Sustainability Q&A
Published on: May 16 2022

While there is now widespread acknowledgement among consumers that we all need to make greener purchasing decisions, there is still a gap between these intentions and how we spend our money. Dr Michael Smith, advisor to CloudArmy has recently published a book on this gap and how it might be bridged - Inspiring Green Consumer Choices: Leverage neuroscience to reshape marketplace behavior (2021, Kogan-Page, London). We sat down with Michael to ask what researchers have learned about this subject and how such questions can be addressed with Implicit and response speed measurement techniques.

Q: Do consumers exaggerate their stated intentions to buy more sustainable products and services and engage in things like recycling?

Well, it does seem to be the case that when asked most people will espouse intentions to engage in pro-environmental behaviors of one form or another, and most of those same people probably actually do engage in such behaviors on at least some occasions. But in many cases such intentions can result in little or no action. This does not necessarily reflect deliberate exaggeration of intentions on the part of the people being asked, nor is this disconnect between stated intentions and actual actions unique to the sustainable consumption realm. Efforts to decode this sort of “say-do” gap are endemic to many areas of research on consumer decision-making.

A large body of social science research on behavior change has indicated that, for many reasons, even honestly stated intentions can often fail to lead to corresponding actions. As a result, several prominent frameworks have been put forward in efforts to understand how behavioral intentions form and are transcribed into corresponding actions.

One such widely cited framework is Icek Azjen’s Theory of Planned Behavior. Azjen characterized behavioral intentions as having a probabilistic rather than deterministic relationship to action, with a stronger intention generally increasing the likelihood of an associated behavior. In his view several key psychological factors contribute to the strength of an intention. First is simply one’s attitudes towards the behavior in question – would it be consistent with one’s self-concept, beliefs, values, etc. A second factor is more closely related to perceptions of the prevalent social norms around the behavior. Do you observe other people doing it (or not doing it)? Or do you think that your friends and peers would approve or disapprove of you doing it? And a third key factor relates to feelings of self-efficacy. That is, do you think that your actions would matter in the big picture? If you have concluded that whatever actions you take are inconsequential, your motivations for engaging in those actions may be undermined. Collectively, in any given situation, these psychological factors are thought to work together to either reinforce or undermine behavioral intentions, which in turn increase or decrease the likelihood of corresponding actions.

A large body of research from many different domains has accumulated which has shown that Azjen’s framework can provide a good fit to data concerning stated beliefs and attitudes, stated intentions, and either self-reported or objectively measured actual behaviors. Such research has also highlighted some gaps in the framework that may also be important for understanding discrepancies between intentions and actions.

Q. Are there any examples of research in the area of sustainable consumer behavior that support this perspective?

Yes, there is a growing body of evidence that similar factors influence pro-environmental consumer choices. Let’s first consider the notion of the sustainability intention-action gap itself. Global survey data has in recent years documented a growing public concern over the environmental consequences of consumer behavior, with a majority of respondents stating an intention to modify their behavior in response. Yet only a more dedicated minority indicate that they have stronger or more confidently held intentions of this nature and frequently try to make “green” choices. So, there is self-reported evidence of an overall intention-action gap in the population. And less subjectively, despite a majority stating an intention to change their behavior, consumption, waste, and environmental degradation are growing steadily year-over-year. So, there is considerable physical evidence of an overall intention-action gap as well.

Over the last decade several empirical studies of green consumer decision making, inspired in part by Azjen’s framework, have been conducted across global markets and reported in the peer-reviewed literature. These studies have combined survey methods with consumer choice measurements and statistical causal modeling methods in efforts to better understand the factors that contribute to the sustainability intention-action gap. In general, these studies find that psychological factors such as pro-environmental attitudes, beliefs about prevailing social norms, and beliefs about self-efficacy in terms of personal behavior are positively related to both the strength of “green purchase intentions” as they are sometimes called, as well as actual choice behavior where such data were available. But this research has also shown that the correlations between these factors are often moderate at best, and that they explain only a portion of the variance in outcome variables. They have also shown that some underlying psychological variables not included in the canonical framework can negatively impact such outcomes. For example, factors such as low levels of general environmental knowledge or general suspicion of sustainability-related product claims and the potential for greenwashing have both emerged as factors contributing to lowering green purchase intent and actual product choices.

And a variety of factors that are less psychological in nature have also been identified as barriers to pro-environmental decision-making and hence as contributors to the sustainability intention-gap. For example, if a “greener” product choice is also much more expensive than a more traditional choice, many consumers may not be able to afford that differential despite their best intentions. Similarly, the effort (in terms of time, search costs, etc) required to adopt a more sustainable alternative may be judged as too great relative to the perceived benefit of the choice. And a related issue is just the availability of the alternative and any infrastructure needed to support it. For example, I might want to have a heat pump installed at home for heating and cooling purposes, but if there are no local HVAC specialists in my community who have experience with the technology and I’m not handy enough to do it myself I might not be able to get one despite my intentions. Or one might want to switch to an electric vehicle but live someplace that lacks adequate charging infrastructure to support it. To acknowledge such factors researchers in this domain have begun to build on the basic motivational framework implied by Azjen’s work to develop models that also attempt to quantify the ability and opportunity for consumers to act on their intentions in order to create a more fleshed-out view of the factors that conscribe pro-environmental choice behavior.

Q. Aside from a desire to look 'good' or socially acceptable are there any other reasons why answers in regular, traditional style surveys might be misleading when it comes to green consumption issues?

Well, all the typical litany of criticisms about the limits of traditional survey methods for understanding consumer behavior certainly apply here. It is becoming broadly understood that people have little in the way of direct and veridical introspective access to their cognitive processes or implicit emotional reactions. One consequence of such limited self-insight is that overtly stated responses to survey queries about the drivers of behavior can often reflect post hoc judgements of plausibility rather than any personal truth. People also tend to underestimate how much of their routine day-to-day activities are guided by automated habits rather than deliberate intent, and those habits may or may not reflect pro-environmental behaviors. Consumers further tend to underestimate how much physical or mental effort they are actually willing to expend to change their existing behavior patterns. Finally, there has been widespread discussion in recent years about the wide variety of mental heuristics or cognitive biases that routinely color judgment and decision-making in ways that typically go completely unnoticed by the people making those judgments and decisions. These biases demonstrably influence the manner in which people respond to questionnaires, including questionnaires that gauge intent to engage in pro-environmental behaviors.

Since the issue of how cognitive biases influence choice behavior has been discussed at length by the marketing and behavioral science communities, I won’t belabor it here. Except to consider it with respect to the sort of prosocial response predisposition you allude to in your question. So, when reviewing results from a nationally representative survey on attitudes towards food waste conducted in the US some years back, it seemed clear to me from the results that respondents had concerns about the environmental and social impacts of food waste, understood common ways to reduce it, and that they professed a desire to minimize their own such waste. All very reasonable. But when asked to gauge how much food they wasted relative to other people, around 75% of respondents claimed they wasted less food than the average person – which just on the surface of it can’t accurately reflect the reality of the situation. This struck me as an interesting observation, and so I dug into the literature on this “better than average” effect and it turns out it seems to be a pretty common observation when people are asked about their own sustainable consumption activities across a variety of types of pro-environmental behaviors. One interpretation of this is just to attribute it to a basic prosocial bias on the part of respondents, who strive to look good both in their own eyes and in the eyes of whoever is doing the questioning. But maybe there is more going on.  

For example, back in the 1970s Amos Tversky and Daniel Kahneman identified a cognitive bias they called the “availability heuristic”. This heuristic refers to the relative ease by which relevant information can be retrieved from memory or otherwise brought to awareness, and it reflects an automatic inference that if something comes to mind relatively easily when considering a specific topic or problem, then that conceptually salient information must be relatively important and representative of the facts of the real world. The better than average effect might at least in part reflect the influence of the availability heuristic. If you are asked to judge how likely you are to engage in some pro-environmental behavior, you might easily remember that you tossed something into recycling yesterday or that you brought a re-usable shopping bag to the store this morning rather than opt for a single-use plastic one. In contrast, when asked to compare your own tendencies to that of some abstract average “other”, you don’t have available examples in mind of such behaviors on the part of that vague comparison point. So, reporting that you are more sustainably virtuous than average might just be a default judgment when the availability heuristic is at work rather than something that solely reflects social pandering.

Q. How might implicit and response-speed research techniques help to better understand and overcome these effects?

Research techniques that use indirect methods to infer the attitudes and beliefs of respondents can be critical complements to more direct queries in that they can help circumvent cognitive biases and the limits of conscious introspection and thereby provide information that consumers are unable to accurately articulate. The value of such methods is clear for most types of consumer insights problems, including the problem of understanding drivers of, and barriers to, sustainable consumption.

There are lots of examples in which reaction time-based measures in particular have been used to glean insights into this problem domain. For example, one agency recently reported that they were able to use fast choice methods (as a proxy for decision confidence) to segment a representative UK sample into a group that had strong pro-environmental beliefs and were likely to engage in related behaviors (based on rapid agreement with associated statements), a group that was in agreement with the statements but apparently were less confident in that agreement, and a group that didn’t identify as pro-environmental. They were then able to propose messaging strategies especially suited to the proclivities of each group.

