Is I/O Psychology Ruining Human Resources?

In a recent issue of Human Resource Management Journal, Godard1 provides a provocatively-titled opinion piece: “The psychologisation of employment relations?”  The central arguments of this paper are that 1) human resources management (HRM) is interdisciplinary, 2) industrial relations has historically been an important part of HRM, 3) organizational behavior has taken over HRM, pushing out industrial relations, 4) I/O psychology has taken over organizational behavior, pushing out traditional organizational behaviorists, and 5) these events have conspired to reduce the overall value of HRM. Clearly, such a paper requires a response, which was likely its intention.  There is no way for me to respond to Godard’s entire argument in a blog post, but I thought I’d hit some highlights.

Before going into Godard’s specific arguments, I think it’s important to note recent related goings-on within psychology to provide some context. Psychology is in general undergoing somewhat of a crisis, as the replicability and thus overall value of our research literature has being seriously questioned from within. Belief among US citizens that psychology even qualifies as a science is rather low. Within I/O psychology, the scientist-practitioner gap has been widening, driven partly by published research in mainstream journals (e.g. Journal of Applied Psychology, Personnel Psychology) becoming increasingly esoteric and impractical to apply in the real world, and driven partly by organizational belt-tightening, increasing the pressure on practitioners to use their time in a more billable fashion. These days, policies providing compensation to practitioners for publishing research are becoming increasingly uncommon, reducing the already-low publication rate by I/O practitioners. So things are not all well in I/O psychology, and I don’t think one can seriously claim otherwise.

Even given all that, Godard’s view is not reasonable to me for reasons I will explain momentarily. Here is Godard’s central concern:

What makes this potential takeover of particular concern is that it has been occurring at the same time that the study of labour relations (and trade unions) has continued to be narrowed and marginalised. Again, this has been especially so in business schools, where it has now increasingly come to be viewed as, at best, a subarea of HRM. Also important, and less noticed, has been the long-since-completed takeover of OB by I-O psychologists, and the corresponding displacement of the more sociological and ethnographic orientation associated with its main progenitor, the human relations school (Whyte, 1987). The result may be the gradual psychologisation of the study of not just HRM, but of employment relations in general (e.g. Sparrow and Cooper, 2003).

Let’s assume for now that this “takeover” has, in fact, occurred. Why does it matter? Again, I will let Godard tell you:

For example, a 2003 text by Paul Sparrow and Cary Cooper, entitled The Employment Relationship: Key Challenges for HR, has only 1.5 pages on employment law, nothing (that I could find) on labour law, trade unions or collective bargaining, and nothing on conflict other than the ‘intense emotional experience’ associated with a ‘breach’ in the ‘psychological contract’, and a brief (50-60 words) discussion of its
implications for exit, voice and loyalty (pp. 43-45). This text also contains only about a dozen lines on ‘employee involvement’ systems, all of which are in the context of their implications for commitment compared with those of job design. Although this book may be the exception rather than the rule, it is at minimum illustrative of the potential consequences of psychologisation should it continue.

So here lies the heart of the problem. With such an abundance of I/O psychology research driving OBHRM, more traditional (i.e. older) areas of HRM are being crowded out. The reason that this has occurred, Godard argues, is because 1) IO psychology’s attempts to simplify what are at their heart very complex problems more readily produces “answers” for management, which make it more attractive to both business school deans and managers-in-training, and 2) IO psychology has all the trappings of science but, in reality, is not a science. Instead, it is a sham. Another quote of interest:

Their research has six important components, all of which are largely consistent with this paradigm and ultimately with instrumental narcissism: (a) multiple authorships, with an extensive division of labour; (b) small-scale research questions; (c) extensive reliance on experimental research or survey methods, typically using students; (d) fixation on data analysis techniques, creating the appearance of scientifc sophistication; (e) extensive citation of other work; and (f) an absence of reflexivity…The consequence is that I-O psychologists do ‘better’ than their more traditional labour relations and HRM counterparts, who have traditionally done ‘messier’ institutional research that takes longer to come to fruition, and who are not as obsessed with citation rates. This is the case not just once they are hired, but also while they are in their graduate programmes, which seem to be designed mainly to generate publications rather than to ground students in substantive knowledge.

Ouch. Sort of fair – in that these have been and continue to be problems in parts of the I/O literature – but it is definitely overstatement, especially in regards to graduate training. The core of this statement is that “I/O psychologists oversimplify complex problems, using statistics that aren’t warranted to make their research appear meaningful when, in fact, it is not. They then pass this fake knowledge on to future students, who perpetuate meaningless research.” It is amusing to me that Godard would criticize oversimplification with such an oversimplification.

This type of oversimplification is common in Godard’s arguments, attributing to all of I/O psychology what is in reality much more complex. For example:

For example, selection and training courses increasingly focus on ‘soft skills’ having to do with attitudes and interpersonal qualities rather than technical and intellectual capabilities having to do with the actual ability to get things done. The result is a world in which employees are pleasant, but few have much of a clue as to what they are doing.

