In 2013, Angela Caputo, then working for the Chicago Reporter, looked for "million-dollar blocks" in Chicago—single city blocks where, over a decade, the imprisonment of residents of those blocks cost the state at least a million dollars. She found almost one thousand of them in the city.

Now two researchers, Daniel Cooper (co-executive of the Institute of Social Exclusion at Adler University) and Ryan Lugalia-Hollon, and the local civic-tech company DataMade have mapped those million-dollar blocks (and ones that don't rise to that level, but still represent hundreds of thousands of dollars in prison sentences). They looked at a shorter timeframe, just five years, but the numbers are still astonishing—851 million-dollar blocks, 121 of which rang up $1 million in sentences just in non-violent drug offenses. By Caputo's calculations, prison sentences for residents of one census block in Austin cost four million dollars over a decade; by the calculations of the DataMade collaboration, the six of the seven blocks between Madison and Jackson on the west side of Cicero in Austin, around where Caputo was writing about, cost more than one million dollars each over five years. Two of them cost more than two million dollars.

The million-dollar blocks are scattered throughout the city, but the vast majority are concentrated in one area on the West Side that runs from North Lawndale and East Garfield Park west through Austin. Not surprisingly, this matches up with maps of 2014 arrests for marijuana (12,483 over 10 months) and heroin (5,294), which cluster in the same area.

Why the West Side, and not so many other troubled neighborhoods in Chicago? A major piece of the puzzle is almost certainly the west-side drug markets that have surrounded the Eisenhower for decades, which provide easy access to suburban heroin users.

There are alternatives; the new governor has been saying the right things, and recently the Illinois Criminal Justice Information Authority released a report evaluating the county's deferred-prosecution program, which is meant to keep first-time non-violent felony offenders (mostly retail theft, possession, burglary, and theft) from going to prison. And despite problems—especially funding—the program has, in some ways, worked:

An average of 35 individuals per month are admitted to DPP each month since the inception of the program. Examining a sample of those (695) in the impact evaluation, 68.6% (477) successfully completed the program and, in turn, avoided a criminal conviction. Thus, although the re-arrest rates for DPP participants and comparable defendants adjudicated through traditional prosecution were the same, these successful DPP participants avoided the stigma of a felony conviction. The impact of this cannot be overstated–felony criminal convictions can significantly impact an individual’s ability to find employment, stable housing, and advanced education. Thus, although DPP may not reduce the likelihood of re-arrest, DPP significantly reduces the future collateral consequences of a criminal conviction for all individuals who complete the program.

In short: it doesn't work better than locking people up, at least in terms of preventing future arrests, but it doesn't work worse; it's cheaper, and fewer felonies might have better long-term outcomes for the participants.

But its numbers are small—just 35 diversions a month, which added up to 477 people avoiding a felony rap over the one year the ICJIA looked at—and eligibility is pretty selective in criminal terms, cutting out anyone with a prior felony conviction, "violent offense" misdemeanor, or charges for delivery, intent to deliver, or manufacturing of controlled substances. Another conclusion from the million-dollar blocks map: That's clearly a lot of people.