However, there are several limits of opt-out policies:
- Since decisions are made through inaction, they are less likely to result in the kind of committed follow-up that is often useful for implementing decisions. This is particularly an issue for choices which require ongoing engagement, rather than those which just require a one-time decision.
- Opt-outs are most likely to be useful when there is a single optimal cause of action for most people – when different options are best for different people, the opt-out can be much less beneficial.
- Opt-outs may be considered unethical or illegal in some cases – e.g. many employees would consider it unethical if employers offered something like an opt-out ideal weight plan in their benefits package
This is best illustrated through a couple of examples, which come from the four experiments they run in the paper.
Flu Shot Example
The first example, comes from a lab experiment with 55 employees from an educational institution, in which participants were told “We would like you to imagine that you are interested in protecting your health. Your employer tells you about a hypothetical program that recommends you get a flu shot this Fall and possibly save $50 off your bi-weekly or monthly health insurance contribution cost”. The three choice structures tested (and the result) were then:
- (Opt-in) “Place a check in the box if you will get a Flu shot this Fall – 42% ticked the box
- (Active choice) “Place a check in one box: I will get a flu shot this Fall or, I will not get a flu shot this Fall”: 62% ticked the box for a flu shot
- (Enhanced Active Choice) Place a check in one box: I will get a Flu Shot this Fall to reduce my risk of getting the flu and I want to save $50 or, I will not get a Flu Shot this Fall even if it means I may increase my risk of getting the flu and I don’t want to save $50: 75% ticked the box for a flu shot.
This experiment was done as a field study with 9950 CVS pharmacy customers. The treatment was done as a recorded telephone message sent to participants, to see whether they wanted to enroll in an automatic prescription refill program. The two treatments were:
- (Opt-in) press 1 if they wanted to be transferred to Customer Care or to press 2 if they were not interested in enrolling in the automatic refill program: 15% took-up the new program
- (Enhanced active choice) press 1 if they preferred to refill their own prescription by themselves each time or to press 2 if they preferred the PBM to do it for them automatically: 32% took-up the new program
What makes Active Choice, and particularly Enhanced Active Choice, effective? The authors note several channels at work. First, the authors note that cognitive dissonance and regret is lower among people who are automatically enrolled into programs because defaults allow them to avoid cost considerations – so enhanced active choice can lead to greater loss aversion for the new opportunity. Second, they note that reminding people of what they will lose if they opt for the non-preferred alternative can have a powerful impact on choice because decision makers are unlikely to seek out information about the costs of remaining with the status quo without prompts, especially if such thoughts evoke negative emotions like anxiety and regret. i.e. Active choice is effective because it prompts regret aversion - the desire to minimize future regret from not complying with the option preferred by the communicator (e.g., getting a flu shot). Finally, they note that other literature suggests that active choice may produce higher levels of perceived responsibility and satisfaction than opt-out, leading to longer-term rates of ongoing adherence.
So here is another example where relatively small modifications to the way a choice is presented can have large impacts on the take-up of a program. While, like much of behavioral economics, it seems obvious ex post, low take-up is a problem we face in many development programs, and not nearly enough attention is devoted to the “choice architecture” or way in which we present these choices to individuals. This seems a ripe area for experimentation in many of our World Bank projects, changing the way we offer business training programs, HIV screening, financial literacy courses, and the like. Any TTLs who would like to try this, let me know.