespective of time and place …”. Dismissing those who argue that people’s economic activity might be shaped by social relations like politics, religion and other empirically and historically evident forms of power, he sums up the implications of his assumptions on page 51 “Society, i.e. peaceful cooperation of men under the principle of the division of labor, can exist and work only if it adopts policies that economic analysis declares as fit for attaining the ends sought”. Hmm … if only. But what kind of Utopia would this be anyway?

    The issue here, aside from von Mises’s extraordinary political message, is manifest in the remarks to which Nicolai refers, the idea that theory is ‘real’ knowledge. Hence there is a sense in which we now all know, post the sociology of knowledge/science etc., that there is no ‘pre-theoretic’ position. To look at the world and separate the relevant from the noise we must have a ‘lens’ or theory to use as a data filter or selection mechanism. Absent this we are overwhelmed in ‘meaningless’ sense-data.

    But this is a peculiar – in the sense of idiosyncratic, not in the sense of being ‘odd’ – modernist epistemological position.

    It happens that I was also reading:

    Cook, H. J. (2007). Matters of Exchange: Commerce, Medicine, and Science in the Dutch Golden Age. New Haven CY: Yale University Press.

    A much fatter book which is also much more fun.

    Cook, of course, is dealing with the very period of the ‘botanists of antiquity’ about which von Mises seems rather less well informed. Cook tells of the prevailing assumption, basically Aristotelian, that human beings can have two types of knowledge – which have little to do with each other. Or rather, how they are brought together becomes the arbitrary choice of the epistemologist (or religious) and determines the intellectual fashion of the time. Those who have looked into Vico find much the same.

    One kind of knowledge is that deduced from axioms. This knowledge is essential linguistic, depending on language and logic, to use Ayer’s terms. The second kind of knowledge is that of the senses. The distinction, which is clear in Locke, is familiar to us today from James’s ‘knowledge-of-acquaintance’ versus ‘knowledge-about’ – acquaintance coming from practical ‘sensual’ engagement with the material to be known, while ‘knowing about’ was contingent on ‘mentalist’ logical deductions from taken-for-granted axioms, those neither needing nor offering any other warrant (in the same way that stories – like Harry Potter – contain their own meanings).

    Cook points out the 15th and 16th century European botanists were denying the utility and validity of axiomatic and deductive ‘speculation’ and were intent on sticking to the ‘facts’. Indeed their whole purpose was to stand against the speculative science of the time. ‘Facts’ incidentally come from the Latin ‘factum’ meaning ‘a thing done’ i.e. a notion that is essentially empirical rather than theoretical.

    The botanists, and the apothecaries and medical professionals who were their ‘cousins in practice’, were less interested in quasi-axiomatic ’causes’ than in direct descriptions and the ways in which things changed through time and evolved. In this sense they were forbears of Bridgman and other modes of operationalism.

    Their cast of mind led directly to ‘case histories’ and the Harvard Business School pedagogy of four centuries later. My guess is that von Mises would not have been happy at HBS … and surely no-one teaches cases as if they are natural experiments and rigorous demonstrations of the Theory of the Firm (whatever that might be). On the contrary, cases are best understood as the grand-children of the botanists’ method. They are to surface heuristics useful to those with responsibilities in the particular, not to those writing in journals. They were, of course, the pedagogy of the German Historical School that von Mises is so intent on rubbishing.

    All the petty politics aside, the deep question here, of course, is what the term ’empirical’ really means. I offer these comments by way of suggesting that there used to be an empirical method that was quite detached from theorizing.

    Science, many will argue, only begins as we bring these two kinds of knowledge into some kind of intimate and significant contact. By significant I mean that knowledge of one type can have irresistible impact on knowledge of the other type. After the rise of logical positivism and the hugely influential work of Popper, Bacon’s notion of empirical science was transformed from being the collection of facts about the world and our doings within it, and became only a vehicle for testing hypotheses i.e. the empirical tradition became subordinated (prostituted?) to the task of warranting theoretical speculations. This is the essence of Popper’s notion of ‘falsification’.

    Yet we now know – from the work of Wittgenstein, and even Rorty – that this marriage of knowledge types has never been consummated, that neither verification nor falsification ‘work’ as the philosophy of science proposes. Indeed the entire modernist epistemological enterprise has been disposed of, as far as epistemology is concerned. There is no cheese down that tube.

    This is not a matter of saying there is no possibility of human progress – for we can clearly do things we could not before ‘science’ started work. But this turns out to be a comment about our practice, not about ‘what we know’. We may be in grave danger of confusing what we can do with what we believe we know about the ‘reality’ in which we assume we are doing it.

    Bottom line, then, is that if by empirical we mean what we know, historically, we can do, then there is surely much to be said for publishing that – as the vast bulk of the managerial literature is already doing. Case histories of doubtful ‘theoretical’ import they may be. But hey seem useful enough to attract most managers’ attention.

    But if we are really intent on subordinating ’empirical’ to mere hypothesis testing, it may be better that we do less of it.

  • 3. spostrel  |  27 August 2007 at 8:09 pm

    Interesting regularities are just that, and if they haven’t been noticed before, they’re worth publishing. The experience curve example is a pretty good one–who would be against publishing that? I would apply the same reasoning to the 80-20 rule or the diffusion curve.

    Note that all of these examples are a) pretty well-established though not universal and b) theoretically overdetermined. Collecting such things provides grist for follow-on research that tries to mill out the contingencies for when one set of causal factors versus another operates.

