16 January 2013

Handle With Care! - Ethics in Applying Behavioral and Decision Sciences


In the past ten years behavioral and decision sciences have gained a lot of ground in both research and practice. Some very popular books that have “translated” the academic knowledge have helped a lot in this process. Now, 10 years is not that much for either science or practice. It is not exactly the very beginning, but for sure there isn’t a vast history behind using behavioral and decision sciences in practice.

In this post I will address several issues that have to do with the ethics of applying behavioral and decision sciences. Since academia has its own ethics codes and regulations I will not address the research related issues. However, there is no real ethics code of applying behavioral and decision sciences in practice. Let’s take a look at some potential issues.

Can applying the insights from these sciences be avoided? Take for example choice architecture, can it be avoided?  

The fast answer to this question is “Yes” and the reason for this is that in our minds it is always possible to not do something. In other words, if something can “be done” it also can “not be done”. In reality, however, things are not that straightforward.

Let’s think for example of the so often encountered “agree to terms and conditions” that we find on websites. The “check box” can be empty (not checked) as a default or it can be full (checked) as a default. We know that “the default option effect” is very powerful, thus being or not being checked makes a huge difference. But can we actually avoid the use / occurrence of this effect?

As you probably have figured, avoiding this effect is a lot harder than we initially thought.

Things are quite similar for other “tools” of choice architecture. For example the mere existence of a choice set influences the person’s choice. People don’t usually choose something that is not in the choice set. In other words, whenever a person is presented with a choice set there are elements of choice architecture present even without any intent.

Another example is the way in which elements of choice architecture are hard to avoid is the way in which we present options to be evaluated. We can present them in single evaluation mode (one at a time) or in joint evaluation mode (all at once). The options could be presented in a mix of both ways such as first present them one at a time and then allow for comparisons by showing them all at once. However, even this might influence the choice of a person.

I guess you got the main idea. Even if one would want to not use choice architecture (or other means of influencing people’s choices or decision) it is not really possible to fully avoid it.

Not being possible to fully avoid using elements of choice architecture does not mean that we have to use the entire set of tools form behavioral and decision sciences. In fact it would be wrong to use everything, but that’s another story. For example if we can’t avoid having a “checkbox” with a default of “checked” or “not checked” this does not mean that we have to use more tools such as peer or  authority recommendations.

The question now is “should we use more than just the things we can’t avoid?”. My answer is “yes”, but we have to “handle with care”. Tools from behavioral and decision sciences have significant effects on human judgment, decisions and behavior. Sometimes these effects are not necessarily large, but they exist and more or less they affect everyone. Using these tools responsibly is not an easy job.

Can we “overdo it”? The answer is “For sure!”. One of the areas in which tools from behavioral and decision sciences are used is marketing. Companies and their marketers want to sell more, increase profits and so on. Everything is legitimate. At the same time using too much the insights from these sciences can back-fire big time.

One way in which using too much of behavioral sciences insights can back-fire is customers returning products. For example, if a (on-line or brick and mortar) retailer increases its sales by using too much of behavioral insights it is not unlikely that some of the customers will return products that were bought under the influence of the behavioral tools. Sure, not each and every customer will return what she bought, but it is enough that a critical number of clients do so. An avalanche of people returning (slightly used) products will create considerable troubles with logistics and cash-flow. Again it is not necessary that all customers return products; a couple of hundred is enough to create problems.

In an earlier post, I have discussed the issue of Regret in decision making and subsequent behavior. Regret is a very powerful negative emotion that comes from the realization that “I could have done better”. Some people might regret buying stuff out of impulse (or mindless shopping) and this negative emotion will be linked with the brand of the seller. This does not apply only to physical stuff, but to services also.

In essence, a marketer who uses behavioral insights to increase sales should keep in mind that there is a need for a balance between present outcomes (sales increases) and future outcomes. Again, it is not necessary that a large number of people to react in a negative way. Considering the fact that many markets are more or less stagnating and that profit margins are getting smaller and smaller it is enough to just “overdo it just a bit” to get bad results.

On a more moral philosophy note, there is an issue of morality in using too much of behavioral sciences insights. I remember reading an article which advocated “reducing consumer surplus”. Consumer surplus is considered to be the difference between what a person would be willing to pay for something and what she actually pays. This consumer surplus can be seen in a different light such as opportunity to save more for one’s children’s education. Companies what to “reduce consumer surplus” and from their point of view it is legitimate to want higher profits. At the same time there is a bit more to life than just increasing profits. Anyhow, that’s my opinion.  

In a similar note, in a consumerist world we are surrounded by “Stuff”, most of which we don’t really need. If you think about the “personal storage space” business sector, the only reason it exists is that people buy a lot of stuff that they don’t need. Of course, not over-buying is due to the use of “tools” from behavioral and decision sciences, but using too much of these tools can lead to useless purchases.

Unfortunately there is no “golden rule” that says “from here on it is too much”. What is “too much” is up each and every one of us to decide. As far as I know there is no “code of ethics” in using behavioral insights in practice. There is some (formal) regulations (at least in the EU) on some aspects such as subscribing to newsletters (default option is “no”), but there is a long way from this to a “book of rules”.

Richard Thaler and Cass Sunstein have proposed the concept of “Libertarian paternalism”. This apparent oxymoron is considered to be a “self-regulatory mechanism for using behavioral insights”. In essence, Libertarian Paternalism combines two major philosophies that are in opposition. Libertarianism supports the idea of the rational independent agent. In other words, it claims that leaving people and markets free the optimal outcome for both the individual and society will emerge.  At the other end of the continuum is Paternalism which supports the idea that it is best if “someone” (authority) tells people what to do and makes decisions for them.

The combination of these two opposing philosophies goes like this. We use paternalism to “steer” (or nudge) people into the right direction and at the same time we do not restrict free choice.

For example one can use “the default option effect” to increase the number of people who make the “right” choice, but people are free to change the “default option” into another one at their choice.

Richard Thaler and Cass Sunstein claim that by allowing for free choice abuses can be prevented. By abuse they understand using behavioral insights for “bad” purposes. Indeed, by not restricting free choice people can make their own judgments and subsequently act accordingly. I for one, however, am not fully convinced that free choice is enough to prevent abuses. In the end “steering” people’s decisions and behaviors can be done in either the “bad” or the “good” direction.

I conclude this post with a warm recommendation: “Handle with care!"

Like it?  Spice Up  Your Business


No comments: