As Eric Hanushek has proven in a recent working paper, teachers make huge amounts of money. Maybe not for themselves, but certainly for their students and for their country. They can help students in classes of 20 generate marginal gains of more than $400,000 in present value of future earnings, and even more if they teach in larger classrooms. And if we get rid of the worst-performing five-to-eight percent with just average teachers, we can move America near the top of the world’s math and science rankings, bringing in $100 trillion in present value of future gains.

The calculations behind these conclusions may be complicated but the message is simple: effective teachers are extremely valuable.

We all feel intuitively that good teachers provide more value to our children than poor ones, but now we have economics to match our emotions. We also have a major policy implication for school reform: a logical way to improve our education system would be to fire our worst teachers and replace them with average teachers.

But how would we pull this off?

It’s All in the Numbers

Identifying the worst five-to-eight percent of teachers is possible but it’s not possible to be completely accurate about it. To make sure we got the worst teachers out of the classroom, we’d have to identify them at a higher rate, maybe as high as 10 percent.

With a conservative estimate of three million active teachers, that would mean firing 300,000 of them and replacing them with 300,000 teachers who were significantly better. Where would we find an additional 300,000 teachers? And given that we want average teachers as replacements, we might need to find twice as many, or 600,000, in order to throw half of them out.

Obviously, we’re not going to do this all in one day.

Our worst teachers are often our youngest teachers. Mr. Hanushek and others acknowledge that it takes most teachers three years to hit their stride. But many of our long-serving teachers are also among the worst. This too must also be acknowledged. So we probably need to hold off on the pink slips for new arrivals. This means letting more experienced people go and then hoping new hires reach at least average performance levels three years later. It’s hard to predict the success of a teacher before he or she starts teaching. We might end up in a cycle of removing low-performing teachers, recruiting potential replacements, and then removing many of the replacements who didn’t reach average levels of performance.

But it gets even more complicated. It’s also important to consider how individual schools or districts might be affected. The lowest-performing 10 percent of teachers in a ritzy suburb might be a very different population than the lowest-performing 10 percent of teachers in a remote rural area.

Obviously, we’d need a standard for teacher quality and a uniform evaluation system that cut across schools and districts. Then, when we find out what we already know—that some schools have better teachers than others. Forced transfers may be needed to keep classrooms filled as schools filled with low-performing teachers are disproportionately affected.

Mr. Hanushek’s findings, and the obvious policy implications that arise from them make perfect sense. But they might not play out so sensibly in the field. We have over 15,000 school districts in our country and over 100,000 schools. That makes it hard to balance supply and demand.

We must also look at how it looks for kids. While Mr. Hanushek’s work is macroeconomic in nature, and his conclusion is stated in terms of teachers we’d need to fire and replace, it also has a microeconomic dimension that can be expressed in terms of the life potential of a single student. Having a string of better teachers, as opposed to a string of worse teachers, makes a huge difference to a child. (Dropout Nation has elaborated on this in its collection of reports and podcasts on teacher quality issues). There’s quite a risk, then, to our children in simply leaving weak teachers in the system.

Building Better Teacher Recruiting, Training and Evaluation

With a reliable way to evaluate teachers nationwide, and tens of thousands of reasonably talented people to replace our poorest performers, who wouldn’t rush to implement a policy of “ire and hire” onsistent with Mr. Hanushek’s work? But such conditions don’t exist at present. At the same time, Mr. Hanushek’s work makes clear that allowing children to languish in the classrooms of low-performing teachers isn’t fair. We know the cost to both kids and country. We can’t continue to turn a blind eye.

There are risks on both sides. We tend to focus primarily on the risk to children but adults are at risk of losing their jobs. And even if they haven’t performed those jobs well, they were certified, hired, and favorably evaluated, sometimes for many years. Most would have no idea they were some of the least successful educators in the system.

The challenge, then, lies in balancing the risk to children with the risk to adults—between taking on the immense challenge of replacing our nation’s worst teachers and allowing the many children whose lives they touch to be negatively affected.

So how do we do this? Here are three suggestions:

Residencies for Early-Career Teachers: Most teachers, and much research, tell us that years one through three are the crucial years in a teacher’s career. These are the years that often determine whether teachers stay or go and how effective they become later on. So we need to adapt the techniques we use to prepare doctors. After working hard in school, they work even harder in the field under the guidance of experienced professionals. They are, technically, doctors. But until they successfully complete their residency, they can’t move on in their careers.

