Let’s first explore the relationship between service delivery and trust—along with how behavioral science can provide us with a powerful lens for interpreting their interplay.

As someone who has led government transformation initiatives, I have witnessed plenty of qualitative evidence pointing to a deep erosion of trust with even seemingly minor issues in service delivery. To support these observations with hard evidence, I have led several studies that delve into the data to better understand the drivers of public confidence in government services. 

For example, research I led at Qualtrics examined satisfaction levels across a set of highly visible state services, such as DMV, unemployment, Medicaid, and taxes. When respondents reported their experience improved over the course of a year, their trust in the government also grew, increasing 4.5 points for every 1-point bump in satisfaction. However, when respondents reported their experiences declined over a year,  each point decrease in satisfaction corresponded with an 11-point drop in trust—a dip over two times larger than the equivalent rise.³

Simply put: while improved performance enhances trust, declined performance erodes it at far greater levels.

Behavioral scientists were among the first to predict this type of outcome.

The fathers of prospect theory—Daniel Kahneman and Amos Tversky—studied how people make decisions under conditions of risk and uncertainty. Perhaps the most durable principle within this school of thought is loss aversion, which asserts that people tend to fear losses more than they value equivalent gains. Bottom line, losing $100 feels a lot worse than gaining $100.⁴

Applied to government services, loss aversion means that constituents will only mildly offer their trust when provided good service, but will quickly (and even more firmly) take it back when that service is compromised. Re-enter the FAFSA, a historically reliable service that incrementally built trust among its customers and the broader public. However, when that same satisfaction is compromised due to, let’s say, a failed modernization, trust drops precipitously—deeper, faster, and longer.

Three additional principles stemming from prospect theory can help us further explain why trust is so fragile in the public sector:

1. Reference Dependence

We tend to evaluate outcomes relative to a reference point rather than based on absolute outcomes. Whether something is seen as a gain or a loss depends on this reference point.⁵

For example, if the DMV successfully reduces wait times from 60 minutes to 20 minutes, but somehow jumps back to 60 due to a system outage or staffing shortage, people will perceive a frustrating “decline” in service quality. What was once “normal” is now a “loss” relative to customers’ new reference point.

2. Diminishing Sensitivity

The value we assign to gains and losses diminishes as the total amount increases. For example, the difference between gaining $100 and $200 feels larger than the difference between $1,100 and $1,200—even though both are $100 differences.⁶

Let’s think about this in terms of passport processing. If a government shortens processing from two months to one month, this feels like a huge improvement compared to reducing the process from eight months to seven months. The latter doesn’t generate the same public goodwill, even though it’s still a month faster.

3. Probability Weighting

We often overweight small probabilities and underweight large probabilities—meaning we assign too much importance to unlikely events and not enough to very likely events.⁷

Data breaches, for instance, might be one-off and relatively infrequent. They are, however, still highly publicized, creating an outsized public fear and mistrust in the government's ability to handle data.

Thanks to prospect theory, these principles offer robust explanatory power to help us understand why trust, even when widespread, is so fragile in the public sector.

Why does this all matter? When an agency loses trust, it can sap morale among the mission-driven workforce, stifling creativity and productivity. Service interruptions and incidents typically yield oversight from legislatures and boards. While well-intentioned, this actually compounds the issue by drawing resources away from fixing the problem and instead toward reporting and compliance. Critically, a deficit in public trust makes it difficult to try new things out of fear of further diminishing what little trust remains.

And trying new things? That’s the definition of innovation.

In the commercial sector, successful firms face what Harvard Business School Professor Clayton M. Christensen has called “the innovator’s dilemma,” referring to the disincentive to innovate due to prior success. Take Google, the absolute powerhouse of search engines. Search has been Google’s “cash cow” for years. Developing new products that will likely cannibalize existing revenue streams that are well-established makes innovation calculus especially difficult to solve. This perceived tradeoff sparks incrementalism and a preference for the status quo, ceding research and development to other firms with less at stake.

In this context, it’s no wonder that Google is considered late to the AI game, which could threaten its current $175 billion search advertisement business. But as Google is learning, ignoring this breakthrough technology puts future earnings at risk, as its dominance in search shows signs of waning.

For the government, the stakes are much higher. Whereas private sector organizations optimize for profits, governments optimize for a range of outcomes—many of which are tied to our physical and economic livelihood. The complexity and breadth of government programs heighten the tension between moving incrementally and moving transformatively, begging the question: how do we resolve this tension?

Imagine a typical government agency providing quality and stable services to customers with two “extreme” strategies to choose from. On one end, it can choose to remove and replace everything, even processes that are working well. Let’s call this the  Innovate Everything Everywhere All At Once strategy. This would be an extremely risky proposition (and one that it appears the current US administration is taking). 

Now, imagine that the same agency employs the opposite strategy. Let’s call it the No Change, No How agenda. Such an agency may make improvements at the margin, but is generally averse to any major policy or structural shifts from business as usual. This is a more inherent, “default” approach in government. Why change a thing when the trains are on the tracks, arriving on time, and customers are generally happy? But just like the “innovate everywhere” approach, a status quo approach can be just as risky in the long run.^8,9 How?

• Rising customer expectations: As technology and commerce evolve, so too do customer expectations. Seamless experiences, delightful digital interactions, and frictionless service have become the new norm due to technological advances in automation—especially in AI. Governments that fail to keep up will also increasingly fail to deliver on their mission in the eyes of the public, even under otherwise static conditions.
• The legal and political reality: Few agencies have the luxury of sitting back and resting on their laurels, given the need for elected officials and policymakers to show responsiveness to the electorate. Still, a status quo agenda, by definition, slow rolls change through the many tools administrators have—budget and bureaucracy being the two largest cudgels.
• Technological deprecation and security: Sure, the train may be running on its tracks, but what if those tracks are getting rusty? And what if train robberies are increasingly common? Countless agencies have been caught flat-footed as cybercriminals adopt more sophisticated methods to access government and taxpayer data. Many government systems are built on legacy platforms with processes that only made sense in a highly analog, paper-based world. The status quo means that every day, there is an increased risk of system failure. This is not only a costly proposition. It also cripples agencies’ ability to maintain, let alone improve, service levels their customers have come to expect.

Below, we offer three broad recommendations using the metaphor of a surgeon (public sector innovators) operating on a patient (government agencies) to illustrate when, where, and how to “operate” and innovate.

1. Before operating on a patient, diagnose the problem and apply the appropriate remedy.

No surgeon would put someone under the knife without a deep understanding of their medical history and underlying diagnosis. Just as a doctor would look at multiple levels of a patient’s health from the molecular to the systemic, public sector innovators should consider their agency’s diagnoses at multiple levels:

• Know thy customer: At the most basic level, practitioners should understand what drives customer behavior, what is most important to them, what moments matter most, and what drivers turn the needle on trust. This may seem obvious, but customer experience diagnostics is anecdotal at best. Government innovators can apply multi-modal methods (like surveys, focus groups, intercept interviews, digital diaries, and unstructured data analysis) to get to the heart of their customers' wants and needs.
• Know thy employee: According to the COMB-B Model, behavior is the result of capabilities, motivations, and opportunities. For the government professional, this might translate to skills, incentives, and responsibilities. In diagnosing the agency’s appetite for change, innovators should consider how these motivators are managed. For example, several thought leaders and scholars have argued that the federal bureaucracy is too focused on a “rules-based” order. This incentivizes individuals not toward creativity and innovation, but instead to focus on compliance and process at the risk of violating one of several policies, procedures, laws, rules, and regulations that govern their work.¹⁰
• Know thy organization: Finally, look at the system as a whole. As the Swiss cheese model tells us, risks become more likely when there are many “holes” (or vulnerabilities) that are lined up among different “slices” (or defensive layers). Thus, one might ask regarding their government agency: where are the underlying risks and vulnerabilities, and at what layers? Knowing that potential system risks occur from a combination of factors rather than a single root cause will help innovators avoid the biggest “complications” when operating on their patient.

2. Operate iteratively over time so that if mistakes are made, they are not fatal.

My sister-in-law has recently experienced significant issues due to a Chiari malformation, which is essentially when the skull places pressure on the brain, causing migraines, dizziness, nausea, and other more serious issues.¹¹ The doctors neither said “sorry we can’t do anything” (a “No Change, No How” approach) nor did they say “great, let’s open your skull up and try every procedure possible” (an “Innovate Everything Everywhere All At Once” approach). Instead, they offered a series of interventions, beginning with the least invasive and potentially most effective treatments, and ending with the most invasive and potentially risky procedures. 

The same approach can be applied to government interventions and innovation. More often than not, a more agile, progressive approach that involves feedback and continuous improvement both lowers the risk of potential system failure while also not shying away from doing hard things. That is, it gets you out of the “status quo” bias.

An example: in my own work at the Massachusetts Department of Revenue (DOR), taxpayers were flooding the call centers asking about their refund. While we already had a “Check My Refund” site, it was underutilized, leading to a deluge of calls. To fix the problem,  several argued for a complete product overhaul, a communication protocol to proactively text taxpayers, or a costly marketing campaign. Instead, we started with the most elegant and least involved solution before pursuing those more drastic interventions: we created a bright red “check my refund button” and moved it to the DOR homepage. This reduced calls by 75%.

3. Focus on patient outcomes, not process.

Finally, any good surgeon would not call it quits after the operation. There would be follow-ups, typically focused on patient outcomes, health, and quality of life. As government leaders, we can easily get distracted by process, by compliance, and by oversight. Returning to our original question, the FAFSA would have been much better served had it recognized its launch would not improve the outcomes baked into its mission—that is, helping more people attend college with financial aid. It reflects a government-serving-government mentality (e.g., to meet Congressional deadlines). 

Above all else, innovators should never forget that the very essence of public service is, of course, serving the public. 

In a world where public trust is fragile and citizen expectations are rising, innovation in government is not a luxury—it’s a necessity. But unlike the private sector, where disruption is often rewarded, the public sector must operate with prudence, accountability, and empathy.

The challenge, then, is not whether to innovate, but how to do so responsibly—by diagnosing before acting, iterating with intention, and staying relentlessly focused on outcomes that matter to the public. Done well, public sector innovation can build not just better services, but stronger civic trust. Because when government gets it right—when it innovates thoughtfully, delivers consistently, and centers the people it serves—it doesn’t just meet expectations. It reshapes them.