Why Crime Policy Should be as Complex as Crime Itself

The Need for a Macro Perspective on Crime

This article first appeared in Vital City.

Crime, we are told, played a key role in the 2024 election. Voters were concerned about rising crime (although national crime appears to have declined in every major category) and cast their ballots for the candidate they believed could address it. Contrary to those who seem to believe that the answer to rising crime is simply hiring police and charging them with more vigorously enforcing the laws, however, crime is a complex challenge — and understanding it demands a comprehensive theoretical framework.

There are many analogs in other disciplines. In physics, scientists must reconcile the behavior of subatomic particles with the structure of the cosmos. In economics, it means integrating theories of individual decision-making with models of entire economies. In health, allowing individuals to make their own healthy choices is juxtaposed with the need for collective action against collective contagion. For the most part, these fields fail to unite the large with the small, and the small with the large, but they learn important lessons in the attempt.

Criminology, by contrast, typically lacks a parallel focus on the “macro” scale. While there are countless theories about individual behavior and community dynamics, the study of crime as a large-scale phenomenon remains fragmented. Unlike macroeconomics or cosmology, there is no dedicated field of “macro-criminology.” Instead, criminology relies on extrapolations from smaller-scale studies, with large-scale assumptions, leaving a significant gap in our understanding of crime at the societal level.

This absence limits us. Macroeconomics is the study of whole economies and their interaction with other whole economies. So, it is the study of inflation, unemployment and GDP; of exchange rates, capital flows and interest rates. The study yields a range of related policy responses: monetary and fiscal policy, trade policy and the like. Without macroeconomics, we wouldn’t understand the trade-offs between inflation and unemployment, interest rates and economic growth.

Similarly, macro-criminology could illuminate new ways to approach crime policy. At the top of that list is understanding that, broadly speaking, there are different policy mechanisms available with different consequences that solve different problems. And there are unavoidable trade-offs of every policy intervention. Without this perspective, we miss opportunities to develop holistic, systemic responses to crime.

It’s time to start thinking about crime with the same breadth and ambition as we do other societal challenges.

A simple, long-term display of the two main crime categories—property and violent—raises a hundred questions. What are the big social forces driving the three-decade spike in crime? What are the social forces driving the three-decade decline? Is a three-decade cycle the natural rhythm of crime, or were the 1970s and 1980s anomalous? Why were property and violence so highly correlated in the 1960s and 1970s, but property rates peak (and flatten) a full decade before violence peaks? Why were the declines in the 1980s and 1990s so similar but the 2010s so different (violence flattens, but property continues a rapid decline)? What can this tell us about the future?

Here, I want to make four quick points about the value that would be added by the emergence of macro-criminology as a discipline in its own right.

• First, I want to point out that crime, like macroeconomics, is cyclical — meaning that it tends to rise or fall over time, like many other phenomena.

• Next is the idea that the forces acting on people and places — especially cities — are complex and intertwined, and policy should be considered in that milieu.

• Another point is that there are sectors in crime, just as there are sectors in economics, and a macro approach can help to understand their shared and unique patterns.

• Finally, macro-criminology would provide useful insights for cities and regions in fighting crime, which must deal with the competing challenges of problems specific to their locale and the larger forces of national economic and public health that exert pressures locally.

Crime is cyclical, and crime policy should be as well

The place to begin is with the cyclical nature of crime, which typically receives little notice. The long-term trends above show some obvious cycling. But drilling down a little, to different types of crime, shows different patterns and cycles. And cities experience their own cycles—some aligned with the national tides, some unique to local influences.

In other fields, there is a clear focus on these cycles. When economic times are good, we raise interest rates and slow down government spending to keep the economy from overheating. When times are bad, we drop interest rates and ratchet up spending. Macroeconomics teaches us that the details of what is good or bad about the economy matter enormously in deciding which course to take.

