CSR, Management, Politics & Society — March 6, 2015 at 1:02 am

Individual Social Responsibility: Why We Don’t Act in Line With What We Say

by

Just over 15 years ago, I was working with a global NGO helping them to engage with multinational corporations (MNCs). As part of this activity, I ran workshops in which MNC managers interacted with NGO managers on matters of mutual interest. These interactions were interesting in that many of the MNCs had quite enlightened policies in areas related to production – driven predominantly by their need to secure their supply chain, keep costs down and productivity and quality high. However, when it came to the vexing area of consumers, the common refrain was that the vast majority of consumers were simply not motivated by the fundamentals of how products were produced but that they focused on value, price, quality and more idiosyncratic product features that appealed to their personal use and interactions with the products. While MNC managers had a fairly clear line of sight between how they addressed concerns about workers and dealt with local governments and community groups and the performance metrics on which they were incentivised, addressing how consumers valued the ethical, moral and social linkages with the products they purchased simply fell into the too hard basket.

I was quite surprised by this, as I remember seeing quite a bit of discussion in the press and elsewhere on the rise of ‘ethical’ consumption. However, when I delved into the specifics of what was being discussed, I was very surprised to see that there was little, if any, substantive social scientific research that had been done addressing the psychological nature of ‘ethical’ consumption that went beyond superficial descriptors. The most sophisticated work was survey based and, as I will discuss, survey work on value related issues is quite problematic. Most research entailed asking individuals whether or not they would purchase products made under specific circumstances, such as items produced in unsafe work environments or using children in the production process. Unsurprisingly, this work found that nearly everyone asked said that his or her purchase would be influenced by such information. The validation amounted to little more than a passing note that various forms of ‘ethical’ product consumption was increasing, that there were more organic products in the grocery and people were purchasing more from charity shops and retailers marketing themselves as pro-social, anti-animal testing and so on. Some studies also found that individuals who had more ‘moral’ personalities based on psychological tests (also survey based), were more likely to indicate a tendency to answer that they would be influenced by information about the product’s provenance.

However, there was a massive disconnection between the magnitude of the purchase intention survey responses and the actual uptake of ‘ethical’ products in the market. Such surveys normally indicate that 70% or more of consumers are influenced by ‘ethical’ factors, yet such products rarely account for more the a few percentage points of actual market share. One manager stated it succinctly: consumers were “radicals when taking surveys and conservatives at the checkout line”. What we see here is a classic psychological and survey response bias known as the attitude-behaviour gap.

The attitude-behaviour gap can arise for many reasons but the most common reasons are two-fold. First, surveys are not ‘incentive compatible’. Incentive compatibility is a fancy way of saying that the information is extracted in a way that reflects the incentives that would exist when the individual is engaging in the actual behaviour; in this case, when purchasing a product. The second reason we might see problematic survey responses is that surveys are typically context free. A very good example of the first is the lack any trade-offs being made in surveys; essentially, one value is not being traded for another and there are no effective prices. We tend to think about individuals as having an innate sense of fairness that is immutable. However, this is simply not true. Research has shows that that fairness has both supply and demand characteristics, such that the more of it in the market the lower it is valued and the higher the price of behaving fairly the less people behave fairly. The classic example of the influence of context is the Stanford Prison Experiments. Despite the fact that individuals were randomly assigned roles as prisoners and guards – hence there is no psychological bias from people’s innate tendencies towards cruelness or mercy – the participants took on the characteristics demanded by the roles (context) into which they were placed.

There are, more problematically, more concerning issues. One is that perhaps the psychological measures of things like ethical disposition are erroneous in that they are based on stated survey responses rather than actual behaviours. One example of this is a study on trust done by Glaeser, Laibson, Scheinkman, and Soutter. What they found was that there was no relationship between psychological survey measures of trust and trustworthiness and how people played economic trust games, where trust is revealed by how much you are willing to leave for others in one period where they can reciprocate in the next period by leaving money for you. However, what they did find was that past ‘trusting’ behaviours forecast how people played in such games.

All of these factors together hint that a degree of scepticism is in order when one is dealing with issues of ‘ethical’ consumption. For my colleagues and me, the issue is not that individuals who take into account social aspects of their consumption do not exist but that they are likely to operate in niches rather than as the market norm. The reality is that most of what individuals look for in products is intangible – pleasure, brand, image, coolness, etc. – and there is no reason to believe that one of those attributes cannot reflect a need to meet pro-social norms or to be used to reflect outwardly what the individual want to be recognised as being, such as, ethical, caring, communitarian, etc. However, from what we know about consumer behaviour we also understand that consumption is an individual-level as well as a group-level phenomenon and that individuals are heterogeneous and the circumstances in which they consume highly varied. So the idea that there is an ethical consumer who is led by their morals in consumption is a rare thing. Yes, in specific circumstances, with specific products at certain prices, an individual may reveal themselves to be an ‘ethical’ consumer; but in different circumstances or with different prices or different products that same consumer may be anything but an ‘ethical’ purchaser.

We can see how this plays out by looking at a host of studies we have conducted across the globe in the intervening 15 years since my first meetings I mentioned at the beginning of the article. Unlike prior survey-based research, we focused on using structured experiments and ethnographic research where we could not only force people to make more realistic choices where prices and trade-offs existed but also observe them and understand why they did what they did (or didn’t do what they said they would do).

Three of my favourite examples of why it is difficult to make grand statements about ‘ethical’ consumption are given in three studies.

In the first study, we replicated a popular poll conducted by Christian Aid Abroad and the Mori polling organisation about what influenced people in making purchasing choices. The results of that poll (and our replication) showed that overwhelmingly, people said they were influenced by the conditions in which the products were made and the impact of those products on the environment. People also stated that they were willing to pay more for products that were more ‘ethical’ on these dimensions. However, like the study on trust discussed earlier, we found that these statements were virtually meaningless, with a twist. People who stated in the survey that did not care at all, behaved consistently with that viewpoint. Those who stated in the survey that they cared about how the product was made and whether or not it harmed the environment were no different than those that did not care. In other words, if people reveal that they don’t care they are more believable than if they say they do care. There is a clear lesson in this: Do not believe what individuals say they will do when there is a clear socially acceptable answer and no cost to giving that socially acceptable answer.

The second study looked at whether ‘ethics’ behaved like any other form of demand. In this study we manipulated two things, the price of the product and whether or not the product’s functionality had to be compromised to get the good ‘ethics’. Individuals were faced with a dilemma only when functionality was in conflict with good ‘ethics’. Across all product categories we saw two effects. The first is that demand for ethical products goes down as the price premium associated with getting the ‘ethical’ option rises. The second effects shows functionality trumps ethics. It shows that when there is a dilemma – i.e., ’good’ ethics requires a decline in the other attributes of the product consumers consider important – demand collapses to almost nothing. No price seems to matter when this occurs. Again, the lessons from this are clear: if you are attempting to sell an ‘ethical’ product you cannot expect individuals to sacrifice any aspect of the other things that matter.

The third study looked at the extent to which individuals who were amenable to some price premium for ‘ethical’ products in one product category were similarly amenable in another product category – where being ‘ethical’ means caring about something different. Here we found that only 11% of people fell into the socially concerned segment for one product category fell into that segment for another product category. This is telling in that every public survey of ethical consumption shows that people who say they are ‘ethical’ consumers invariably say they are ethical in all things. Again the lesson is very clear: it is highly dangerous to assume that knowing whether individuals are more pro-socially orientated in one product category, will tell you anything at all about how sensitive they are to pro-social orientation in another category. Indeed, it may be that such assumptions will be worse than non-informative and will lead to the wrong predictions.

In the past few years, we have taken this work into other realms. Two of our more recent projects examined ethical investing and whether or not social positioning helped in employee recruiting. The results, not surprisingly, revealed results similar to what we have seen in consumer studies. Again, we found that ‘ethical’ aspects of pension investments and workplace environments play a very small role in choice and that simple surveys massively overstate both the overall role that ‘ethics’ plays and very specific characteristics of individuals who are likely to make a choice based on ‘ethical’ factors.

As one looks at all of these studies, there is a tendency to become a bit depressed when examining the reality of human behaviour. However, we are not the only ones to find these types of results. What all of this work reveals is that simplistic notions of humans doing good just because it is good fail remarkably at representing our behaviours as consumers, workers and investors. We are complex creatures and the cognitive rules that we use when faced with the complexity of everyday life are not easily categorised into pleasant sound bites, however inconvenient that may be for policymakers and those peddling their own values as those best for the society at large.

Policymakers in particular need to be worried about the sort of work I have been discussing. There is a tendency to believe in simplistic solutions for hugely complex problems. For example, most policies aimed at energy conservation – e.g., smart meters – have been shown to fail with the population at large because they do not fit into people’s natural decision making logics. More recently, health policy experts have been arguing for adding calorie counts to fast food menus as a way of tackling obesity. Yet experimental and field studies clearly show that while people say it is a good idea, they ignore the information when purchasing. Indeed, few, if any, labeling experiments have shown that labels influence behaviour much at all since the information being added is not part of the consumers’ set of salient decision criteria. The logic of ‘nudging’, popular with the current UK government, operates based on the fact that individuals do not notice the nudge and eventually internalise the behaviour as part of a natural habitual response.

My recommendations are encapsulated in the lessons outlined throughout the article. However, one more is worth closing on. Individuals have spent years developing heuristic decision rules that help them deal with complex situations. One of the reasons that we find what we find is that when a person is shopping they are motivated by the goals and decision heuristics they have found worked for them when shopping. The same is true with looking for a job or investing in their pension. If one is going to change behaviour there are only two real options. The first is to get people to change the heuristics they use. In the short term could be done by adding in new options that account for the social aspects of consumption. In the long term it requires training from a very young age, where new heuristics are embedded in people’s upbringing. A good example of this is the attitudes to things like car safety, recycling and smoking. The second is to change the context in which those decisions are being made so that different decision rules are called into play. In one study we found that just by shifting the context from consumption for the family to education for their children we could get people to make very different choices, because the goal associated with the task of consumption were changed. This is an extremely difficult task to achieve outside a controlled environment, as the number of purchasing contexts is virtually infinite.

In The Myth of the Ethical Consumer, my colleagues and I talk about “consumer social responsibility” rather than ‘ethical’ consumption as it better reflects the idea that what matters is “individual social responsibility”. Focusing on consumption alone implies that it is the act of consumption that is generating non-ethical outcomes. When the reality is that what is ethically neutral are our cognitive rules for dealing with reality.

This article is a longer version of an article to be published by the UK Royal Society of Arts.

Comments are closed