Beyond carrots and sticks How state and local governments can use nudge thinking to improve outcomes

Article Sections

​Government agencies charged with handling small offenses and oversights can look to a largely untapped resource: the behavioral economics playbook. Learn how “nudges,” if executed effectively, can often yield better results using fewer resources than traditional approaches would.


Learn more

View the Behavioral Economics collection

The idea is simple: It is better for state and local governments to help citizens “do the right thing” in the first place than to go through the painful process of penalizing those who misbehave.

Government agencies seek to promote socially desirable citizen behaviors in many ways. Consider these examples:

  1. Children, brush your teeth. In some states, daycare centers are required by law to have children brush their teeth.1 With centers at risk of losing their licenses, heavy-handed enforcement can be costly and difficult.
  2. Recycling in Chicago? Only 9 percent of Chicagoans properly recycle. Recognizing that incentives would have to be substantial to change behavior and penalties are not feasible, the city government is considering the best methods to change behavior.2
  3. Hire the nanny on the books. In 1993, “Nannygate” derailed not one but two of President Clinton’s potential nominees for US attorney general, when it was found that both had used questionable and/or “under-the-table” methods in hiring domestic help.3

No doubt “carrots and sticks”—incentives and punishments used to change behavior—have their place. Economic incentives play a critical role in our market economy. In law enforcement, it makes sense to punish perpetrators of serious crimes—and to incur the cost of apprehending, adjudicating, and incarcerating those lawbreakers.

But for more minor offenses or oversights, trying to drive better social outcomes through punitive actions can be costly, and officials are rightly reluctant to unleash big-time punishments on small-time violators. As the Nannygate example shows, even individuals generally fit to serve as the nation’s top law enforcement officers can sometimes stumble on minor regulations. Beyond rule breaking, many government agencies also seek to improve suboptimal behaviors such as smoking and aggressive driving. Overall, policy makers who limit themselves to enforcement may be missing a huge opportunity to help improve social outcomes.

Enter the emerging field of behavioral science, which explains how “nudges”—carefully designed prompts and activities that encourage better outcomes by leveraging how people naturally think and feel—can be a more effective (and often, less expensive) alternative. Nudge thinking is a stark departure from the traditional “carrots and sticks” approach. It relies on the idea that small changes to the “choice environment” can encourage large changes in people’s actions. Part of the appeal of nudges—for both those seeking change and those who are being asked to change—is that instead of mandating behavior, nudges offer people the ability to make their own decisions.

There are many areas where nudge thinking may prove more effective—and potentially much more cost-effective—than the carrots-and-sticks approach. And yet most state and local governments are just beginning to scratch the surface of this approach; they have not fully explored or realized its potential benefits, both in terms of cost and outcomes. This article explores three broad areas in which state governments and municipalities could apply nudge thinking to improve people’s habits and behavior:

Following the rules. Government agencies often work hard to encourage citizens to follow public benefit programs’ rules and regulations. Here, the goal is compliance. From tax laws to labor laws, from programs such as Supplemental Nutrition Assistance Program (SNAP) to unemployment insurance, the rules surrounding these programs can be complex, difficult to follow, and easily misunderstood. Given the complexity, participants can also be tempted to cut corners. In lieu of ineffective enforcement efforts, these agencies can use nudges to boost compliance at little cost.

Getting with the program. For those who rely on social services, particularly public health and return-to-work programs, there is a mutual interest in encouraging behaviors that promote better choices by participants. Using nudges to improve program adherence could promote better outcomes for citizens while limiting program costs.

Being good citizens. Nudges could be used to encourage participation in a broad array of positive and helpful behaviors, from environmental stewardship (such as water conservation, recycling, or carpooling and other forms of ridesharing), to proactive health measures (such as getting vaccinations), to good neighbor behaviors (such as courteous driving, voting, or organ donation).

To illustrate the viability of nudges in these areas, we’ve outlined a number of successful campaigns that incorporated insights from behavioral science to improve outcomes.

Following the rules: Using nudges to increase program rule compliance

Motivating compliant behavior is very different from punishing those who commit major crimes.4 When it comes to taxes and benefit programs, the opportunity for noncompliance, and the difficulty in detecting these corner-cuttings, can far outstrip the resources of government. The fact is, government relies heavily on voluntary compliance. Just as an ounce of prevention is worth a pound of cure, a small investment to boost voluntary compliance can be far more effective than trying to police every instance of noncompliance. (This can also be true for internal initiatives. See the sidebar, “Nudging public employees” to learn more.) The good news is that well-designed nudges can strengthen voluntary compliance, influencing everyday citizens to follow the rules set by government programs.

Nudging public employees

“How do we adopt new technology? Painfully.”

—Massachusetts CIO Bill Oats5

Whether it’s adopting a new technology or establishing a new process, any organization can find it difficult to change its culture. Evidence shows that 70 percent of all large-scale change initiatives fall short of their long-term objectives.6 As the quote above shows, tech adoption in the public sector can be particularly challenging. Low adoption rates can result in large IT investments failing to deliver hoped-for benefits.

Public-sector leaders looking to drive change likely face even more hurdles than their private-sector counterparts for a few reasons.7 First, few government leaders have experience driving large change efforts, particularly in organizations of the scale of government. Secondly, public-sector change leaders are often constrained by red tape and restrictions around procurement, personnel, and budgeting. Union contracts typically limit their ability to use the financial incentives leveraged by corporations.

With these constraints in mind, it is particularly important for agency leaders to work with the stakeholders they have the most leeway in influencing—the employees charged with carrying out change initiatives. How can public-sector managers overcome the status quo and instill new attitudes and behaviors?

One way is to employ behavioral insights as part of any change management effort. Change management research suggests that these initiatives are more likely to succeed if they incorporate more intrinsic motivators. These include adopting practices that would provide employees with a greater degree of autonomy, opportunities for growth, and a sense of meaning.8 Other nudge strategies might include sharing goals for adoption, lowering barriers to adoption, or providing information about the adoption by peer groups.

To learn more about how behavioral science can improve change management outcomes, refer to the Deloitte Review article, “Humanizing change: Developing more effective change management strategies.”

Ontario employer tax collection: Make it easier to follow the rules with implementation intentions

In the Canadian province of Ontario, every year, employers are required to file and pay an Employer Health Tax (EHT), a payroll tax on health services provided to both current and former employees. In 2014, the government of Ontario was struggling with a number of businesses filing their EHTs delinquently;9 more than 7,000 employers filed their returns late that year. Recognizing that many small businesses were struggling to comply with the mandate, the government of Ontario decided to turn to the behavioral sciences for help.

To assist employers who were running late on filing, Ontario tested new messaging that focused on implementation intentions (see sidebar, “Nudge toolbox: Implementation intention).

Nudge toolbox: Implementation intention

Committing to a plan can help individuals accomplish a specific goal by spelling out “how, when, where” they intend to carry out an action, which is known as an implementation intention. In a number of studies concerning fruit and vegetable consumption, people who expressed their dietary goals in an “if/then” format—for example, “If I am at home and I want to have dessert after dinner, then I will make myself a fruit salad.”)—consumed significantly more healthy foods.10

Similarly, the province of Ontario applied an implementation intention approach to tackle the delinquent EHT filings. In 2015, a subset of employers tweaked their collection letters to focus on where participants could file a return, directing them to websites and the mailing addresses to service centers. Participants also received detailed instructions on how to go about filing the EHT return and a deadline for when the filing was needed.11

A month later, the results were analyzed. Compared to the control group that received the standard delinquent message, the employers using the implementation intention approach increased their filings by 13 percent (53 percent vs. 46.9 percent). Through a single nudge, Ontario was able to significantly increase the compliance of its businesses.

Unemployment insurance in New Mexico: Nudging compliance with priming

Let’s start with the bad news: Nearly one dollar out of eight distributed under the unemployment insurance programs in the United States went to someone who was ineligible, totaling more than $4 billion in 2014 alone.12 Unemployment benefits are 100 percent financed by employer taxes, which means that every incorrectly paid dollar comes out of a company’s bottom line.

These overpayments may be due partly to a poor understanding of the rules, or may be due to “small rule breakers,” in which some claimants under-report money they earn while looking for work and collecting unemployment. Recovering these overpayments through detection and recovery is difficult. In New Mexico, the Department of Workforce Solutions (DWS) is only able to investigate 25 percent of the potential 45,000 cases of unemployment fraud.13 If possible, it is far preferable to encourage accurate answers in the first place.

To reduce overpayments, in 2014 New Mexico’s DWS applied the principles of behavioral insights. They introduced some simple, low-cost nudges to their claims process with promising results.

Behavioral nudges were especially impactful during the weekly certification process, where an unemployment claimant documents any hours worked during the week. And it starts with a question: “Did you work this week?” This is a key question because by the rules of the program, earnings are usually deducted from the weekly unemployment benefit amount.

As one could imagine, there may be temptation to underreport earnings to avoid reductions in their unemployment insurance check. So to address these natural temptations, the DWS applied the behavioral concept priming to the claims process to induce higher rates of honesty (see sidebar, “Nudge toolbox: Priming”).

Nudge toolbox: Priming

When people are exposed to a stimuli that reminds them of the value of honesty, they are more likely to behave honestly. For instance, in one study, asking students to recite the Ten Commandments before a test resulted in fewer cheating incidents than occurred when students were asked to name as many Shakespeare sonnets as possible.14

Here, a group of weekly claimants were primed with a personal attestation asking them to certify that they would fill out their weekly earnings accurately and honestly before completing the form. This was a departure from the more standard practice of attesting to honest behavior after entering information. By using this technique, along with a number of other forms of priming, the DWS doubled the number of claimants who reported earnings vs. those in the control group.15

New Mexico also used pop-up messages that nudged greater disclosure, experimenting with various messages and finding that some messages performed better than others (see figure 1). According to the Deloitte Review article, “Nudging New Mexico”: “Overall, claimants who see pop-up messages reported earnings 25 percent more often that the control group. They reported the same amount of earnings on average, but they reported more frequently.”16

Percentage of claimants who reported earnings in response to different messages

Overall, as a result of introducing behavioral nudges and other changes, New Mexico saw dramatic drops in its overpayment rates and fraud rates.

Whether through application of implementation intentions or priming more honest behavior, a number of state government programs may benefit from similar nudges.

Getting with the program: Nudging greater engagement in work and health programs

Public programs are often aimed at providing citizens with services intended to improve their quality of life. Too often, however, the people enrolled in these programs fail to fully follow through on their responsibilities, and thus fail to realize the full benefits of these services. Here are a few examples:

  • An expectant mother on Medicaid fails to take advantage of free prenatal visits, leading to health complications and higher costs.
  • A laid-off worker fails to take advantage of job placement and resume-writing classes, dimming his prospects of finding a job.
  • An injured worker fails to attend scheduled re-training classes, then struggles to find alternative work in a timely fashion.
  • A diabetic Medicaid patient fails to regularly monitor his insulin level and disregards dietary guidelines, resulting in poorer health—and higher costs for the state.

In all of these cases, individuals are failing to behave in their own self-interest—that is, incentives are simply not driving desired behaviors. In some of these cases, behavioral nudges may facilitate better outcomes and rein in costs.

New South Wales helps injured workers return to work faster: Redesigning the choice architecture to improve outcomes

New South Wales wanted to improve the Return-To-Work (RTW) results for injured workers. In partnership with Australia’s Behavioural Insights Team (BIT), New South Wales redesigned its RTW program to make the process of finding employment easier.17 At the heart of the redesign was changing the choice architecture.

In short, the old process was time-consuming and confusing. During the first 21 days of entering into the RTW program after an injury, participants were expected to read 20 pieces of paperwork. These documents were filled with legalese that failed to engage participants; most never made it past the third letter.18

To address these issues, New South Wales streamlined the process and simplified the language. Similar to implementation intentions, they also provided injured workers with commitment devices (see sidebar, “Nudge toolbox: Commitment devices”).

Nudge toolbox: Commitment devices

Psychology tells us that when a person formally commits to do something, she is more likely to carry through with her plan—even when there are no consequences for noncompliance. Commitment devices offer a vehicle to address implementation intentions. For example, during the 2012 election, President Obama’s campaign provided would-be voters with a “commitment card.” People would sign the cards, which included a smiling picture of President Obama, stating that they would go to the polls on Election Day. In addition, these voters were asked to make a plan regarding how and when they would get to the polls.19

New South Wales included a commitment device that asked injured workers to make a plan that included a return-to-work goal. Rather than telling people when they should be ready to return to work, New South Wales asked participants when they thought they would be ready to return. They also asked them to make a plan regarding how they would return. Following this change to the choice architecture, these workers returned to work 27 percent faster than the control group did.20

Other governments have emulated these successes in improving return-to-work outcomes. Norway’s social security administration redesigned its process to help those who were out of work on medical leave for more than six months.21 Specifically, it found that holding an inexpensive meeting with the employee, employer, and treating physician led to better practices to help the worker gradually return to his job. For every meeting these groups held, employees returned to part-time work an average of 20 days faster and full-time work 10 days faster.

Better health outcomes at lower costs through salient messaging and prospect theory

Public health programs spend a disproportionate amount of time and money treating preventable diseases; 75 percent of all health care costs are associated with the treatment of these diseases.22 For this reason, public health services are continually looking for new ways to preemptively serve people before they need to be rushed to the emergency room. (Medical providers can be nudged as well—see the sidebar, “Increasing generic drug prescriptions.”)

Increasing generic drug prescriptions

Patients are just one of the players in the health care ecosystem for whom nudges can lead to more efficient treatment. Doctors can also be encouraged to make better choices through changes to the choice architecture. For example, in some cases, doctors may prescribe more expensive brand-name prescriptions rather than generic drugs. This is often out of habit rather than for medical reasons, or because it is often easier to quickly recall the brand name vs. its generic equivalent.

Nudges have been shown to influence this decision, resulting in more appropriate utilization of lower-cost generic prescriptions. How? In one study conducted by Penn Medicine, the simple but powerful technique of changing the default significantly altered behavior by the prescribing physician.23 For 10 common medical conditions, the electronic medical system doctors used to enter prescriptions automatically defaulted to the equivalent generic drug—even when a brand name was originally typed. To override this default and keep the brand name, which is medically appropriate for certain patients, the prescribing doctor merely had to check a box labeled “dispense as written.”

The results were compelling. Before establishing the new default, brand-name drugs were prescribed for 1 in 4 patients. With the new defaults built into the system, brand names were prescribed for just 1 in 50 patients—achieving significant cost savings.

Not surprisingly, medication adherence—taking medications as prescribed, in terms of dosage, timing, and frequency—is one of the best defenses against future medical problems. Backed by a number of studies, research suggests that the additional amount spent on medications due to stricter adherence is easily offset by the reductions in emergency room visits and hospitalizations.24

For a number of reasons, though, people still frequently skip their medications. This can be due to lack of motivation, perceived ineffectiveness, or simple forgetfulness.25 The behavioral economics field, however, offers some promising methods to help people adhere to their medication schedules.

Consider one promising study involving statins, a highly effective type of medication that helps reduce heart-disease-related outcomes such as strokes.26 Statins are only effective if used regularly. This is particularly challenging since many people who are on statins, though at high-risk for heart disease, may not have experienced any symptoms; they feel the same if they take their medications or not. To motivate better statin adherence, the Oklahoma Health Care Authority’s Medicaid program tested several behavioral interventions, which proved highly successful: The program increased its statin prescription adherence by 78 percent over the control group. It accomplished this by employing a combination of small financial incentives for scheduling cholesterol appointments ($5 gift cards) along with salient messaging (Post-it note reminders) describing the consequences of not taking the medication.

In another experiment, patients taking Warfarin to treat blood clots had their pill dispensers connected to a lottery system.27 Essentially, the foundational behavioral economics concept of prospect theory informed this intervention (see sidebar, “Nudge toolbox: Prospect theory-based lotteries”).

Nudge toolbox: Prospect theory-based lotteries

Prospect theory suggests that even a small chance of winning a big prize is enough of a motivator to invest in a course of action that may not financially make the most sense.28 This is why so many people will purchase a lottery ticket even when the investment doesn’t make rational (financial) sense. However, prospect theory also suggests that people are risk-averse—or, in this case, regret-averse: We often make decisions not because they yield the greatest economic utility but instead, we try to minimize feelings of regret.

For the patients taking Warfarin, their pill bottles included a sensor that indicated whether they had taken their medication. Every day, patients had a significant chance of winning $10 or a small chance of winning $100. (On average, a prize would be awarded every fifth day.) There was a catch, however: If a patient failed to take his medication, he was disqualified from the following day’s lottery. To bump up the regret factor, disqualified patients were still notified if they would have won had they taken their medication. The cost of running the lottery was only about $3 per day per patient, but this small incentive radically altered outcomes. Incorrect dosages per day dropped from 22 percent to 3 percent.29

In all of these cases, program designers were able to nudge greater compliance simply by embedding principles from the behavioral sciences.

Good citizenship: Nudging efficient energy usage and water consumption

State and local governments often seek to encourage better behaviors that aren’t necessarily associated with a specific rule or program. These “good citizen” efforts might include basic civility (such as courteous driving, or giving up your seat on the bus for a person with a disability), environmental awareness (encouraging energy conservation or promoting recycling), or other civic concerns (such as voting and anti-smoking campaigns). In most of these cases, society would like to see more of these good behaviors, but may want to avoid the heavy enforcement hand to enforce them. Indeed, there exists a whole category of laws that are a half-step away from suggestions passed by legislatures—the hardly-ever-enforced laws that routinely get ignored by citizens and police alike. Think about it: Do you know anyone who has ever actually been cited for loitering, or jaywalking, or violating a watering ban, or fishing without a license, or . . . ? You get the idea.