  • 4. JC  |  27 August 2007 at 9:23 pm

    Interesting regularities – or ‘stylized facts’? If, as Nicolai’s link suggests, along with Connie’s piece, that ‘stylized facts’ are ‘generalizations’, then how are they not ‘theories’? How can one ‘generalize’ from the particular without theorizing? A fact is a particular, a ‘happening’. It cannot be generalized without departing into a different mode of thought – which Bacon called ‘reflection’. Is the product of such reflection no more than a summation of the accumulated ‘facts’ – such as a computer might perform? Is this what statistics is supposed to do for us? And if the process of reflection adds something to the facts which is not in the facts (echoes of Kant’s synthetic and analytic here) in what sense can one advance a notion such as a ‘stylized fact’? For starters, is it a singular or a plural?

  • 5. JC  |  27 August 2007 at 9:28 pm

    OOPS forgot to ask Steve if he considers the ‘experience curve’ a theory or a ‘stylized fact’.

  • 6. Nicolai Foss  |  28 August 2007 at 3:33 am

    Steve, With all respect, you seem to be merely restating Helfat and Miller. Miller and Helfat argue in favor of publishing “interesting” “stylized facts” as is, apparently without any theoretical discussion. The points I was trying to make are these (plus some more):
    1) “Stylized facts” are presumably “Interesting” on the basis of some already existing background knowledge. Wouldn’t it be necessary to state the relevant background knowledge in a journal article?
    2) The observed empirical “regularity” may be spurious. To convince the reader that it is not, it may be helpful with a theoretically informed discussion of the mechanisms that produce the regularity.
    3) Paradoxically, it may be harmful to publish the regularity. If the regularity is some correlation between profitability and market share, well, we know that using this as a prescription for action may destroy value. (Would it have taken longer for the Philips curve relationship to break down if economists had kept it secret for policy makers? ;-)).
    4) I can see why there may be a role for reporting stylized facts as is in a atheoretical phase of research (the Fleming penicilin example is a good one here). But isn’t it the case that we are almost over-loaded with theory in management? Is it a harsh request to ask of a author that he does not “just” report a stylized fact, but also come up with a theoretical story that can support the finding? I think not.

  • 7. srp  |  28 August 2007 at 11:37 pm

    Surely this is a case-by-case matter.

    JC: The experience curve has been measured in many, many industries and firms. The power-law form seems to be pretty well supported as fitting the data. There are many potential processes that can generate power laws. I’ve seen chunking from cognitive psych, Kaufmann’s complexity model, and spurious confounds with volume economies cited as possible causal mechanisms. The question is still open as to what mechanisms operate in what situations.

    Peter: If we’re over-loaded with theory, why do you want people who report new regularities to drag still more theory into the arena?

    But that’s just a polemical point, because I don’t believe we’re overloaded with theory, at least with plausible, useful theory. I am certainly not advocating “dust-bowl” empiricism, but I don’t think our net of theory is so fine that only regularities clearly confirming or challenging that net deserve to be hauled in.

    Heck, what about the literature on Gibrat’s Law? Consigned to the flames for not being theoretical enough? .OK, I’m on a roll now. What about variance decompositions of profis by business unit, industry, corporation, etc.? What about industry life-cycle work, where the empirics have clearly led the theory?

    As for the argument that publishing a regularity may lead managers or policy makers to rely upon it inappropriately, you’ve got to be kidding.. Incorrect and partially digested ideas get warped and misused all the time, whether theoretically grounded or not. Just look at reengineering or core competence, which surely had very narrow empirical bases compared to their theoretical appeal. And in any case, the idea of referees and editors as Platonic guardians over what facts are too dangerous to promulgate is both impractical (how could they predict?) and damaging (potential for abuse and bias).

    Theory and observation are two feet that need to alternate if science is going to walk. If one foot always leads the other, then we can only skip or moonwalk, not particularly dignified or practical modes of travel over extended distances.

  • […] 23 August 2007 at 2:25 pm Nicolai Foss 7 comments […]

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 )

Connecting to %s

Trackback this post  |  Subscribe to the comments via RSS Feed


Nicolai J. Foss | home | posts
Peter G. Klein | home | posts
Richard Langlois | home | posts
Lasse B. Lien | home | posts


Former Guests | posts


Recent Posts



Our Recent Books

Nicolai J. Foss and Peter G. Klein, Organizing Entrepreneurial Judgment: A New Approach to the Firm (Cambridge University Press, 2012).
Peter G. Klein and Micheal E. Sykuta, eds., The Elgar Companion to Transaction Cost Economics (Edward Elgar, 2010).
Peter G. Klein, The Capitalist and the Entrepreneur: Essays on Organizations and Markets (Mises Institute, 2010).
Richard N. Langlois, The Dynamics of Industrial Capitalism: Schumpeter, Chandler, and the New Economy (Routledge, 2007).
Nicolai J. Foss, Strategy, Economic Organization, and the Knowledge Economy: The Coordination of Firms and Resources (Oxford University Press, 2005).
Raghu Garud, Arun Kumaraswamy, and Richard N. Langlois, eds., Managing in the Modular Age: Architectures, Networks and Organizations (Blackwell, 2003).
Nicolai J. Foss and Peter G. Klein, eds., Entrepreneurship and the Firm: Austrian Perspectives on Economic Organization (Elgar, 2002).
Nicolai J. Foss and Volker Mahnke, eds., Competence, Governance, and Entrepreneurship: Advances in Economic Strategy Research (Oxford, 2000).
Nicolai J. Foss and Paul L. Robertson, eds., Resources, Technology, and Strategy: Explorations in the Resource-based Perspective (Routledge, 2000).

%d bloggers like this: