Behavioral Insights: A New Tool for Performance Management

Performance management and innovation teams across the country, at all levels of government, are helping to deliver higher-quality services using fewer resources. As part of these efforts, these teams are beginning to incorporate behavioral insights and randomized controlled trials (RCTs) as essential components of their performance management frameworks.

Process improvement techniques, like Lean, have been used increasingly over the past several years to drive innovation in the public sector. Lean offers government workers structured ways of scoping and solving problems. It pushes them to ask questions like: What service to do I provide? Who is my customer? Where does waste exist in my process of delivering that service, and how can I work with my team to eliminate it?

At workshops such as Kaizen Events, civil servants learn frameworks to map out processes, quantify resources, and hone in on concrete ways to be more efficient or effective in delivering government services, while using fewer resources like time, money, or steps to get things done. They’re saving taxpayer dollars, spurring innovation, and bringing a renewed focus on the citizen as a valued customer.

Now, through What Works Cities, an initiative launched by Bloomberg Philanthropies to help mid-sized cities use data and evidence to improve decision-making and results, innovators in municipal governments are taking the next step in process improvement, integrating behavioral science into their toolkits and rigorously testing these insights to understand what works.

These efforts center on helping improve service delivery by taking the science of human decision-making into account. While we know intuitively that people are not always entirely rational actors, too often policy is designed as if we were.

Behavioral science teaches us that our environment and context influences the choices we make and sometimes causes us to act in ways that go against our best interests or intentions: a business owner may accrue a code violation because the rules are buried at the bottom of a difficult to access document; a homeowner might forgo a tax credit because the process of claiming it is poorly advertised or appears too complicated; a family has their water shut off, not because they can’t afford the bills, but because they lost track of the letter in a stack of mail.

Knowledge of the situational factors that may push people to make adverse choices has helped cities uncover new approaches to tackle longstanding challenges.  

For example, building on research that suggests making people feel unique can prompt action, the city of New Orleans, sent out behaviorally-informed SMS messages that increased the number of low-income individuals agreeing to schedule a doctor’s appointment by 40 percent.

In Denver, the city increased the rate of businesses filing taxes online—a major savings compared to filing by mail—by 67 percent, simply by reframing a letter to highlight a pervasive social norm: that the majority of their peers already have an online account.

Brian Elms, the director of the Denver Peak Academy, is helping institutionalize these practices within the City’s acclaimed performance management training program. “The Peak Academy teaches every Denver Black Belt about behavioral economics in our classes,” says Elms. “We believe choice architecture and process improvement complement each other incredibly well.”

Elsewhere, the city of South Bend, Indiana, is currently working on a variety of projects that employ behaviorally-informed messaging strategies. They’re making it easy for business owners to address fire code violations, low income homeowners to qualify for tax credits, and utility customers to pay their bills earlier so their water isn’t shut off.

“We tended to think that systems and processes can be changed but that human interaction was fixed,” says South Bend Chief Innovation Officer Santi Garces. “But with behavioral insights, we’ve seen that human interventions can be measured and form a lever that we can pull as well.”

Behavioral science teaches us that context can play an outsized role in determining how decisions are made. That’s why cities like New Orleans, Denver and South Bend have been so careful to test how these innovations work locally by using the gold standard of evidence-based policymaking: the randomized controlled trial (RCT).

Random assignment of individuals to either a treatment group, which receives the new intervention, or a control group, which receives services-as-usual, helps ensure that there are no systematic differences between the two groups with the exception of what is being tested. Cities can then be confident that any difference they observe in outcomes between the two groups is due to the intervention itself and not to other incidental factors.

With training provided by the Behavioral Insights Team through the WWC program, Elms has helped build out Denver’s capacity to run multiple RCTs. “These techniques,” he says, “push us to be on the cutting edge for government service delivery and innovation.”

Performance management in government is evolving fast. More and more, city managers know that serving the people means building municipal services upon a nuanced understanding of how people actually behave, rather than how we might think people should behave. And, more and more, city managers have data at their disposal to test what works.

“What service to do I provide? Who is my customer?” Civil servants have grown comfortable asking these questions. But now they’re also asking, “how does my customer perceive the service I provide?” and “how can I test my idea for improving it?”