In another example, in our own collaborative research utilizing CloudArmy’s Reactor platform we have used evaluative priming methods to identify the implicit associations that respondents appear to have towards imagery associated with wind and solar energy sources versus imagery associated with fossil fuels. In that case we found that the former was positively associated with a clean future, whereas the fossil fuel imagery was relatively more associated with a dirty past. Such differentiation in associations might be effectively leveraged in marketing campaigns targeting renewable energy adoption. We have had similar success in evaluating the strength of implicit consumer attitudes towards eco-labelling symbols intended to communicate energy efficient appliances.

More complex reaction time-based tests such as the traditional “implicit association test” (IAT) have also been used in this domain. For example, in one published study researchers examined implicit associations to descriptions of traditional products versus product descriptions that included sustainability claims, and found that with respect to expected efficacy traditional products might be implicitly associated with the notion of strength whereas environmentally-friendly product attributes tend to activate notions of weakness or gentleness. A basic finding that we were able to replicate in a new sample using Reactor. Such an insight might be used to optimize communications targeting consumers who are looking for performance on one or the other end of this dimension for some target application. Another novel application of the IAT method has been to gauge the degree to which individuals feel “connected to” or “part of” the natural world, that is, to indirectly measure the sort of personal attitudinal predisposition that might be difficult for an individual to accurately explicitly report on. This is useful because in general harboring a personal sense of being part of nature appears to be a good predictor of one’s likelihood to engage in pro-environmental behaviors, and so having a tool to implicitly measure such a sense of connectedness can be useful in research aimed at identifying interventions effective at increasing such feelings, which in turn might serve as strategic interventions in behavior change campaigns that aim to promote some type of related target behavior.

Q. How might behavioral science concepts, like 'nudges' help with creating solutions?

Behavioral science-informed interventions have been widely employed to encourage pro-environmental behaviors, in both commercial and noncommercial settings. One challenge with encouraging people to engage in pro-environmental behaviors is that asking them to do things differently often also means asking them to do things like expend extra effort, overcome existing habits, incur greater expense, switch from a trusted brand to an unfamiliar competitor, and so forth. All things that people are generally loath to do. However, we know from the behavioral science literature that if you can just make the target behavior or desired choice easier in some way by removing barriers to doing it while increasing motivation to engage in it, then change can occur quite readily.

A great recent example of a sustainability marketing effort that in my mind seems to exploit this basic fact is that of P&G’s 2021 #TurnToCold campaign for its’ market leading Tide brand. It encouraged consumers to start washing their clothes in cold water by default and that by doing so with Tide they could expect cleaning efficacy equivalent to warm water washing. And the campaign highlighted the direct personal benefits of doing so, including significant financial savings from both reduced energy use and by helping clothes to last longer. That is, P&G didn’t ask people to switch their brand allegiance from a trusted product or to do anything more difficult or expensive than what they were currently doing, and they provided a simple path to direct personal rewards to increase motivation for making the change. Of course, the reduced energy use has the side benefit of reducing greenhouse gas emissions, and the longer lasting clothes keeps textile waste out of landfills. But in this marketing approach the impersonal environmental benefits can be seen as complements to the direct personal benefits enjoyed by consumers rather than trade-offs.

It is also important to realize that reaction time-based approaches such as those we were discussing earlier can help inform the design and optimization of behavioral science inspired interventions. As a result, the two fields of research and associated practices are highly complementary and can work synergistically to develop great behavior change campaigns. Let me give you a hypothetical example. As I noted earlier, prominent theories concerning the psychological factors that can reinforce behavioral intentions often incorporate judgements as to whether or not a certain course of behavior is congruent with prevailing social norms. As a result, many behavioral science-informed behavior change interventions have used marketing communications that convey information about prevailing social norms as nudges intended to shape the way consumer choices are made. And in general, prior research has shown that such interventions can sometimes be quite effective. But how does one know if normative concerns are even an issue for a specific target population in a particular context? And if they are, how does one know how to best phrase a communication about prevailing social norms in order for it to have maximum impact? Reaction-time based measurement tools can be used to reach some clarity on both of these types of issues. That is, such tools could be used to both uncover implicit concerns about social appropriateness that people may harbor even if they aren’t consciously aware of them, as well as to help identify messaging frames that are most impactful in terms of addressing such concerns. There are lots of other analogous types of problems where such approaches can be similarly complementary. Accordingly, I think that marketers who want to be successful at creating effective campaigns to promote pro-environmental behaviors would be well advised to combine both types of approaches when developing and executing such campaigns.

Michael E. Smith, PhD is an advisor to CloudArmy and author of the book Inspiring Green Consumer Choices: Leverage neuroscience to reshape marketplace behavior (2021, Kogan-Page, London)