This week is all about transparency: how to build it, how it affects consumer behavior, and its potential benefits (and downsides). Notably, the papers use a blend of lab experiments, natural experiments, and field experiments to illustrate how transparency builds trust—and when it can do harm.
Operational Transparency and Government Trust
Across three studies, this research paper examines an important context for applying operational transparency: government services.
In the first study, the researchers collaborated with the City of Boston, following over 20,000 residents over a two-year period. During this time, residents submitted service requests for fixes (i.e., filling potholes, painting over graffiti) to the city’s Public Works Department via a smartphone app. The city decided to launch a new version of the application, with a new layer of transparency: images of the work that was completed by government employees. Due to the rollout of the new app, some city workers were able to post the images, and others were not. Since the posting of photos was essentially randomized, the data acted as a natural field experiment. The researchers analyzed how behavior changed after residents were exposed to an ‘operationally transparent’ image. They found that operational transparency increased citizen engagement significantly over 13 months, resulting in residents submitting 59.8% more public service requests in 37.7% more categories.
In their second study, the researchers partnered with Code for America to create three versions of “The Daily Brief,” an open source website which provided transparency into resident-submitted service requests. The ‘control’ version of the website just listed the service requests in each category. The ‘responsive transparency’ version of the website represented each service request open and closed the previous day with a pushpin graphic, overlaid on a map of the city. Finally, the ‘unresponsive transparency’ version of the website visually depicted the backlog of unresolved requests. The researchers recruited Boston residents to a behavioral lab to interact with the websites. Participants randomly assigned to the ‘responsive transparency’ condition reported higher levels of trust in government relative to the control. Interestingly, participants in the ‘unresponsive transparency’ condition were no less trusting of the government relative to the control condition, suggesting that drawing attention to negative information (e.g. a service backlog) doesn’t lead to a backlash. These results were replicated in a final online study.
Why does this research matter? Given that public trust in government is at a record low, interventions that have the potential to increase civic engagement and rebuild trust are more important than ever.
Buell, Ryan W., Ethan Porter, and Michael I. Norton. “Surfacing the Submerged State: Operational Transparency Increases Trust in and Engagement with Government.” Manufacturing & Service Operations Management (forthcoming). https://www.hbs.edu/faculty/Publication%20Files/14-034_16ccb2b4-1a24-47b5-8bcf-dca2720bb96f.pdf
The Benefits of Cost Transparency
A substantial body of work in social psychology suggests that disclosure, especially of information typically kept private, is associated with heightened relationship quality. In this research paper, the researchers show that when a company discloses information that a customer perceives to be sensitive – specifically, the costs to produce a good or provide a service – this can increase purchase intention and feelings of trust.
The paper begins with a natural field experiment conducted with an online retailer selling wallets. The retailer decided to add an infographic that included the costs incurred to produce the wallet to their website. The infographic was implemented for only certain wallet colors and not others, a mistake that was overlooked for five weeks. This resulted in a natural experiment, since the difference in daily sales before versus after the infographic was introduced could be compared. Revealing cost information increased sales for the cost transparent wallets by 22.0%.
The researchers then ran a more controlled field experiment, in a university dining hall. Next to a station selling chicken noodle soup, customers saw one of two different signs. One sign listed the chicken noodle soup components, while the other sign revealed the costs of making the soup. The field experiment was run over five consecutive weeks at lunchtime, with the signs shown in alternating order. Cost transparency led to a 21.1% increase in the probability customer would buy a bowl of chicken noodle soup.
The subsequent four online studies in the paper shed light on when and why the beneficial effect of cost transparency emerges. The researchers show that cost transparency is perceived as a particularly sensitive form of disclosure. The positive effect of cost transparency is primarily driven by increased perceptions of firm trustworthiness, as opposed to merely feeling that prices are more fair. Moreover, cost transparency boosts purchase interest only when voluntarily revealed by the firm, as opposed to involuntarily (i.e. required by law).
So what’s the key takeaway? This research suggests that akin to when people reveal sensitive information, when firms voluntarily reveal their costs, they can engender trust and deepen relationships with their customers.
Mohan, Bhavya, Ryan W. Buell, and Leslie K. John. “Lifting the Veil: The Benefits of Cost Transparency.” Marketing Science (2020).
The Pitfalls of Transparency
In this paper, the research team examined the relationship transparency and bias. Prior research has documented the prevalence of bias in the sharing economy. To combat bias, a number of ride-sharing platforms have removed information about a rider’s gender and race from ride requests sent to drivers: drivers therefore can only see a profile after accepting a rider. This research paper examines whether this deliberate change in the timing of customer transparency removes bias from ride sharing platforms.
The researchers conducted a field experiment on a ridesharing platform in Washington, DC. While pickup and drop-off locations were held constant, they randomly manipulated three dimensions of rider characteristics: gender, ethnicity, and support for LGBTQ rights. The researchers used a computer program to generate rider profiles, randomly manipulate profile pictures, and request rides. At the moment the ride was requested, the driver could only observe the general location of the rider. Once the ride was confirmed, the rider’s name and profile picture were made visible to the driver. as well as the exact pickup location. The researchers waited three minutes to allow the driver to cancel, before canceling the ride themselves. They requested an equal number of rides during peak and non-peak hours, for a total of 3,200 ride requests.
The key measure of interest was the cancellation rate: were drivers more likely to cancel ride requests from certain types of riders, after seeing their profile picture and name? The researchers found no evidence of bias against female riders. However, underrepresented minorities were 2.6 times more likely to be canceled than Caucasians. Those who supported the LGBT community in their profile pictures were 1.6 times more likely to be canceled than those who did not. Thus, changing the timing of customer transparency did not remove bias from the ridesharing platform.
Why is this important? This research suggests that when changing policies related to transparency, an organization must carefully examine unintended behavioral consequences.
Mejia, Jorge, and Chris Parker. “When Transparency Fails: Bias and Financial Incentives in Ridesharing Platforms.” Management Science (2020).