It seems everyone is down on bad teachers these days. But the truth is that simply removing the bad apples won't fix our education problems. After all, it's not as if there's a large pool of superstar teachers waiting to replace them. Our best hope to improve education broadly and deeply is to strengthen the programs that develop and prepare the vast majority of the nation's teachers.

Unfortunately, many of those systems are in sorry shape. Research has found that teacher preparation programs frequently have low admissions requirements, low exit criteria and a lack of academic rigor, and provide training disconnected from what teachers will experience in actual classrooms.

The result is that beginning teachers often walk into their new schools with very little idea how to handle and teach a classroom full of kids.

This lack of preparation is unlikely to be measured — or even noticed. Few teacher education programs measure their graduates' success as teachers, or ask graduates or their employers to evaluate the quality or relevance of their preparation. Most school districts have no system in place for evaluating how one program compares to another.

Although states are supposed to identify and assist teacher training programs that are low performing, they rarely do. Out of more than 1,400 colleges of education across the nation, states identify only 38 as having low-performing programs. That doesn't square with the widespread quality concerns raised by outside observers.

About a dozen years ago, Rep. George Miller, D-Concord, led a federal effort to establish a better quality-control system for teacher preparation programs. He succeeded in getting a federal requirement that education schools must report certain outcomes, such as how many program completers pass teacher licensure tests. But that kind of information, besides being wildly inconsistent across states, says almost nothing about whether graduates are effective teachers.

That would mean tracking more meaningful measures of success, including information about whether graduates actually get (and keep) teaching jobs, what beginning teachers and their employers think of the quality of their preparation, and whether their teaching has a measurable effect on the achievement of students in the classroom.

Sound impossible? It's not. Louisiana has developed a strong data system that allows it to track where graduates of different preparation programs go and what kind of effect they have on K-12 student achievement. The state also asks new teachers to rate how well their teacher education programs prepared them for their first year of teaching.

In California, the Cal State system voluntarily surveys program graduates about the quality of their preparation, both at graduation time and during their first year of teaching. The university system also surveys employers and supervisors of graduates and works with several large districts to determine whether the graduates of some programs are more effective teachers than others.

States should also ensure that universities use this data to improve teacher prep programs. For starters it could include tightening standards for program entry and exit.

Programs should also focus more on clinical training, and universities should dedicate more tuition revenue to teacher preparation. School districts should carefully collect data on how a program's graduates perform in the classroom. And if, year after year, the data show dismal results for particular programs, then it's fair for the state to shut them down, as Secretary of Education Arne Duncan said should be done.

But again, as with teachers, getting rid of a few bad programs isn't nearly as important as strengthening the programs that will continue to educate our teachers. With unemployment so high, it may be hard to imagine a time when many more new teachers will be needed, but as baby boomers retire over the coming decade, we'll need new teachers. And for the sake of children, we need people who are ready for the job on Day One.