16 March 2011

The Japanese Business Model For Teacher Pay

There are lots of differences between the prevailing American model and the Japanese model for K-12 education. One of those differences is that Japanese schools often have class sizes that many U.S. schools would consider unacceptably large, but use these larger class sizes to make it feasible to pay teachers higher salaries and generally, to build up the prestige of teaching as a profession. Also, many tasks that would be done by paid staff in American schools, like a significant share of the janitorial work and photocopying, is done by students in Japan, again, freeing up resources for higher teacher salaries. American law professors routinely teach classes of ninety students. Almost no high school teachers in the United States do that, although such large class sizes are unheard of in Japan.

Nicholas Kristoff argues, reasonably, that teachers are underpaid:

Until a few decades ago, employment discrimination perversely strengthened our teaching force. Brilliant women became elementary school teachers, because better jobs weren’t open to them. It was profoundly unfair, but the discrimination did benefit America’s children.

These days, brilliant women become surgeons and investment bankers — and 47 percent of America’s kindergarten through 12th-grade teachers come from the bottom one-third of their college classes (as measured by SAT scores). The figure is from a study by McKinsey & Company, “Closing the Talent Gap.”

Changes in relative pay have reinforced the problem. In 1970, in New York City, a newly minted teacher at a public school earned about $2,000 less in salary than a starting lawyer at a prominent law firm. These days the lawyer takes home, including bonus, $115,000 more than the teacher, the McKinsey study found.


He also cites some studies that argue that good teachers produce good student results, although the general trend is that instructional inputs are related only dimly as second or third order effects to student academic performance and life success.

But, part of the reason that demonstrated results from instructional inputs are so equivocal, is because almost everybody is using the same basic model: 180 days a year of school for days of less than eight hours, modest amounts of homework, teachers mostly drawn from conventional teacher certification programs in public schools with reasonably similar class sizes, and low career long pay prospects for someone with a college degree. Private religious schools don't pay teachers very well either. There are differences, but those differences are mostly modest.

A few elite secular private schools pay teachers very well, but given the elite character of the incoming student body which is overwhelmingly either upper middle class, rich, or extremely academically talented, it is hard to tell if good teachers or good students are driving the results. Also, these schools typically have not only highly paid teachers but very modest class sizes, which make their model unattainable without immensely larger K-12 budgets.

The public will simply does not exist to pay for a system of public schools where the average student-teacher ratio is 8:1 and the average teacher makes $90,000 a year. The more typical story is like this one from today:

The fifth-largest school district in Colorado this afternoon expects to announce $30 million in cuts — including jobs and student activities, as well as increased class sizes — to meet its 2011-12 budget.

Today at 3:30 p.m., Adams 12 Five Star Superintendent Chris Gdowski will announce the proposed cuts, including an anticipated 185 job cuts. Last year, the district cut 188 full-time positions. Over the past five years, the district has cut $38 million from its budget.

For the 2010-11 school year, the operating budget was $268.8 million, about 88 percent of which was devoted to employee salaries and benefits.


Even if we did, we wouldn't get the results you see at Groton in classes full of kids who are two grade levels behind by the third grade, may not get a square meal on the weekends, have an older sibling in a gang, live in a single parent home with a parent who dropped out of high school, and are intermittently homeless. Government stinginess towards children in the United States is not limited to education spending.

On the other hand, the Detroit model: dramatically increasing class sizes while paying their teachers 30% below the going rate for teachers in the state, despite the fact that they are teaching students with the greatest educational needs, and simultaneously increasing commuting time for students on their way to and from school, is pretty surely not a good way to improve educational quality either. To some extent, you get what you pay for.

Nobody has ever tried in recent history for a sustained period to set up a K-12 education system in which entry level teachers teaching ordinary average neighborhood kids are paid as much as entry large law firm lawyers in the same area, and can expect to receive that kind of compensation for their entire careers, but teach very large classes by American standards with very little staff assistance.

Of course, it isn't at all easy to break out the impacts of particular parts of the Japanese or American models. The Japanese spend more days in school. A larger share of American kids are college bound than in Japan. Cram school for college entrance exams to supplement ordinary high school education are the norm. Japanese educators are less reluctant to use drill and kill instruction methods, and more liberal in putting social pressure on kids to perform though public disclosure of academic performance. Extra-curricular activity participation is more of an expectation in Japan. Many American kids have more educational choices that comparable Japanese kids. American higher education, particularly at the graduate level, has a better reputation for rigor than Japanese higher education. American teachers have a much larger proportion of very low income students and far more students who are non-native speakers of the language of instruction. Japan starts foreign language instruction sooner and has very little sex education.

Also, somewhat surprisingly, given the tendency of K-12 education systems to pay teachers of older students more than teachers of younger students, the evidence on educational performance and education seems to indicate that the importance of educational quality is greatest when kids are youngest and lower when kids are older. Kids who need remedial work upon entering college were mostly behind already by the time they were in the 6th grade.

If we are to set compensation based on potential lifetime educational performance impact (and no school system in the world does this, to the best of my knowledge, although many do mandate smaller class sizes for younger students), we should be paying the highest salaries to pre-school and kindergarten teachers, the next highest salaries to elementary school teachers, the next highest salaries to middle school teachers, and the lowest salaries to high school teachers. Instead, the pay of teachers seems more closely related to the proportion of the teachers who are men, which is greatest for the older ages. As usual in the American labor market, the supply of similarly skilled labor is more important in setting compensation levels than the value of what workers produce.

But, moving in the direction of the Japanese business model for teacher pay, and paying early education teachers much more than we do, is certainly something worth considering.

No comments: