By David Leblang
Governor Abigail Spanberger opened her term as Virginia’s new governor with a clear-eyed rejection of immigration enforcement as political theater and an affirmation that public safety must be built on evidence, not spectacle.
On January 17, the day she was sworn in, Spanberger rescinded her predecessor Glenn Youngkin’s Executive Order 47—a 2025 mandate that forced state and local law enforcement into formal 287(g) agreements with federal immigration authorities that deputize local law enforcement to perform federal immigration functions.
By returning discretion to local agencies, Spanberger’s Executive Order 10 prioritizes tangible safety: solving violent crimes, earning community trust, and building cooperative policing—proven strategies that truly keep Virginians safe. [W(1] Predictably, the backlash was instantaneous.
Republican leaders, including outgoing attorney general Jason Miyares, declared that “By directing our local law enforcement to stop working with federal law enforcement agencies, our streets have become less safe…” House Minority Leader Terry Kilgore branded the move “reckless,” arguing it will invite a surge in gang violence and “shield” criminals.
Many conservatives hail 287(g) agreements as an indispensable shield against crime. However, an objective examination of the empirical evidence reveals a stark reality: focused policing, not federalization, delivers real public safety.
The “sanctuary” myth vs. empirical reality
The core claim in the 287(g) debate is that “sanctuary” policies—limits on local cooperation with federal immigration enforcement—lead to higher crime. This claim, however, unravels under scrutiny.
A 2020 study published in the Proceedings of the National Academy of Sciences (PNAS) and highlighted by the Washington Post examined data from more than 3,000 counties over a 15-year period. This study found that sanctuary policies have no detectable effect on crime rates, concluding that the presence or absence of a 287(g) agreement is not a lever that moves the needle on public safety.
However, it is vital to distinguish between sanctuary policies and the removal of 287(g) mandates. While the PNAS study shows that “non-cooperation” is safe, specific research on 287(g) partnerships goes further. Recent longitudinal data (2024-2025) found that the activation of 287(g) task force agreements not only did not reduce violent crime but also, in some cases, significantly increased violent victimization risk for immigrant populations. This suggests that adding 287(g) does not decrease crime, implying that its removal is unlikely to trigger the “wave” predicted by critics.
The real cost: resource diversion
Spanberger’s rationale for EO 10 rests on “opportunity cost.” Proponents tout 287(g) as a “force multiplier,” yet evidence shows it often drains resources rather than multiplying them.
The fiscal drain: 287(g) agreements are not federally funded mandates. While ICE covers training, local taxpayers pay for personnel, overtime, and administrative supplies. A 2012 Brookings Institution study found that Prince William County had to raise property taxes to cover the millions in costs associated with its 287(g) program. Similarly, Maricopa County, Arizona, saw a $1.3 million deficit in just three months due to 287(g)-related overtime.
The policing paradox: By diverting officers to civil immigration tasks—which largely result in detainers for minor misdemeanors and traffic offenses—local law enforcement risks diverting time and energy away from solving violent crimes.
The trust gap and underreporting: The most significant resource any police department has is community cooperation. When residents fear that contact with a uniform could lead to deportation, they retreat. This creates a “dark zone” where criminals can operate with impunity because victims are too afraid to call 911.
The politics of symbolic safety
If data consistently highlight the fiscal and social costs of 287(g) agreements, why does rhetoric surrounding immigration enforcement remain so fierce? The answer is clear: symbolic politics. For many, immigration enforcement is a blunt tool for asserting “control.” Even as crime drops, the mere display of federal-local cooperation becomes a public badge of order—no matter what the numbers say.
Spanberger’s Executive Order 10 represents more than a recalibration of local law enforcement priorities—it’s a rejection of immigration enforcement as political spectacle. The evidence for this is now overwhelming: massive federal operations in Democratic cities appear designed less to reduce crime than to create visible displays of federal power in jurisdictions that oppose the administration’s agenda.