In these cases, getting people to “do the right thing” isn’t always easy. How do you encourage people not to litter when fines are low and catching someone is probably not worth the time of writing the ticket? When California is stuck in a drought, how do you persuade people to consume less water? Once again, nudges may help.

Increasing conservation with social proof and real-time feedback

Seeing how our peers perform can profoundly influence our own behavior. Though you may have thought peer pressure stopped influencing you after ninth grade, you may be surprised to see how persuasive it still can be. If you’ve ever felt self-conscious about being last to arrive to a meeting and made sure it never happened again, then you understand this.

In the behavioral sciences, nudging with peer pressure is referred to as social proof (see sidebar, “Nudge toolbox: Social proof”).

Nudge toolbox: Social proof

Social proof messages help people benchmark their own behavior against others to make changes for the better. Simply making people aware of how they are doing against a norm can prompt them to make positive changes. In the New Mexico case discussed earlier, social proof was used to encourage compliance with messaging that stated (truthfully), “Nine out of ten people in <your county> accurately report earnings each week.”

In one example, the World Bank’s Governance Global Practice and the OECD’s Latin America Caribbean Unit teamed up with Ideas42, a nonprofit specializing in behavioral tools, to reduce water consumption in Belén, Costa Rica.30 To curb household water usage, a number of peer comparison messages were mailed out to inform people how much water they were consuming compared to their neighbors. Accompanying their water bills was a brightly colored sticker: If their water consumption was below the neighborhood average, the sticker had a smiley face and a note of congratulations. Those who consumed more than their average neighbor received a frowny face sticker, along with a special message informing them that they were using more water than most.

This little nudge created disproportionately positive results. Compared to the standard letters, consumption fell between 3.5 percent and 5.6 percent for those receiving a social-proof-inspired sticker.

Results such as these have the potential to be carried even further through the integration of smart devices. One study out of the United Kingdom found that equipping houses with smart meters that provided real-time feedback on energy consumption, like the earlier example involving pill bottles, could powerfully influence behavior.31 By showing citizens not only how they compared to other households but also informing them with real-time alerts, the British Office of Gas and Electricity Markets was able to significantly reduce energy usage.

Whether it be energy consumption or engendering better citizenship, nudges can speak to our intrinsic desire to “do good.”

Taking these insights back to your government program

State and local governments work very hard to alter the behaviors of their citizens. In many cases, “behavior modification” is synonymous with heavy regulation, fines, and sometimes, incarceration. Nudge thinking, meanwhile, offers an effective, low-cost alternative that addresses many of the nuance issues facing government agencies.

There is a legitimate public policy debate about how much influencing we want our government to be engaged in, and governments need to consider the ethics of nudging, including transparency around both goals and means. But it is hard to get too upset about a government program where the worst consequence is getting a frowny face sticker.

These nuance issues often take shape in one of three forms: getting people to comply with rules and regulations, especially regarding benefit programs; getting people to more fully “buy-in” to the services that government programs offer, such as preventative health care measures; or simply inspiring better citizenship. In each of these cases, nudges have a demonstrated track record of making government interaction easier (see table 1 for a review of the concepts covered).

Review of nudges and future opportunities

For program designers looking to embed these insights into their own initiatives, consider these three steps:

  1. Identify your objective. Does your initiative need to mitigate small-time rule breakers, increase program buy-in, or encourage better citizenship? If not, more traditional approaches may be appropriate. If so, your program may benefit from carefully designed nudges.
  2. Choose your nudges. Start with considering your choice architecture. Is it intuitive or filled with technical and legal jargon? Make “simple” the mantra.
  3. Test and learn. Not all nudges work and sometimes they need to be tweaked. With any new initiative, lean on randomized control trials to validate the efficacy of your program. It is hard to predict how people will actually respond beforehand, so measuring impact is essential.

Make no mistake—nudges are no panacea. But we believe that there many opportunities for state and local governments to improve outcomes with minimal expense and without the heavy hand of punishment or the heavy cost of incentives. Indeed, these efforts can often pay for themselves by reducing expenses in other areas.

We believe that states have only begun to scratch the surface on employing nudge thinking. Particularly as data analytics increasingly allows us to target the most critical subset of a population, understanding how to influence behavior becomes especially important. From reducing fraud in benefit programs to improving the efficacy of health services, from protecting the environment to encouraging kids to stay in school, governments have a huge opportunity in applying the insights of behavioral economics.

The beauty of nudging lies in its simplicity. Starting small can set a government agency on a journey to uncovering big results. So what are you waiting for? After all, everyone else is doing it. (Nudge, nudge!)