Wonders never cease. With all the fanfare of a Capitol Hill briefing attended by various politicos, Teachers College released a report this week advocating five "proven" educational interventions as means to reduce the high school drop out rate by half--and saving the country $45 billion annually.

What are these five magic bullets? Two relate to better preschool. Two more call for small learning communities and reducing class size in the early grades. The final one is giving teachers a 10 percent across-the-board salary increase.

We'll leave the first four to others to parse out, but the news that raising teacher salaries 10 percent could lower the dropout rate set us scurrying to the footnotes. Unfortunately, the research supporting the interventions isn't in the footnotes. You have to download the technical appendix to find out that the authors (led by Henry Levin at Teachers College) based this assertion on a single study. The study, while obtuse, was credible, done by solid researchers Susanna Loeb and Marianne Page, but it is one study up against pretty high stack of others asserting the reverse. Do consensus findings mean anything?

Levin et al overlook a key point in Loeb and Page's study. They found that raising salaries 10 percent in one district, makes that district more competitive for quality teachers. They don't examine the impact of raising salaries across the board in all school districts, as Levin et al propose, thereby reducing, if not eliminating any market impact from the intervention. Just a minor detail, to be sure.