Consider the observable patterns. President Donald Trump has explicitly directed Immigration and Customs Enforcement to prioritize enforcement in “the Democrat Power Center” while claiming without evidence that Democratic cities use immigrants to “cheat in Elections.” Yet Republican-controlled states have murder rates 32% higher than Democratic-controlled states, and cities such as Memphis, St. Louis, and Kansas City—all in red states or with Republican governance—have higher homicide rates than Chicago. Despite this, federal enforcement surges have concentrated in Democratic-led cities: Los Angeles, Chicago, New York, Minneapolis, Portland, and Memphis under Democratic mayors.
The deployment of National Guard troops alongside ICE agents—often over the objections of state governors and without consultation with local mayors—has effectively militarized immigration enforcement in a way that blurs traditional distinctions between law enforcement, immigration policy, and political confrontation. In Chicago, the administration attempted to claim credit for the city’s historic crime decline even though the reduction occurred before Operation Midway Blitz began in 2025. In areas where ICE was most active, crime reductions were actually smaller than the citywide average, and 911 calls dropped precipitously—suggesting enforcement operations may undermine rather than enhance public safety.
This creates a fundamental question for states like Virginia: Should local law enforcement resources be conscripted into a national political project that lacks empirical justification? The research answer is clear. Studies consistently show that 287(g) agreements do not reduce crime, while substantial evidence documents their costs: diverted resources, damaged community trust, and reduced willingness among immigrant communities to report crimes.
The 2025 crime data is decisive: homicides fell 20% across all types of jurisdictions, no matter their approach to immigration enforcement. The lesson is unmistakable: immigration policy is not a public-safety lever. Virginia can and must decline to waste resources on a federal agenda that delivers only politics, not safety.
David Leblang is the Miller Center’s Randolph P. Compton Professor and director of policy research. He is also the Ambassador Henry J. Taylor and Mrs. Marion R. Taylor Endowed Professor of Politics and professor of public policy at the University’s Frank Batten School of Leadership and Public Policy.

Clarifying “Sanctuary” Policies, Enforcement Discretion, and Virginia’s Direction
With a new governor and recent executive orders rescinding some prior directives, there’s understandable confusion about what has—and has not—changed in Virginia’s approach to immigration enforcement.
This report argues with valid studies that Governor Abigail Spanberger’s decision to rescind mandatory participation in federal 287(g) agreements reflects an evidence-based public-safety strategy, not political symbolism.
A key part of the confusion lies in the term “sanctuary.” It’s widely used—but legally imprecise.
Here’s what matters:
🔹 Federal immigration law still applies everywhere.
No city or state can legalize unlawful presence or stop federal authorities from enforcing federal law.
🔹 Enforcement vs. enforcement priority are not the same.
All levels of government routinely decide which violations to prioritize based on resources, public safety, and policy judgment. This is true across many areas of law—traffic enforcement, drug policy, code violations, and tax audits.
🔹 What so-called “sanctuary” policies actually do:
They typically mean local governments choose not to:
• Deputize local officers as federal immigration agents (e.g., 287(g))
• Hold individuals solely on civil immigration detainers without a judicial warrant
• Ask about immigration status when it’s unrelated to a criminal investigation
These choices:
• Do not invalidate immigration law
• Do not prevent ICE from acting
• Do not shield people accused or convicted of crimes
They reflect how local law-enforcement resources are allocated—not immunity or obstruction.
🔹 Why this is lawful:
Under the Constitution’s anti-commandeering principle, the federal government may enforce federal law, but it cannot require states or localities to use their personnel or budgets to do so. Cooperation is permitted; compulsory participation is not.
🔹 Why “sanctuary” is a misnomer:
The term implies refuge or immunity—neither is accurate. A more precise description is:
Jurisdictions that lawfully choose not to prioritize certain civil violations, while leaving the underlying law fully enforceable by federal authorities.
Bottom line:
• Enforcement discretion is routine and lawful
• Immigration enforcement follows the same logic as many other areas of law
• “Sanctuary” is rhetoric; discretion is the reality
I look forward to your comments, corrections, or thoughts