Over the past few years, the discussion of crime and violence in the United States has focused on police brutality, mass incarceration and the sharp rise in violence in cities such as Baltimore, St. Louis and Chicago. This is appropriate. Any spike in violence should attract attention, and redressing the injustices of our criminal justice system is a matter of moral urgency.
But it is also worth reflecting on how much the level of violence has fallen in this country over the past 25 years and how widespread the benefits of that decline have been.
From the 1970s through the early 1990s, the murder rate in some American cities rose to levels seen only in the most violent, war-torn nations of the developing world. In the years since, violent crime has decreased in almost every city, in many cases by more than 75 percent.
For well-off urbanites, the decline of crime is most visible in sanitized, closely guarded city spaces where tourists and others can now comfortably wander. Far more consequential have been the changes in low-income, highly segregated urban communities. Indeed, my research has shown that the most disadvantaged people have gained the most from the reduction in violent crime.
Start with lives saved. Though homicide is not a common cause of death for most Americans, for African-American men between the ages of 15 and 34 it is the leading cause. This means that any change in the homicide rate has a disproportionate impact on them.
The sociologist Michael Friedson and I calculated what the life expectancy would be today for blacks and whites had the homicide rate never shifted from its level in 1991. We found that the national decline in the homicide rate since then has increased the life expectancy of black men by roughly nine months.
That may not seem like much, but it is exceedingly rare for any change in society to generate such a degree of change in life expectancy. For example, researchers have estimated that if the obesity epidemic in the United States were wholly eliminated, life expectancy would increase by a similar amount.
The drop in homicides is probably the most important development in the health of black men in the past several decades.
The decline in violence on city streets also occurred inside public schools, creating environments where students could learn without fear of being victimized. Analyzing statewide tests of academic achievement, I found that test scores have risen the most, and the gap in the average scores of black and white children has narrowed the most, in those areas where violence has fallen most sharply.
The drop in violent crime has led better-off families to move into poorer city neighborhoods, thus reducing the concentration of poverty in urban America. Though gentrification has become a problem in a few prominent places, in most cities there is no good evidence that poor families have been pushed out of their neighborhoods as violence has fallen. In fact, as research I conducted with the doctoral student Gerard Torrats-Espinosa shows, the crime decline has improved the prospects for upward mobility for the poorest American families.
The everyday lived experience of urban poverty has also been transformed. Analyzing rates of violent victimization over time, I found that the poorest Americans today are victimized at about the same rate as the richest Americans were at the start of the 1990s. That means that a poor, unemployed city resident walking the streets of an average city today has about the same chance of being robbed, beaten up, stabbed or shot as a well-off urbanite did in 1993.
Living in poverty used to mean living with the constant threat of violence. In most of the country, that is no longer true.
Critics will note that cities, in their efforts to reduce crime, have relied heavily on controversial tactics such as aggressive policing and an expanded prison system. That is correct. But reducing violence does not have to rely entirely on the police and prison, nor has it: Nonprofit organizations, my research has found, have played a critical and underappreciated role.
To understand how cities have changed since the 1990s, I gathered data with Torrats-Espinosa and the doctoral student Delaram Takyar on city demographic characteristics, public and private security forces, business establishments and public institutions. As expected, we found that police forces expanded and the imprisonment rate skyrocketed and that those changes probably contributed to the crime drop. But we also noticed that there was a huge increase in the number of nonprofit organizations developed specifically to build stronger communities or to confront violent crime.
To find out whether these types of organizations had an impact on crime rates, we looked for situations in which anti-violence nonprofits were formed not in response to a rise in violence but because new sources of funding became available to community groups and leaders. (In those situations, we can more confidently assess whether the newly formed organizations had a causal impact on the level of violence.) We found that in a typical city with 100,000 people, each additional nonprofit devoted to confronting violence led to a roughly 1 percent drop in the city's murder rate. Considering that this segment of the nonprofit sector grew by about 25 organizations for every 100,000 residents in New York and elsewhere, community-based organizations appear to deserve more credit than they get for contributing to the fall of violence.
These findings suggest a new model for combating urban violence. While police departments remain crucial to keeping city streets safe, community organizations may have the greatest capacity to play a larger role in confronting violence. Working directly with law enforcement and residents, these organizations are central to the next stage in the effort to make our cities even safer.
Patrick Sharkey is a professor of sociology at New York University and author of the forthcoming book "Uneasy Peace: The Great Crime Decline, the Renewal of City Life, and the Next War on Violence." He wrote this article for the New York Times.
