  This urlLink article is one of many critical indictments of charter schools that only casually and somewhat indirectly mentions what is the central point. Students whose parents move them out of traditional public schools into charter schools do so for a reason - either there is something perceived to be bad at their 'old' schools or they believe there is something desirable about charter schools. Further, many charters come about solely to serve "at risk" students, so comparing charter test scores to traditional public school test scores is misleading at best. From a research design perspective, what is needed is to treat students as units of analysis (not schools) and randomly assigning equivalent students to charters and to traditional schools and tracking their performance over time with repeated equivalent measures.
Absent this, be very suspicious about these kinds of voodoo reports because charter kids are by definition not equivalent to the larger public school population. Public education is fraught with problems - not the least of which is that they are immune from the competitive pressures of the market. Ineffective teachers are very hard to fire or demote yet receive the same pay as "teachers-of-the-year". Pay is low, mitigating the ample supply of talented labor. Discipline is a major problem and the 'behaviorally challenged' consume significant school resources.
No Child Left Behind has some market-like incentive structures but they are school- (not teacher-) based. So the idiot teachers who are teachers because they have no other labor market options are lumped in with the high-achieving, inspirational, and innovative teachers. Plus there are too many administrators and in my view, too few parents willing to play an active role in the education of their children. A mess. 
