Giving More Credit than is Due

Simon van Zuylen-Wood has a nice piece on the Web site for The New Republic about Newt Gingrich’s “Earning by Learning” plan for education reform.  This is the one that involves giving kids cash incentives for such things as reading books over the summer.  Zuylen-Wood offers up a reasonable summary of hard research data, from which one can conclude that such incentives work some of the time on some of the subjects.  Indeed, the data seem to indicate just when it does work and for whom.  This allows him to conclude with the punch line:

Newt might have been on to something.

This seems to imply that someone who comes up with a sweeping generalization, disproven in its full generality, “might be on to something” for the special cases that confirm the usefulness of the recommendation.  That kind of approval is dangerous.  All we need is more shoot-from-the-hip generalizations that will do more harm than good if not scrupulously examined.  When was the last time anyone knew Newt to do anything scrupulously?