Authors:
Forrest Briscoe – Pennsylvania State University
Chad Murphy – Oregon State University

Interviewers:
Ivana Naumovska – Northwestern University
Johan Chu – University of Michigan

Article link:       http://asq.sagepub.com/content/57/4/553

Question 1. You investigate the diffusion of controversial innovations–reductions in retiree benefits in Fortune 500 companies–and find that the adoption of easy-to-understand transparent practices (i.e., benefit cuts) reduces subsequent diffusion, while opaque practices (i.e., spending caps) increase diffusion. This is an interesting result! What happens, though, as opaque innovations increase in prevalence, perhaps attracting more attention and becoming better understood (less opaque)? Are there negative consequences for the further diffusion of these “opaque” practices? Do you see any strategies used to keep diffusing practices opaque?

Thanks! You raise an important issue for thinking about opaque practices: if an increase in their prevalence leads to more public attention placed on them, then the resulting scrutiny and resistance should indeed put the brakes on diffusion. And yet, we think that under certain conditions, practices can increase in prevalence *without* attracting more attention (in other words, without becoming more transparent). How is that possible? Two features seem critical, both having to do with the role of professional experts:

a) Causal ambiguity. First and foremost, professional experts construct the innovation in a way that (intentionally or not) makes it hard to identify it as a cause of subsequent events. For the cap innovation we studied, it’s not that the innovation was “hidden,” but the cap’s design made it hard to claim that accelerating cost increases born by workers and retirees were caused by the cap. Because of the cap’s inherent design, if you try to explain this connection, you’d have to use statistical and/or probabilistic concepts, and accounting and/or actuarial logic — and you’re quite likely to lose your audience by the time you’re done!

b) Professional networks. The second key feature is the structure of the professional community. When the professional community that develops a controversial innovation is structured so that it crisscrosses the corporate world, this provides a conduit for the innovation to spread across corporate “interiors,” without gaining much visibility on corporate “exteriors.” For example, nearly all large U.S. corporations have relationships with the benefits consultants who promoted the cap innovation we studied—and many firms also employ internal staff who are similarly trained. These agents-of-diffusion can understand the implications and consequences of the innovation, enabling them to make the internal “business case” for adoption… but at the same time, that understanding of consequences need not extend beyond firm decision makers to members of the affected groups (be they employees, retirees, or other kinds of stakeholders).

How often does this situation arise for other types of controversial innovations in the corporate world, beyond the case we studied? We suspect these two features we identified are actually widespread, and can be applied to the wider ecology of professional services swirling around modern firms. Lawyers invent corporate takeover defenses (e.g. poison pills), accountants invent tax shelters (e.g. corporate inversions), PR specialists invent reputation-repair strategies (e.g. astroturfing), political consultants invent congressional-influence pathways, HR consultants invent employee monitoring systems… in all these domains, we think there are examples of innovations where the “interior” understanding from within the expert community includes consequences that are functionally opaque to the negatively affected parties. That sets the stage for the diffusion of a potentially controversial practice, minus the attention or contention..

Question 2. Your data show that adoption of benefit cuts increased the incidence of interest group protests and negative news reports. You suggest that this was because transparent practices, which are easily described, are susceptible to cogent negative perception by adversely-affected interest groups and the media. Also, you manage to convey two antecedents of opaqueness (p. 557-558): (1) complexity in design and implementation, and (2) complexity in language when announcing or explaining the practice (i.e., framing). The former being related to substance, whereas the latter mechanism would be more in line with established theories of symbolic management and decoupling, i.e., when legitimacy is attained through symbolic actions that disguise non-conformity (e.g., Meyer and Rowan, 1977). Could it be that controversial opaque practices enable adoption by being ambiguous enough to allow companies to deploy their own and distinct framing strategies? Do you observe that in your data? (You mention, for example, that a retiree interest group allowed their company’s Benefits Office to explain the new policy to them.)

Great point! Each firm’s customization can contribute to an innovation’s practice opacity on all four of the dimensions we identified in Table 1 (“What is being done?” “To what degree is it being done?” “When is it being done?” and “Who is doing it?”). Greater variation across firms on those dimensions would logically make it harder for audiences to see the big picture—that is, to perceive the innovation’s overall consequences at the field level. That, in turn, would make it less likely that the innovation would spark higher-level opposition from policymakers, journalists, NGOs or activists.

Importantly, though, we are struck by how much the “upstream” design of the innovation matters, relative to the “downstream” framing processes that have dominated a lot of our literature to date. When practices have opacity baked into their design, the consequences are difficult to register no matter how you frame, package, or otherwise explain them. This suggests that invented practices can have essential qualities that shape the extent to which one can communicate about their impacts effectively. In particular, practices that are invented within the abstract theoretical domain of a particular area of professional expertise can have qualities that are readily understood through the architecture of that expertise–but which are effectively opaque through the everyday cognitive structures of lay audiences. Think of the mortgage innovations that spread across banks in the lead-up to the financial crisis, or the user agreements adopted by Apple, Facebook and other web companies. No matter how they are “framed,” such practices may remain unclear for all but the most indoctrinated.

Question 3. In your empirical setting, members of a retiree interest group expressed frustration with the large and confusing premium increases coming from the opaque spending caps (p.559). However your overall results suggest that possible frustrations about opaqueness are less consequential compared to grievances about clear cause-effect relationships. Is there then a level of “optimal opaqueness” – just opaque enough to blur potential negative consequences, yet not too opaque as to lead to vagueness-related contestation? Or is more opaqueness (to possible antagonists of the innovation) always better?   

It is interesting to wonder if vagueness itself could serve as a basis for contestation–perhaps with oppositional sentiments and collective action roused under the rallying cry, “End the Opacity!”

On the one hand, vagueness-related contestation seems quite plausible, as when powerful actors have been caught deliberately obscuring the consequences of the innovation they are adopting or endorsing. The term “obfuscation” comes to mind — just the kind of thing that would generate lots of negative press for adopting companies and therefore cool diffusion. Indeed, some of the retiree interest-groups leaders we interviewed held the view, and some journalists covering retiree benefits cuts conveyed this message as well. Beyond our study, recall objections to financial instrument complexity during the Occupy Wall Street movement, and earlier during the Enron scandal.

And yet, in our study, it did not really seem that vagueness or confusion was *itself* an important foundation for organizing resistance. Perhaps grievances based on vagueness or complexity have a hard time gaining traction with audiences because such grievances are themselves difficult to communicate succinctly. Or perhaps would-be activists are inhibited by their own embarrassment at not being able to understand the complex practices. Or perhaps it’s the barrier of proving that the opacity was deliberately concocted to deceive, and not just an incidental feature of the innovation. This would be an interesting topic for future research.

Of course, the point we raised earlier is critical here as well: a practice can be opaque on the “exterior” of the corporate landscape but transparent on the “interior” – implying that optimal opaqueness isn’t only a question of overall perception, but also a question of opaqueness *to whom*!

Question 4. Your data shed light on important questions of how the labor market changed in the United States, and what drove that change. Do you plan more studies with the data set? What other intriguing directions have the data suggested? 

How can we understanding the changing U.S. employment relationship and labor market overall? This is a major challenge, a “big hairy problem” that a growing number of scholars from management, economics, and sociology are all engaged in, in various ways.

For us, the Sleight of Hand study reinforced a growing suspicion that we need to approach these changes using a political lens. In particular, we need to apply a political lens to the firm’s stakeholders, defined as the constellation of actors that can affect or be affected by the firm’s decisions (investors, customers, suppliers, regulators, creditors, employees and unions, retirees, community groups and NGOs, and so on). For example, the corporate retirees who were negatively affected by the benefits reductions we studied are a weak stakeholder group, lacking in resources, power and organization vis-à-vis their former employers. In contrast, shareholders and executives, who had much to gain from the reduction in liabilities and increase in cash flow and operational flexibility, are well-resourced and powerful stakeholder groups.

All these major stakeholders are vying to influence the nature of changes in the employment relationship, through pressures they exert, directly or indirectly, on decision makers in the firm. So while it may be a truism that employment practices (like retiree benefits) will inevitably change, the nature of those changes is not pre-determined. The changes will create winners and losers among the stakeholders, depending on how they are designed. The process at its core thus involves stakeholders angling for influence, trying to expand (or at least preserve) their rewards as changes are implemented. We see this as another ripe area for future research.

Question 5. What question did we miss? Please ask yourselves a good question, and answer it.

Briscoe and Murphy (interviewers): Contentious diffusion is a cool topic, but there sure have been a lot of inter-organizational diffusion studies published already! After writing this paper, what future angles do you think are left for researchers to make a fresh contribution in this area?

Briscoe and Murphy (interviewees): Wow, good question. First off, we need studies that figure out how to incorporate a realistic political model of stakeholder pressure on firms (like the one we just described above shaping corporate employment practices) with a diffusion framework. The task is difficult, because stakeholders are a diverse group and they operate at multiple levels to influence firms, industries, and fields. On the other hand, data and analysis tools for this purpose are becoming ever easier to come by. And a toolkit of mechanisms may be emerging among scholars sitting at the intersection of social movements, stakeholder theory, and institutional analysis.

Second, we need work on how the *characteristics* of practices affect inter-organizational diffusion. Scholars need to dimensionalize and then measure these characteristics. Our findings concerning opacity may be useful by suggesting how a practice’s characteristics shape perceptions of its consequences, thereby affecting reactions to it. Future scholars could also integrate more insights from the psychology of diffusion (see “Switch” and “Stick”). But the key empirical need here is to find research settings where variations in practice characteristics can actually be observed and compared in their impacts on diffusion patterns.

Finally, we see great opportunities to understand the role of professional experts in constructing and spreading innovations across organizations. In particular, what’s the process by which professionals design and implement innovations that are externally opaque? In a sense, it’s the mirror image of occupational boundary spanning: instead of creating artifacts that help people communicate and understand one another across boundaries, these are innovations that erect barriers between professional and lay audiences. Take the financial statement footnotes we studied, which are often constructed in ways that disguise the causal agent of change (see example on page 568 in our paper). Researching this innovation-crafting process would probably require real-time access to the practitioners involved in it.