The end of year three is the most appropriate place to determine whether or not a new teacher should remain in the system. A three-year residency, with ample support from master teachers, would be the best way to develop high quality teachers for long and successful careers.

New Evaluation Systems: We can argue all day about the quality and fairness of evaluation systems, but we can’t make much progress without them. The traditional approach, where virtually all teachers receive satisfactory ratings based cursory observations and little or no valuable feedback, is of no help to anyone.

Evaluation systems need to be more rigorous and more comprehensive. They need to include more inputs like multiple classroom observations by multiple observers, student achievement data, interviews, personal statements, and student work samples. We also need to consider the value a teacher brings to his or her colleagues. A school is a community and a teacher is more than just an isolated set of statistics.

A  Fatter Pipe of New Talent for Career Classroom Teaching: If we’re going to be moving some teachers out, we have to have new teachers to replace them, and those teachers have to be better. This requires two things: a larger pool of talent to begin with and individual members of that talent pool who are committed to career classroom teaching. It does little good to bring talented new people into the classroom if they only stay for two or three years. We can’t raise the quality of teaching overall on a national level with a revolving door approach. (My Dropout Nation colleague, RiShawn Biddle, slightly disagrees, arguing that this problem is easy to solve with robust recruiting and training, along with more-rewarding and socially-entrepreneurial career paths.)

These three recommendations have three things in common: they require more people, they require more money, and they suggest a national dimension to the teacher quality issue. Residency programs require master practitioners. Evaluation systems require larger numbers of well-trained evaluators. And a fatter pipeline of talent means getting more, and more talented, people into teaching—people who actually want to be career teachers.

What’s the national angle? Part of Mr. Hanushek’s thesis is about the economic value of raising average student performance throughout our country by raising the proportion of average teachers throughout our country. If only a few districts apply these ideas, little benefit accrues. Although I will acknowledge that many cities, including New York City and Indianapolis, already have residency and recruiting programs of their own, it’s hard to imagine every district building their own evaluation systems. The cost of operating all three on their own (or even with partners such as the New Teacher Project, which runs many of these efforts) would be too high. Economies of scale are needed.

National leadership and direct federal involvement in teacher recruitment, training, and retention is a reasonable response. Most of the countries whose education systems we admire have national systems of teacher preparation. This would provide a national benchmark for teacher quality and help balance supply with demand.

School has always been a state issue. It’s difficult to see ourselves doing anything meaningful in education related to teacher quality through coordinated federal leadership. But we are coming closer to realizing the necessity of national approaches, and Mr. Hanushek’s research is of many reasons why.

Ten years ago, the notion of national curriculum standards and national testing didn’t appeal to anyone. Now, more than 40 states have adopted the Common Core State Standards Initiative, and by 2015, two multi-state assessment consortia will give us a near-national testing system. So even when we run up against states’ rights and local control, we can live with national approaches in education that make sense to us. Knowing the incredible value of improving teacher quality, why couldn’t we live with a national approach in this case?

Glass Half Full or Half Empty?

As the saying goes, “If you think education is expensive, try ignorance.” Now we know just how expensive ignorance is—and how teacher quality factors into that equation. Mr. Hanushek’s findings may look like the perfect justification for mass firings. But they are also a stunning validation of good teaching. In addition to developing more rigorous teacher evaluation systems, investments in developing more good teachers are worthy of consideration. On a national scale, teaching quality is a vital component of economic success.

We can’t just focus on mass firings (although I must also acknowledge that most reformers aren’t calling for that). In fact, from where I sit, it’s a bad idea. We should choose to focus on mass improvements instead. We should certainly be more restrictive about letting extremely low-performing teachers enter and stay in the field. More important than that, we must support teachers in their first three years so we can increase the value of all teachers operating above minimum levels of effectiveness.

As overall teaching quality improves, it makes sense to raise pay for teachers, and to raise it more for exceptional teachers. If consistent and permanent performance-based pay grades replaced inconsistent and temporary merit pay schemes, teaching might become more economically viable for the next generation of teachers.

As a result of Mr. Hanushek’s work, the value of a teacher just went up—sharply. Economic value isn’t the only measure of a teacher’s a worth. But it’s a powerful metric in a free-market society such as ours.

Teachers are valuable and we know how much they are. We can now do more to bring better teachers into the field, support them so we can improve their effectiveness, and keep good-to-great teachers in our classrooms.