On the closure of Microsoft-Research Silicon Valley Lab

The closure of Microsoft-Research Silicon Valley Lab on September 19, 2014, seems to have taken the TOC community by surprise.

Having held deep concerns regarding corporate research labs for decades, I was not surprised at all, and had to ask my friends why they were surprised. It seems that the common answer is that the surprise is not at the mere (research-oblivious) decision to close such a lab but rather at the immediate implementation of the closure decision. Needless to say, several corporate research labs were closed during the last couple of decades, and the closure decisions were always related to "global" considerations of the corporations and not to the merits of the research done at these labs. The immediate implementation of the closure decision is indeed unprecedented when referring to (corporate) research labs, but is common practice in the behavior of these corporations. So what exactly is surprising here? Is it surprising that corporations treat the implementation of their strategy regarding research exactly as they treat the implementation of their strategy regarding other matters? Furthermore, is it a surprise given that they do base their strategy regarding research on the very same principles the are the basis of their strategy regarding other matters?

It seems to me that the research community can save itself some surprises by stopping to ignore the sociological aspects of its endeavors and the social frameworks that are relevant to it.

In the context of research institutions, one should ask what are these institutions; that is, what are their goals, what is their structure, and what are their ways of operation. By definition, universities are committed to research, since promoting research is one of their primary goals. It is also understood there that research is conducted by individual researchers and that one cannot manage the research activity itself (and should not even think of trying). Hence, in principle, universities have a non-hierarchical structure, at least when it comes to the research activities of those defined as PIs (i.e., "Principle Investigators"'; indeed, the definition of PIs may vary (cf., e.g., Europe versus US)). Needless to say, in reality, these principles (as well as others) may be violated, but it is important to take notice of the principles, since they determine or at least affect the "normal operation".

In contrast, the goals of corporations when establishing research labs is never clear. For sure, promting research per se is never the reason d'etre of these labs; the eason d'etre is always derived from main goal of the corporation, which is making money. Now, at times the corporation claims that it maintains the lab as a service to society, and by such declarations it hopes to benefit in terms of public relations. At other times, the corporation may reason that a good research lab is a way to attract good product-developers, and that the latter can be assisted by "internal consulting" from the research lab. And, yet, at other times the corporation may attempt to harness its researchers to work on product development. These three different attitudes may appear in mixed form (rather than in the pure form stated above), but the point is that these are fundamentally different attitudes and they imply fundamentally different ways of managing the labs, which in turn imply different operations of these labs. Furthermore, the shift from one attitude to another is always initiated by powers external to the labs; it may have to do with a change of the high management of the corporation or with a change in its "business environment". Note that only the first set of attitudes guarantees a (relatively) free research environment, which is reminiscent of the academic environment. But such a situation is quite unstable; it is often threatened by pressures from the higher management of the corporation, where the latter has an absolute decisive power.

To realize how unstable this situation is and how much different it is from an academic environment, one may read an account written in 2007 by Ron Levin (the MSR-SV lab director at the time). When reading this text, note how dependent this situation (or "model") is on the identity of the lab director. Also ask yourself how would the high management of the corporation react to such a text.

For sure, a stupid and/or nasty department head, faculty dean, or university president may cause the decline of the freedom of research (and research itself) in the relevant unit, but the effect will be several orders of magnitude milder and slower.

Sociological analysis is not exhausted by the analysis of corporate research labs. Indeed, there is much to say about the way the research community operates and about its relationship to the external world, but this seems a topic for separated posts. In particular, I have already expressed my opinion On Struggle and Competition in Scientific Fields and On Intellectual and Instrumental Values in Science.

The above does not mean that it is impossible and/or undesirable to benefit temporarily benefit from the confusion of these corporations regarding their goals; such temporary benefits are possible both on the individual and communal level. It is just that one should be aware of the temporary nature of this arrangement. Sure, life is temporary in nature (and universities may close too, or "just" close departments), but there is a fundamental difference between such terminations, where this difference reflects the difference in the defining goals stated above.

Back to Oded's page of essays and opinions or to Oded's homepage.