Does Corporate Strategy Still Matter?
Posted: October 23, 2011 Filed under: competitive advantage, Corporate strategy, Mergers and acquisitions, teaching 5 CommentsTeppo recently asked us whether the fundamental questions of strategy have changed since Rumelt, Schendel & Teece’s classic work. Relatedly, Mike wondered if strategy has lost sight of foundational questions and is now ceding territory to Economists.
One critical shift has been away from corporate strategy (multi-business firms and M&A). To an extent, this was fueled by debates over whether industry matters (see classic articles by Rumelt and McGahan and Porter) as well as the rise of the resource-based view. Lost in the shuffle were the prospects for corporate strategy research…
This has practical implications. I’m about to begin a module on corporate strategy (can you tell it’s my teaching semester) and this question looms large. I need to justify to my students why, given miniscule corporate effects, I am spending so much time on this topic. I think it points to a fundamental flaw in the way some of this research has been interpreted.
Consider McGahan and Porter’s finding that the corporate parent effect is a paltry 4% — hardly worth our time, perhaps. This means that in a very small number of cases, all of a firm’s business units seem to perform better (or worse) presumably due to the management of the corporate headquarters. This logic works fine if the HQ has already worked it’s magic on the business units (like GE selling off any units that fail to be #1 or #2 in their industries).
What if the portfolio includes units that are in various stages of transformation as the HQ applies their capabilities? In this case, variance decomposition methods would ascribe much of the variance to the business unit level but that might vastly understate the actual contribution of corporate capabilities. Indeed, very few theories of corporate value would lead us to expect that all units would simultaneously out-perform their competition. As new businesses are acquired, operational synergies are sought or inefficient businesses are transformed. At any given time, the portfolio should include great variance in performance.
Selection bias may be a problem as well. If a firm acquires under-performing firms and increases operational efficiency, a common strategy would be to sell off the best performing units and use the cash to repeat the cycle. The best performers are systematically culled and our research methods may lead us to conclude that corporate parents are not adding value.
Despite hundreds of COMPUSTAT-driven diversification studies, there is still much to learn. Looking around us, it should be clear that firms continue to struggle with corporate strategy problems. We continue to see some firms splitting off into smaller units even as others bulk up with unrelated diversification. Consider Netflix’s recent decision to split the company only to reverse itself a few weeks later. And what’s with HP anyway?
All this points to a need to return to this fundamental, but unanswered, question. What would a new research agenda for corporate strategy look like in a post variance decomposition world?
StrategyProfs.net 
Presumably, it would involve empirical methods that look very carefully at causal identification. Methods have advanced significantly beyond those used in the classics — diff-in-diff, instruments, field experiments — many cites in the Syverson paper point the way.
Russ, great questions. Following on Mike’s comment, I’d suggest looking to the empirical corporate finance literature on diversification, which has been dominated over the last decade by concerns about identification. We know, for example, from work by Judy Chevalier, Belen Villalonga, and others that parents often acquire underperforming subsidiaries and make them perform better than they would have had they not been acquired, even though their post-acquisition performance is worse than that of non-acquired industry peers. This is the kind of corporate effect that doesn’t show up in variance-decomposition without controls for unobserved heterogeneity and the endogeneity of the parent-subsidiary relationship.
Sorry, make that “variance-decomposition studies.”
Hi Russ, as they often say, teaching is a good way to remind oneself of what one does or does not think about a subject. As I get back into writing about and teaching strategy, which I do in Paris, I have been trying some re-thinking too.
I see the basic question ‘does strategy matter?’ as narrowly tautological – perhaps oxymoronic.
Strategy matters by definition. It is always about what matters – if it doesn’t matter it is not strategic. More precisely, strategy is about the consequential that is in your power to change i.e. it is about (a) having a choice, and (b) that choice mattering to you. If there is no choice, there is no strategy. If there is choice, but the outcome is inconsequential to you, then the choice scarcely warrants being called strategic – given that so much of what goes on in one’s life is strategic. All the time and everywhere what we do ‘has consequences’.
The challenge is to find a description or articulation of our own real-world situation that identifies those consequential choices. Strategy is about us, not the situation, which cannot ever be fully articulated.
I’m reminded of a line from Love’s Labor’s Lost, which we saw last night in an uproarious production:
“A jest’s prosperity lies in the ear of him that hears it, never in the tongue of him that makes it”
So I think ‘doing strategy’ must always begin with one’s definition of the situation (which can never be complete, objective, unambiguous, or non-idiosyncratic). The definition also implies or articulates the choices that seem to be open within it.
One problem with some of our literature is that what is strategic is defined in terms of what ‘top management’ does – as if their choice of shoe or airline was strategic. Likewise our literature often neglects the consequential nature of the everyday choices being made by non-executives far from HQ – even though other disciplines, such as the law, recognize the crucial and consequential nature of precedent.
The fundamental question – which can scarcely have changed since the year dot – is not about ‘does strategy matter?’ but about whether we – as individuals, business managers, or firms – have choices.
As I see it, any talk of strategy spins around not fully knowing a situation. But what is this situation and where are its choices? It follows that for managers ‘strategy’ may be something very different from what it is for game theorists. The latter seek out the ‘complete’ rules covering their game – that is their chosen ‘unknown’ – and they call their answer that game’s ‘strategy’.
Business people’s strategizing likewise relates to what they do not know, and that, in turn, relates to the choices they feel they must, or are able to, make. But who are the strategists that interest us? The answer relates to some specific function within the firm. But given we have no generally accepted theory of the firm, we are still in some doubt about what the management’s contribution is, and so where in the firm its strategizing (consequential choosing) is located.
I think strategy is about the specific choices that seem to need to be made to meet the situation’s immediate needs, and these are not always obvious. We normally presume business strategizing is about keeping the firm going, maximizing its growth opportunities and so on. But is this really the case in real-world firms? Do all firms want to grow? Do all the firms in any particular sample have the same notions about discounting anticipated risks? And so on. Our field’s normal assumption is that the strategy is about optimizing the resource-allocation decisions made by the top management team. But can we really be of one mind on this? Is the movement of funds between product/markets the same kind of choice as that involved in M&S decisions? Are R&D decisions strategic in the same way that granting supplier credit to a new customer might be?
One possibility is that ‘strategy research’ is simply an academic game we play to show our ability, or lack of it, to create some hypotheses based on the presumption that the firms in our sample are in fact comparable when, if we actually got into any of them, we would be forced to admit that they are as different as chalk and cheese – even when they are in the ‘same industry’ – whatever that is supposed to mean. This game, and its ‘fundamental questions’, may have no ‘fundamental’ relation to business, the economy, competition, or to managers’ work.
One of the interesting things going on in our field is the new attention to ‘business models’ – driven, of course, by an appreciation of their variety and complexity, as opposed to the micro-economic tradition which tends to imply managing is the generic if not universal activity of rational decision-making about a very narrow range of issues.
Yet when we look at something like corporate performance how do we really know what we are looking at – strategically speaking? Barbs from researchers like Ijiri & Simon or Hannan & Freeman suggest firms, as holistic entities, may have fewer choices than we think. If we think in terms of business models, why would we think it useful to compare and contrast the results of very different business models in very different industry situations? What do we think we can discover in this research frame? (I’m biased on the subject of ‘industry effects’, of course.)
In short, if you think of a firm as an assembly or portfolio of business opportunities, or business models, where do you want ‘strategy’ to lie, in the portfolio management or in the business operations themselves – and what kind of intentions, perceptions, aspirations and decisions are involved? Why would you think them comparable?
[…] Here’s Russ Coff on whether corporate strategy still matters. […]