The analog works for crime, where the tension is between increasing enforcement and maximizing civil liberties. The two instruments of crime policy that are analogous to monetary and fiscal policy in macroeconomics are the criminal justice system and social policy. In times of high crime, increasing the intensity of the justice system is reasonable, even though it comes with social and economic costs. In times of low crime, using broader social policy to invest in people’s well-being is a better choice.

Today, both increasing and declining crime are said to demand higher investment in police and prisons. This is as nonsensical as demanding low interest rates in both declining and rising economic cycles. In the same way cheap money creates inflation in good economic times, so too does overinvestment in the justice system impose unnecessary economic and social costs when crime is down. But without a macro-criminological lens through which to visualize the cycles, the obvious is obscured.

Crime is complicated, and crime policy should be too

If we had a discipline of macro-criminology, it would be clearer that crime policy does not belong in a silo apart from a thousand other policy choices made in cities and communities. Rather, policies around housing, transportation, education, immigration, labor, public health and more all have a substantial impact on the level of crime in a place. To think otherwise is to miss a great opportunity and commit to unintended negative consequences.

But today, we are faced with the strange situation whereby you would be met with a mild shrug if you were to argue that housing, transportation and labor markets were inextricably related — of course they are. But if you were to add a fourth element — crime — to that list, it would suddenly become a political statement. And in some quarters, a quite controversial one.

Patrick Sharkey and colleagues at NYU conducted a groundbreaking study where they found that each additional community-based organization (CBO) in a community reduced crime by several percentage points. One interpretation of that finding is that cities should invest more in CBOs. A complementary idea is that the study shows that siloed government policy has created massive gaps that must be filled, in multiple sectors.

Sectors matter

If you study crime, one of the first truths you learn is that the label “crime,” which we throw around all the time in public parlance, is of little use. Most reported crime is property crime, so any all-crime graphic is mostly just … property crime, which is not what the general public typically means when they talk about crime. Stranger homicides and stranger sexual assault, home invasions, mass shootings, retail smash-and-grabs and carjackings are the high-risk, out-of-the-blue crimes that people most worry about — but they are barely a blip in the national crime averages. They are extremely high social-cost events, but they are rare. With that said, the rare event sector is worth studying and responding to, because these events drive public sentiment, just as Amos Tversky and Daniel Kahneman’s Prospect Theory predicts. But here is the critical point: We know almost nothing about why rare events trend. Do local factors drive them, or do they rise and fall due to large, secular forces?

We also know comparatively little about extremely common events, like robbery and assault. Take the graph above, which shows trends in robberies and assaults. The trends moved together for 15 years, 1960-1975. But then the paths diverge, and from 1975 to 1985, robbery rates are pretty flat while assaults rocket up. From 1985 to 1995, the divergence accelerates, and the gap widens further. The trend is parallel for about two decades, but then the gap begins to grow again in 2015.

Why? Drugs? Guns? Or something more prosaic, like a proliferation of alcohol outlets, because more drinking is clearly linked to more assaults? Or some missed opportunity, where some success in reducing robberies has not translated to assaults? This is another topic that macro-criminology would study.

Cities and regions

I noted earlier that there is a difference between trends driven by local context and secular trends, where secular just means most people in most places experience the same change in crime at the same time. We need only look to about six months ago for evidence of the power of this idea. In July 2024, the Council on Criminal Justice released a midyear report, and the report shows broad but heterogeneous declines.

Of the 28 cities with comparable data for the first half of 2023 and 2024, 19 show a decline in homicides. Seventeen cities show a decline of more than 10%, which is an important benchmark because prior to 2023, homicide (nationally) had never fallen more than 10% in one year.

But the takeaway is clear: It is myopic for cities to report on and respond to their own violence declines or increases without considering the larger (secular) trends. Homicide declined in most places in 2024, and by a lot. So national trends matter. But there are also huge differences across cities. There is a lot to be learned from thinking about both of these topics — separately and together.

A path forward

This is not an academic exercise. The way we think about crime influences the way we address it. We think about it far too narrowly. We need to broaden our lens — and then adapt our policy approaches accordingly.