Friday, April 11, 2014

Nudge Database XI

See the left-hand column for links to previous editions of the Nudge Database.

101. The authors tested a voluntary self-control commitment device to help grocery shoppers in South Africa make healthier food purchases. The shoppers in this experiment, were already enrolled in a large-scale incentive program that discounted the price of eligible groceries by 25%, could risk their discount as a motivational tool. 

Of the total sample of 4,073 households, 632 chose to enter the commitment program. They pledged to increase their purchases of healthy food by 5 percentage points above their household baseline for each of 6 months. If they reached that goal, their discount was awarded as usual; otherwise, their discount was forfeited for that month. This ‘committed’ group showed an average 3.5-percentage-point increase in healthy grocery items purchased in each of the 6 months; households that declined the commitment ("Noncommited", N= 2,112) and control-group (N=1,329) households that were given a hypothetical option to precommit did not show such an increase.
The authors say these results suggest self-aware consumers will seize opportunities to create restrictive choice environments for themselves, even at some risk of financial loss.

Schwartz et al. (2014), Healthier by Precommitment, Psychological Science

102. This trial to encourage healthier grocery purchases ran for 5 weeks from 13 May 2013 in a Morrison’s store in Manchester. The store was decorated with a variety of displays, such as a cutout of the area’s local doctor (pictured), at strategically placed locations to prime the idea of healthier shopping. 

Data for a list of target foods was collected for the five weeks before, during and after the experiment, providing 15 weeks of data in total. Comparative data from five other control stores were gathered for the same period. Preliminary analysis indicates the intervention was successful at increasing fruit purchases.

Edwards (2013), Healthier Choices Pilot.

103. The authors used a natural field experiment  to test people's tendency to stick with the default option on printers by switching the printers' default settings in a large Swedish university from single to double page printing. 
The results are dramatic - one third of all printing is determined by the default alternative, and hence daily paper consumption drops by 15 percent due to the change. The effect is immediate, lasts throughout the experimental period, and remains intact after six months. 

Egebark & Ekström (2013), Can Indifference Make the World Greener?, Research Paper

104. The authors examine how taking a financial survey affects future financial decision-making. By exploiting randomized assignment to survey modules in the LISS Panel (a nationally representative panel study of the Dutch population), they find that households who respond to detailed questions on expenditures and needs in retirement reduced their non-household saving rate by an average of 3.5 percentage points. This effect is driven by high-educated, high-wealth households and is possibly due a salience shock, i.e. the people taking the survey didn’t realize they that were saving more than necessary until they took the survey.

Crossley et al. (2014), Can Survey Participation Alter Household Financial behavior?, IFS Working Paper.

105. Before 1997, students in the US applying for universities could send reports of their test scores to three colleges for free. An additional report cost $6 to send. In 1997, the rule was changed to increase the number of free reports to four, causing the number of people sending four reports to jump from 3% to 74%.

The author finds that the biggest benefit from this policy tweak was that after 1997, low-income students became more likely to apply to, and attend, more selective colleges. The lower-income students attending these colleges received a significant increase in their expected earnings.

106. The authors analyze the effectiveness of the the 2009 Credit Card Accountability Responsibility and Disclosure (CARD) Act on consumer financial regulation. Using a panel data set covering over 150 million credit card accounts, they find that regulatory limits on credit card fees reduced overall borrowing costs to consumers by an annualized 2.8% of average daily balances, with a decline of more than 10% for consumers with the lowest credit scores.

Consistent with a model of low fee salience and limited market competition, they find no evidence of an offsetting increase in interest charges or a reduction in access to credit. Taken together, they estimate that the CARD Act fee reductions have saved U.S. consumers $20.8 billion per year.

They also analyzed the CARD Act requirement to disclose the interest savings from paying off balances in 36 months rather than only making minimum payments. They find that this "nudge" increased the number of account holders making the 36-month payment value by 0.5 percentage points, with a similarly sized decrease in the number of account holders paying less than this amount.

Agarwal et al. (2013), Regulating Consumer Financial Products: Evidence from Credit Cards, NBER Working Paper

107. The authors examine the effects of receipts that included personalized ordering suggestions designed to reduce unhealthy food choices at a restaurant chain. They find that customers made most of the item substitutions that were encouraged by the messages, such as substituting ham for sausage in a breakfast sandwich, or substituting frozen yogurt for ice cream, though effects on overall calories and fat consumed were small. The results illustrate the potential of emerging information technologies, which allow retailers to tailor product marketing to individual consumers, to contribute in meaningful new ways to the battle against obesity.

108. The authors conducted a 2 year experiment in Chile designed to increase savings behaviour. Specifically they tested the effectiveness of weekly self-help style peer group meetings, text message reminders and the more traditional approach of a high interest account paying 5% (instead of the standard 0.3%). 

Surprisingly the high interest rate had no effect on the savings behaviour of most people. Those in the peer groups (which involved announcing their savings targets to other members) deposited money 3.5 times more often and weekly text message reminders (operating like a virtual savings buddy) were almost as effective.

109. The authors tested the efficacy of self-declared deadlines as a commitment device (N= 99 MIT students). While the control group received evenly spaced deadlines to complete three papers (about one paper a month), the treatment group were free to choose their own deadlines. There was no bonus or teacher feedback for early completion of papers so there wasn't an incentive to complete one paper quickly in order to improve later ones. 

A rational response would seem to be to declare all three of your paper deadlines for the last possible day - in fact only 27% of students did this. The no-choice group ended up getting better grades than the self-declared deadline group(88.7 v 85.6, p=0.003) and the no choice group also did better on a project that was also due on the final day (86 v 77, p < 0.01). 

Ariely & Wertenbroch (2002), Procrastination, deadlines, and performance: Self-control by precommitment, Psychological Science

110. Motivated by the fact that 9 out of 10 people support organ donation but less than 1 in 3 register as donors, the Behavioural Insights Team ran an RCT in conjunction with NHS Blood and Transport and other agencies. The trial ran for 5 weeks on a website with a sample size of over 1 million. 

After renewing their vehicle tax or registering for a driving license, the website users proceeded to either a 'control' page saying "Thank you. Please join the NHS Organ Donor Register" or one of many different 'treatment' pages which modified the message slightly using principles of social norms ("Every day thousands of people who see this page decide to register"), loss framing ("Three people die every day because there are not enough organ donors"), reciprocity ("If you needed an organ transplant would you have one? If so please help others") and others.

The reciprocity message was the most effective treatment, increasing the number of donors over the 5 week period by 1,203 compared to the control. This increase extrapolates to around 96,000 more signups per year. 
The Behavioural Insights Team, Applying Behavioural Insights to Organ Donation

No comments: