Connecting state and local government leaders
The Behavioural Insights Team, a company once part of the British government, has its sights set on improving jurisdictional service delivery across the pond.
The city of San José reduced illegal dumping by 320 tons since July, earning the recognition of the Keep America Beautiful non-profit organization on Monday. A key part of the city's success has been a program partner, The Behavioural Insights Team. Created by the government of the United Kingdom, which retains majority ownership, the now-private government consultancy works with jurisdictions worldwide to build a more realistic model of human behavior into policies.
BIT’s application of behavioral science is highly empirical, tweaking government project designs at their outsets and running evaluations—generally randomized control trials—to see how effective those adjustments are.
“I’m always struck by how many more similarities there are between the ways different parts of government in different policy areas operate than there are differences,” Elspeth Kirkman, senior vice president of BIT’s recently established North America office, told Route Fifty in an interview.
In San José, BIT worked with the city to A/B test direct mail promoting the city’s curbside pickup of large items like furniture and appliances, which often are dumped illegally.
“Because of our strong internal and community partnerships, we’re seeing visible results that are making our community cleaner and greener and engaging our residents to strengthen our neighborhoods,” Kerrie Romanow, San José Environmental Services director, said in a statement released Monday about the city's award.
Now BIT wants to expand beyond What Works Cities to team with state and local government to develop innovative policies inspired by behavioral science that can be scaled without increasing taxpayer funding.
In Denver, BIT has helped the city boost the number of businesses filing taxes online by 67 percent.
Three trials were conducted around Denver’s revamped online filing portal, eBiz Tax Center, and the letters were sent to businesses encouraging them to use it. The standard “go green” appeal was the experimental control tested against modified messages: a “social norms” letter asserting, truthfully, that the majority of Denver businesses file taxes online and another “loss aversion” missive warning of the time wasted paper filing.
Both modified letters fared better than the original, but, if the social norms message had been sent to everyone in the experiment, more than 500 additional businesses would have registered online.
“Businesses do what we all do Tax Day,” Kirkman said. “They’re filing at the 11th hour and not in the frame of mind to try something completely different.”
Next, BIT and Denver want to get businesses filing online earlier. Working with the Denver Peak Academy, which trains city employees, BIT is working to train the trainers in behavioral science and evaluation.
Kirkman and the BIT team work with cities to develop a pipeline of experimental projects while building capacity simultaneously. At a high level they’re helping city halls confront the challenges that keep them up at night, she said, while on a granular level they’re examining how services are delivered.
One trend among U.S. cities is their desire to bring behavioral science to operations like code enforcement, Kirkman said. The job is high profile, often under-resourced but plays an important role in shaping the look and feel of cities through compliance with building standards.
In looking for low-cost ways to enforce city code as efficiently as possible, BIT advised the New Orleans city government to send a courtesy letter letting homeowners know an inspector was coming around. Voluntary compliance rose 16 percent.
In Chattanooga, Tennessee, the city intervened at a slightly different point in the process, sending letters to past code violators with tips of what not to do—reducing the likelihood of a second offense by 9 percent.
Much of amateur government behavioral science can be done in Microsoft Excel, Kirkman said, and city leaders don’t need to be econometricians to tell when a result is statistically significant.
“We make it so anybody with the enthusiasm, data and hours to spare can realistically run a randomized control trial without learning all the stats underneath," she said. “It’s about turning the mirror and looking inside government to improve the workforce and get people making better decisions that build trust in the community.”
Dave Nyczepir is a News Editor at Government Executive’s Route Fifty and is based in Washington D.C.
NEXT STORY Locked Up for the Holidays