danger
COVID-19 Risk Level for Multnomah County: Extreme Risk

Behavioral Insights and Testing Outcomes

Through What Works Cities, the City of Portland has partnered with the Behavioral Insights Team to apply behavioral science research findings to improve city communications and programs, and test what works using randomized controlled trials.

On this page

What Are Behavioral Insights?

Behavioral insights (i.e. behavioral economics, or nudging) draws on knowledge from the behavioral sciences (psychology, neuroscience) about how people think, behave and make decisions under different conditions. Governments around the world have shown that when we design for people’s busy lives and use proven methods of encouraging behavior – beyond financial incentives and penalties – we can increase participation, effectiveness and efficiency of government services.

Testing what works using randomized controlled trials – the gold standard of evaluation – allows City staff to pilot multiple approaches at once and find out which works best and for whom. Rigorous, real-time testing provides staff with data and evidence to make informed decisions about how to deliver the most effective programs and communications.

Our Work So Far

Between 2016 and 2018, Portland completed over 10 projects with support from the Behavioral Insights Team. These projects were done in partnership with 6 different bureaus and ranged from encouraging bike share use and emergency preparedness to increasing code compliance reducing 911 call-taker burnout.

Portland staff have since designed their own projects, tested with randomized controlled trials, including efforts to improve on-time tax payments, increase transit ridership and biking, and reduce the frequency of false alarms in home and business security systems.

In 2020, Portland is partnering with the Behavioral Insights Team to improve COVID-19 messaging and to increase the wellbeing of city staff who have shifted to remote working.

Featured Project: COVID-19 Grocery Store Poster

Poster showing grocery staff and customer 6' apart, as well as person wearing facemask, washing hands, and an X over a person touching their face.

In April 2020, the City of Portland partnered with the Behavioral Insights Team to create a poster for grocery stores aimed at encouraging customers to keep a 6-feet distance from staff and each other.

The goal was to protect staff’s health and reduce stress by encouraging customers to keep the recommended safe distance to reduce spread of COVID-19.

We designed multiple variations of the poster and used rapid online testing to assess which text and visuals were most memorable and convincing.

The final poster – which used clear visuals of staff and customers staying 6-feet apart and simple text emphasizing customers’ duty to protect staff and one another – was left unbranded, made available for download, and promoted widely at the local, state and national level.

Thank you to the partners who made this project possible: Laura Hall at the Portland Bureau of Emergency Management, a great team of staff from the Behavioral Insights Team, and What Works Cities, who supported and funded this work.

Project details

The need: Change people’s behavior in grocery stores

In April 2020, as the COVID-19 health crises and accompanying health and safety best practices were rapidly evolving, City of Portland emergency managers received numerous reports about grocery store customers not physically distancing or wearing face coverings.

At the time, social distancing – the recommendation to maintain at least 6-feet of distance between people to prevent the spread of the novel coronavirus – was a new behavior for most people. Customers were also stressed about grocery shopping in a time when there were some product shortages. Grocery store signage varied widely between stores and staff reported feeling stressed about their safety. 

In addition to our goal of improving grocery store staff safety, we saw grocery stores as a good place to reach people with other key safety messages, like hand washing and wearing face coverings, since grocery stores were one of the few places people still visited on a regular basis during the Stay at Home order.

The project: Design a poster to encourage social distancing from staff

The Portland Bureau of Emergency Management partnered with the Behavioral Insights Team to design a poster to display at grocery stores to encourage customers to stay 6-feet or more away from staff. We also included additional key health messages: wash your hands when you get home, avoid touching your face, and wear a face covering (the CDC recommendation for face coverings was announced mid-design, so we quickly incorporated it).

Our goal was to design a poster that was:

  • Easy to quickly understand: We knew grocery shoppers’ already limited time and attention would be exacerbated by stress and anxiety about the health crisis, shortages of some food and cleaning products, and economic uncertainty.
  • Accessible: We wanted the message to be understood by everyone, including individuals with low literacy or low English proficiency. Disasters have a disproportionate impact on vulnerable populations, and it is vital that communications be as accessible as possible. To achieve this, we aimed to show the desired behaviors through pictures and reduce text to a minimum.
  • Based in research: The Behavioral Insights Team had just completed testing of 7 different posters promoting hand washing, and found that bright graphics and minimal text were most effective at helping people understand and remember how to wash their hands properly. They had also found that in efforts to encourage people to stay at home, simple, concise messages (“Stay at home”) and a “duty” message (“It’s your duty to stay at home”) appeared to be more impactful than other messages.

Testing: Rapid feedback through online surveys

The Behavioral Insights Team helped us design 8 variations of the poster, and then designed a randomized controlled trial to test viewer response, using their online platform, Predictiv.

We wanted to know:

  • After viewing the poster, how well did people remember key behaviors (stay 6’ apart, wear face covering, etc.)?
  • How convincing were the posters? That is, how likely were people to say they intended to do the behaviors.
  • If people only saw the images – and no text – would they still understand the key messages?
  • Do people understand that a poster showing 2 shoppers staying 6’ apart is also telling them to stay 6’ from store staff? Or does specifically showing a staff person 6’ apart from a shopper make that message more understandable or memorable?
  • What message framing is most effective: simple, duty, or save lives?

Through the Predictiv online testing platform, over 4,000 US adults viewed one version of the 8 posters and responded to a set of follow-up questions. Tests occurred between April 14 – 19, 2020. 

Results: What worked best

All posters performed well, but the most memorable and convincing poster was the one showing a grocery staff member and a shopper, with the message “It’s your duty to protect our staff and other shoppers.”

8 variations of the poster with different text and images, with most effective version circled.

More than 90% of participants rated all posters as easy to understand and credible. Most people (over 80%) who saw any of the posters variations understood and remembered the key messages and said they intended to follow the recommended behaviors (stay 6’ apart, wear face coverings, etc.).

There were variations in how well each poster performed, and the results showed:

  • A message of duty increased the number of people who recalled they should keep their distance to protect other people in the store from coronavirus. It was also most likely to make participants want to protect store staff. 
  • Showing grocery store staff made people think more specifically about staff. Compared to an image with two shoppers, a staff + shopper image was more likely to make people want to protect store staff. This is a good reminder to be cautious in assuming general images or messages will be as effective as specific ones.
  • Descriptive images can be as effective as text: The image-only versions of these posters performed similarly to image + text posters on comprehension and behavioral intent measures, meaning viewers may understand the main messages from images alone. However, all survey participants were English readers, so we cannot say with certainty that the same impact would be seen in the general population.

Poster distribution

Laura Hall, Communications Program Coordinator at Portland’s Bureau of Emergency Management, reached out to city, state, and national contacts to encourage grocery stores to use the poster. In addition to delivering printed copies to local culturally specific grocery stores, she shared the poster with the Oregon Department of Agriculture (who regulates grocery stores in Oregon), FEMA Region X (Oregon, Washington, Idaho, Alaska), and the EPA.

Volunteer delivering posters to Portland store owner.

Find out more

For more information about the City of Portland’s work in behavioral insights and testing outcomes using randomized controlled trials, contact lindsey.maser@portlandoregon.gov.

News and notices