In a letter to the editor published in Southern Maryland Online, a representative from the American Beverage Institute cites studies that show that roving patrols are far more effective in arresting drunk drivers than organized, visible roadblocks.

It is a reasonable assumption that roadblocks are not a particularly effective use of the limited enforcement resources of the police department. Big, showy roadblocks may make people believe that they are helping reduce drunk driving more than standard, quality police work, and now the evidence supports that belief.

The question is, does stopping and questioning people randomly to check for signs of intoxication make more sense that actually trying to identify people who appear to be driving dangerously or erratically?

And evidence shows the answer is that these efforts don’t make sense, and are an inefficient use of police resources in enhancing public safety.

Anyone who has ever driven on a highway any night of the week can tell you that there is no shortage of people who are driving unsafely or too quickly. You could pull over every third car for exceeding the speed limit. And, though more time consuming then simply stopping ever car at a checkpoint, at least those individuals would be actually suspected of doing something wrong.

It is always helpful when individuals and groups come together to honestly and objectively assess the best ways to prevent drunk driving. Too often, the neo-prohibitionist groups like MADD are the only voice in the discussion.
By the way, the American Beverage Institute is an industry group that represents bars and restaurants that serve alcohol, and is an important voice in fighting drunk drivers at the source.