The Perception Paradox: How Misaligned Public Beliefs About Crime Undermine the Full Continuum of Public Safety - R Street Institute

The Perception Paradox: How Misaligned Public Beliefs About Crime Undermine the Full Continuum of Public Safety

Author

Table of Contents

- Executive Summary

- Introduction

- Why Public Perception of Crime Matters

- The Widening Gap: Belief Versus Data

- Crime Is Real

- Drivers of the Divergence

- How Public Perception of Crime Shapes Enforcement

- How Misalignment Distorts Public Safety Priorities

- Treating Perception of Crime as a Core Public Safety Domain

- From Evidence to Practice

- Key Recommendations

- Conclusion

- Appendix A: Documented Enforcement Cases

- Appendix B: Coding Key

Media Contact

For general and media inquiries and to book our experts, please contact: pr@rstreet.org

The current moment offers a rare opportunity—crime rates are falling, crime data is steadily improving, and public fear is finally beginning to ease. Together, these conditions make it possible to institutionalize the strategies that produced the decline and to bring public belief into line with the evidence before the next cycle of alarm reopens the gap. That work is within reach.

Executive Summary

National crime data has moved in an encouraging direction for the past several years. Data from the Federal Bureau of Investigation (FBI) shows that violent crime fell an estimated 4.5 percent in 2024, bringing the national violent crime rate to its lowest level since 1969, while murder rates declined 14.9 percent. Final 2025 data shows the improvement accelerating, with the largest year-to-year declines in violent crime and murder rates since the FBI estimations began in 1936 and a murder rate tied for the lowest ever recorded.

While the trend is genuine and worth crediting, many Americans do not yet feel the difference. In 2023, a record 63 percent of U.S. adults called crime an extremely or very serious national problem, and those who reported being afraid to walk alone at night near their home reached a three-decade high. Perceptions have since moderated: By 2025, the serious-problem figure had fallen to 49 percent and the fear measure to 31 percent. Even so, the gap between what the data shows and what the public believes remains substantial. More important than the direction of the gap is its persistence. Public perception of crime has repeatedly deviated from the measured reality of crime statistics. This discrepancy is not a polling artifact. Because perception drives funding, enforcement priorities, deployment, and trust, a persistent misalignment between belief and data skews all four, which is itself a public safety problem.

This study argues that crime perception is a core public safety variable. Its central claim is that public perception of crime, however loosely tied to the data, functions as an operating input to the public safety system—one that shapes appropriations, drives the deployment of officers and investigators, and conditions the public trust on which the whole effort depends. When that input is poorly calibrated to actual harm, the costs carry through every stage of the continuum, from deployment to the back end of the system. An agency that ignores what its community fears would forfeit legitimacy, so the practical question is one of calibration—whether the resources that public perception sets in motion land where harm is concentrated, and what it costs public safety when they do not.

To show how this works in practice, this study inverts the question that dominates research on policing and public opinion. Most of that research asks how police tactics shape what communities believe; this study asks the reverse—how does community perception shape what police do? Drawing on a national set of 51 documented enforcement actions from 2021 through 2026, each publicly tied to a community concern, complaint, or publicly identified community crime problem, it finds a consistent pattern in which visible, complaint-generating problems reliably draw visible enforcement responses, sometimes well aligned with evidence and sometimes not.

Introduction

Reported crime in the United States has been declining for most of the past three decades, but public belief has rarely reflected that decline. After a significant pandemic-era rise in violence that peaked in 2020 and 2021, crime rates have fallen for multiple consecutive years, with 2025 figures being among the lowest recorded in decades. There is reason to view that rebound as more than good luck. A recent macro-criminological analysis modeled five decades of national crime against 15 gradual structural conditions, from residential turnover and homeownership to civic engagement and guardianship, and found that those conditions kept improving through 2020 and 2021 even as homicide jumped by roughly one-third in a single year; once the disruptions of the pandemic era receded, homicide settled back onto the trajectory those 15 structural conditions predicted all along. The spike, in other words, behaved like a temporary shock to a stable system rather than a reversal of it. Public belief, however, absorbed the shock immediately and has been far slower to recover. For much of the same period, most U.S. adults polled believed that crime was rising and reported feeling less safe.

The distance between measured and perceived realities is not new, but its recent scale is striking, and the mechanisms sustaining it have grown more powerful. A fragmented media environment, an always-on stream of viral incident footage, and the deepening partisanship of crime as a political issue have combined to keep public fear only loosely tethered to underlying data. Genuine confusion within the data has compounded the gap: For several years, the nation’s two principal crime measures—the FBI’s Uniform Crime Reporting (UCR) program and the Bureau of Justice Statistics’ (BJS) National Crime Victimization Survey (NCVS)—pointed in opposite directions, making it reasonable for people to wonder which was accurate.

Whatever its sources, the gap raises a question the literature has largely left unasked. Research on crime and public opinion has overwhelmingly focused on how policing tactics shape what communities believe and feel about crime. This study deliberately reverses that inquiry. It treats community perception as an input rather than an output and asks how that perception shapes the behavior of law enforcement and the broader public safety enterprise: how fear and complaint drive deployment, how political pressure translates perceived danger into command-level priorities, and how the allocation of scarce public resources follows belief as much as data.

This study arrives at a significant moment. Public alarm about crime has eased from its 2023 peak, and crime data itself has become timelier and more accessible than at any other point in the modern era. The FBI now publishes national data monthly rather than annually, and in May 2026 issued its first preliminary annual estimates. In addition, independent projects such as the Real-Time Crime Index and the Live Crime Tracker maintained by NORC at the University of Chicago aggregate hundreds of local open-data feeds into national trend dashboards. Together, those two shifts create a rare opening to recalibrate—to lock in the gains, credit the strategies that produced them, and align future investment with evidence rather than with the anxieties of the moment. This window of opportunity is unlikely to remain open indefinitely. Crime discourse remains deeply politicized, and the mechanisms that sustained the perception gap over the past two decades—media salience, partisan framing, and the amplification of isolated incidents—remain fully intact and capable of reopening it.

The analysis that follows is organized around a single proposition: narrowing the distance between public perception and measured reality is itself a public safety strategy, one with direct implications for the efficient allocation of taxpayer resources, the safety of communities, and the effectiveness of the officers who