Mixed-use neighborhoods that combine residential and business development may help lead to lower levels of some types of violent crime, a new study suggests.

The results were just as true in impoverished as they were in more affluent areas, offering one possible way of improving blighted areas, according to the researchers.

But the findings come with an important caveat. In a sparsely populated neighborhood, increases in business-residential density actually leads to higher levels of violent crimes, at least for a while. However, after building density reaches a certain threshold, some violent crime begins to decline.

"A residential neighborhood needs more than the addition of one or two businesses to see any positive impact on violent crime," said Christopher Browning, lead author of the study and professor of sociology at Ohio State University.

"There needs to be a sufficient density of businesses and residences throughout the community to really see the benefits."

The study appears in the current issue of the Journal of Research in Crime and Delinquency.

The findings are important as more cities across the country consider mixed-used developments as a way to reinvigorate downtowns and blighted neighborhoods, Browning said.

But one issue has been whether these mixed-use neighborhoods have positive or negative effects on . Some people have argued that businesses attract more foot traffic to neighborhoods, and the increased street activity brings more "eyes on the street," which then helps reduce crime.

However, others believe that increasing the numbers of businesses causes residents to withdraw into their homes as they see more strangers in their communities, which makes crime more likely.

This study is one of the first to examine which of these theories is correct, at least in reference to some violent crimes.

To explore the issue, Browning and his colleagues examined data from 184 census tracts in Columbus, Ohio. They determined how much of each census tract was devoted to businesses and how much was residential in 2000. They then examined 1999-2001 rates of homicide, aggravated assault and robbery in each census tract.

They focused on these violent crimes because they are generally more likely than some other crimes to occur in public spaces where "eyes on the street" may play a role in stopping their occurrence.

The researchers found that the number of homicides and aggravated assaults increased in low-density mixed-use neighborhoods as the density increased, but then began to decline after a threshold of density was met. However, robbery levels continued to increase right along with commercial-residential density.

Browning said robbery, unlike other violent crimes, may be a more strategic crime that is easier to hide from potential witnesses, and thus not as susceptible to control by watchful citizens.

But, in general, Browning said the results show that the interaction of customers visiting businesses with residents of a neighborhood really does help cut violent crime - at least when there's sufficient density.

The results can't tell us how much business-residential density is required to cut crime for any individual city, Browning said. However, in Columbus, homicide and aggravated assault levels started to drop when density levels reached about average for the city.

The lesson is that, in areas with low density, a few new businesses may bring strangers to a neighborhood, but without enough people around to keep a watchful eye on what's going on, leading to an increase in , Browning said.

"You can't develop a mixed-use community in a limited way, with just a few businesses in one corner of a neighborhood. You need enough business and enough housing to have a vibrant pedestrian community, with people walking around and watching what's going on around them."

Browning said it was especially encouraging that, even in disadvantaged neighborhoods, mixed-use developments could help cut some violent crimes.

"Some people have wondered if mixed-use developments are only helpful in preventing crimes in more affluent areas, but that's not what we found," Browning said.

"If anything, mixed land use was slightly more effective in preventing crime in disadvantaged areas. That suggests there is the possibility of creating more viable streets and public spaces in blighted neighborhoods with properly planned development," he said.

Browning conducted the study with Ohio State colleagues Catherine Calder, associate professor of statistics; Lauren Krivo, professor of sociology; Mei-Po Kwan, professor of geography; and Ruth Peterson, professor of sociology. Other co-authors included Reginald Byron of Southwestern University in Georgetown, Texas and Jae-Yong Lee of the Korea Research Institute in Daejeon, Korea.