I cannot imagine a well-trained I/O psychologist advocating dropping technical skill requirements from the selection process. That is, to me, literally unbelievable. If technical requirements are dropped from selection procedures that already contain such requirements, I doubt it is the I/O psychologist making that recommendation. In the I/O psychology model, selection devices follow from job analysis, and well-done job analyses will contain technical requirements. End of story. I am not sure what type of I/O Godard has been observing to draw such conclusions.

I am left concluding from all of this that Godard is upset that industrial relations have been ignored in the growth of HRM to include I/O. Here is where we return to his idea that I/O has “taken over” HRM. In my read, despite Godard’s claim that HRM is “multidisciplinary”, this article is predicated on the assumption that there is only room for one perspective in HRM. That I/O has ruined that perspective and should be stopped.

This approach is bizarre to me. If business school deans are the problem, lobby to deans why industrial relations are important. If the research literature is the problem, build a literature more inclusive of the topics you find critical. If perspectives are missing in the I/O literature, add to that literature, altering and testing theory to include the components that are missing. That is the how the scientific method works. It is not an illusion we employ to make our work seem more credible. It is instead a flexible and powerful framework for testing ideas. It does not discriminate. If your industrial relations variables are really so important, add them to the models in the literature and see if you attain the dramatically greater prediction that you claim to be able to attain.

For all of Godard’s criticisms of I/O as explaining tiny amounts of variance that don’t ultimately matter to organizations, nowhere in this paper is the suggestion that our models be expanded to include such variables. Instead, psychology should be dropped entirely from HRM. That is, to me, is “throwing out the baby with the bathwater”.

Certainly I/O is overwhelmingly popular in HRM, but perhaps that is because I/O actually does bring value? Perhaps, despite its many failings (and I am first to agree that there are many!), it still manages to help organizations, managers, and employees to better meet their goals? Perhaps it is popular, and perhaps it is crowding out industrial relations, because the academic study of industrial relations has failed to demonstrate its own value to organizations? I imagine this is not a viewpoint that Godard considered.

More precisely though: Why is the solution to the “psychologisation of employment relations” to criticize psychology rather than to improve employment relations?

There are a couple of examples of this one-sidedness. The first is in Godard’s criticisms of an Academy of Management Annals paper entitled, “Employee voice behavior: integration and directions for future research”. Godard states:

Figure 2, reproduced from this paper, illustrates the problem. First, one is struck by the number of variables identifed, the arrows and boxes connecting them, and the lack of any effort to arrive at any deeper explanation for voice. Second, there is no identification of trade unions as a source of voice; indeed, the concept of voice in the figure appears to be entirely an individualistic one. The author does at one point acknowledge that there is an extensive literature in the field of labour relations and in HRM, but actually dismisses it on the grounds that authors in these fields ‘have not considered discretionary voice behavior, nor the causes and consequences of this behavior’ (p. 381). Third, the sole motive identified for voice is ‘to help the organization or work unit’. In this regard, the ‘integrated conceptualization’ that the author claims to have developed ‘from the various definitions in the literature’ deines voice as ‘discretionary communication… with the intent to improve organizational or unit functioning’ (p. 375). Apparently, neither interest conflicts nor injustice matter. This is only one illustration, but for anyone with an IR background it has to be an astonishing one, especially given that the topic is one that has long been central to IR as a field yet is now being psychologised (for more extensive critiques, see Barry and Wilkinson, 2013; Donaghey et al., 2013).

So where is the criticism of prior models from industrial relations lacking discretionary voice? Why not write a paper on “the industrialisation of HRM” criticizing how models in industrial relations never contain sufficient exploration of psychological variables? Let’s not forget that the purpose of theory is not to exhaustively describe a particular phenomenon, or else all theories would contain thousands of variables. As that classic quote by Box goes: “All models are wrong, but some are useful.” The simple fact that variables are omitted from a theory does not, by itself, condemn that theory. To believe otherwise is to majorly misunderstand the purpose of theory.

The second major example of this one-sidedness:

The argument in this article may be accused of con?ating theory with practice. There are undoubtedly many with I-O psychology backgrounds who are less interested in manipulating workers to managerial ends than in helping create the conditions for a higher quality of work experience and a productive working life. There are also many who do not fully adhere to the scientised research paradigm that seems to have become predominant, and even among those who do, many who may have misgivings about this paradigm – even if they are not sure why.

One can just feel the condescension dripping from that paragraph. “They’re all wrong; perhaps an enlightened few feel how wrong they are but just aren’t smart enough to realize why!” My counter question: Why are the methods of industrial relations so infallible? I suspect only because they are so familiar.

If I/O psychology ignores industrial relations, industrial relations appears just as guilty of ignoring I/O psychology. Neither perspective is helpful. Perhaps organizations (and their employees!) would be better served if these fields learn from each other instead of providing blanket condemnations of a frightening “other”?

1. Godard, J. (2014). The psychologisation of employment relations? Human Resource Management Journal, 24 (1), 1-18 DOI: 10.1111/1748-8583.12030 [↩]