A bummer for Rand Corp.

Are medical marijuana dispensaries magnets for crime? That question matters, because the assumption that such facilities are neighborhood nuisances is propelling a drive by Los Angeles and other California cities to craft regulations that limit the number of dispensaries and where they can operate. So when Rand Corp. came out with a study last month that seemed to arrive at the opposite conclusion, marijuana advocates stood up and cheered.

The cheering stopped Monday, when Rand retracted the study.

Rand isn’t given to sloppy research, which is what makes its junk-science study so surprising. The Santa Monica-based think tank examined crime statistics for 10 days before and after June 7, 2010, when the city of Los Angeles ordered more than 400 illegal dispensaries to close down, and found that crime increased dramatically near the closed facilities. Researchers speculated that dispensaries, perhaps because they employ security measures such as cameras, might actually reduce neighborhood crime rather than increase it. Yet that was so counterintuitive that it should have raised red flags during the peer-review process.

It certainly raised red flags with the city attorney’s office, which pointed out that Rand had no way of knowing whether the dispensaries it studied actually closed June 7, and that crime might have increased because going-out-of-business sales drew criminals to the vicinity. But that’s not why Rand retracted its study. It turns out that researchers relied for their crime data on a websitethat didn’t include statistics from the Los Angeles Police Department.

Rand officials say they’ll redo the study after considering LAPD data. That’s nice, but it still won’t fix the key problem with the study design. Even if the analysis had been perfect, it would have answered a question no one is asking: What happens to crime in the immediate aftermath of a dispensary’s closing? What residents of California really need to know is: What happens to crime in the long term after a dispensary opens?

To find out, researchers would have to look at crime near dispensaries for, say, six months or a year before and after they opened. That would be hard to do because it would require comparing neighborhood crime impacts with changes in crime citywide; moreover, it may be that dispensary owners choose to locate in neighborhoods that are gentrifying, so changes in crime wouldn’t necessarily be attributable to a pot shop’s arrival. But until Rand or somebody else figures out how to balance all the variables, we’ll remain in the dark about the relationship between cannabis and criminals.