Economist Ray Fisman asks in a column why public schools are so bad at hiring good instructors. He finds some partial answers in the research literature of his fellow economists who assert there's little that predicts how effective a new teacher will be in the classroom. Obviously that leads to a problem. If good teachers can?t be picked out in advance, then principals (or somebody) have to have the gumption to look at what teachers are doing once on the job and get rid of the bad ones.

That worked for Anthony Lombardi, the principal of P.S. 49 in New York City, who helped his school go from lackluster results to shining ones by discharging about a third of his teachers. But an even better approach, according to Lombardi, would be an apprenticeship and tough evaluation for new teachers. Under such a system, only good bets would get tenure, so few bad teachers would be floating around with the union protections that make them hard to fire.