Some jails are notorious. Think New York City’s Rikers Island or the Los Angeles County Men’s Central Jail. News stories about overcrowding, violence, and deplorable conditions fuel ongoing public debate about the nation’s two largest jail systems and capture the public's imagination about just what jail looks like. But it turns out urban jails are in decline—there is even a movement to “close the jail” in New York City; Los Angeles is already tearing down its largest jail and building a smaller one—and it is rural America that represents the true picture of U.S. jails today. That's because growth in the jail population is not driven by the largest counties; it has taken root in a thousand very small ones across the United States.

Of course, it wasn't always like this. The nation’s very small counties once had less than half as many people in jail as New York City and Los Angeles combined. Now, it is the very small counties that have double the combined jail population of the two cities. Original analysis of the Vera Institute’s online jail population tool show that jails have grown the most in small counties, not large ones. In the last decade, the outsized jail growth in very small counties has only continued, but jail populations in larger counties have actually begun to decline.

To illustrate this, I conducted additional analysis to compare two groups of counties—each with a population of 18.6 million. The first group: Los Angeles County and New York City, which have a combined resident population of 18.6 million in 2014, and are also the largest—and perhaps most notorious—jail jurisdictions in the United States. The second group: 1,003 very small counties, each with between 10,000 and 30,000 residents in 2014, and also with a combined resident population total of 18.6 million (around one-third of all U.S. counties fall into the 10,000 – 30,000 category). Each group holds 6 percent of the total U.S. population, and has grown at nearly the same rate since 1970.

Of course, there are differences between the two groups. The growth of mass incarceration in local jails is one key difference. From the 1970s to the present, NYC and LA’s combined jail population grew 30 percent, from 23,000 to 30,000 people on any given day. This outpaced the cities’ resident population growth of 25 percent. In contrast, in the very small counties, jail populations started out much smaller. For example, Gonzales County, Texas—with 20,000 residents between San Antonio and Houston—had 2 people in jail in 1970. But very small counties grew far more. The jail populations in these very small counties grew six-fold from the 1970s to the present—from 9,000 to 62,000—and now hold double the amount of people behind bars as NYC and LA. Gonzales County had 87 people in jail in 2013, for a jail incarceration rate twice the national average. Or Marion County, Tennessee— with 28,000 residents outside of Chattanooga—had only 8 people in jail in 1970, and now has 131 in 2013.

Another meaningful difference is in diversity: the combined population of New York City and Los Angeles is about 70 percent people of color, and the very small counties are about 80 percent non-Hispanic whites. To understand the full impact of mass incarceration at the local level, it’s important to understand how it affects people of color. Compared to very small counties, far more people of color live in NYC and LA County. One might expect NYC and LA to have more people of color in jail. But they don’t—very small counties have more people of color behind bars on a given day than NYC and LA.  While data limits mean we can only compare back to 1990, the changes since then are dramatic. In 1990, 33,000 people of color were behind bars in NYC and LA, but only 9,000 were behind bars in the local jails of very small counties. Twenty-four years later, in 2014, very small counties had tripled to 27,000 and NYC and LA had dropped to 25,000. In some very small counties, the change is dramatic: Custer County, Oklahoma held 11 people of color behind bars in 1990 and 114 in 2013—10 fold growth when the resident people of color population had only doubled.

When thinking in terms of populations, might the increasing numbers of people behind bars in small counties be caused by rapidly shifting demographics, particularly in diversifying suburban areas? Though the number of people of color in very small counties has grown, this relatively moderate population growth does not explain the huge increase in jail incarceration. When looking at the changes in terms of rate of jail incarceration, the racial disparities in the very small counties become even more visible. (Looking at rate controls for changes in the population, by taking the number of people of color who are jailed per 100,000 people of color aged 15-64.)

In very small counties, nearly 1,100 out of 100,000 people of color aged 15-64 are behind bars in a local jail on a given day. For NYC and LA, that rate is significantly lower, at just 280. For a national perspective, the jail incarceration rate of people of color is 502 out of 100,000 aged 15-64, which is less than half the rate in very small counties, and significantly higher than the total national jail incarceration rate of 341.

This disproportionate growth is further evidence that the era of mass incarceration hasn’t delivered on public safety. It has, however,  taken a fiscal toll as well as damaged individuals, families, and whole communities. Jails are under the jurisdiction of local stakeholders, and their day-to-day size and operations are not significantly affected by federal or state legislative proposals to reduce prison populations. As we know from looking deeper into the national data, the use of jail incarceration is embedded in the culture and practice of communities nationwide, large and small.

Growing evidence suggests that reform efforts to downsize local jails are catching on in many large jurisdictions. Ways to shrink jail populations safely include alternatives to arrest, expanded pretrial release options, alternative sentencing options, improved drug treatment, and mental health resources. However, in many small communities, there’s little awareness of a jail overuse problem that would spur the adoption of such tools. For national criminal justice reform efforts to be successful, every county will need to understand not only their jail size in relation to historical trends or similar counties, but also the racial disparities it may contain.

With more information about jail trends nationwide—and who they are affecting—small counties can begin the critical conversation about what kind of change is needed in their own backyard.