Thursday, September 14, 2006

Why class size DOESN'T matter...

Over at California LiveWire, ms-teacher, makes the case for why small class sizes are so important in our schools. Her arguments are compelling: better rapport with students, more individualized instruction, quicker identification of students who need help, fewer problems with classroom management. As she points out, anyone with a modicum of common sense can see the benefits of reducing class sizes. It really does make a difference!

Except, it doesn't.

The thing that is so often overlooked in this discussion is that there is a shortage of good teachers. In fact, there is such a shortage of teachers of any quality that, for decades, districts across the country have been forced into hiring inexperienced, uncredentialed (so called "emergency credentialed") teachers just to fill out their staffs. Obviously, teacher quality suffers as a result of this; it's tough to be choosy when there isn't a lot to choose from.

To be clear, it's not that we don't have good teachers. There are plenty of them--just not enough to go around. And therein lies the rub.

Smaller classes translates to more classes, and that translates to more teachers--a lot more. Let's say our average class size is 30, and we'd like to reduce it to 20 (very close to the example ms-teacher gives). This one-third decrease in class size means we'd need to hire 50% more teachers--50%, when we're already scraping the bottom of the barrel just to get warm bodies in the classrooms.

Where are these teachers going to come from, and how much more will we need to relax our hiring standards so that we can still pretend they are "qualified?" Would you rather have your kid in a class of 30 with a good teacher, or a class of 20 with a poor one?

And while you're pondering that, let's look at some real world evidence. Class size reduction isn't a new idea. We've been trying it for decades with poor results. As Abigail and Stephan Thernstrom explain in their book No Excuses: Closing the Racial Gap in Learning,

Over the past thirty years, the number of teachers employed in our public schools has increased by 48 percent more than the number of students they have to teach...

In fact, pupil-teacher ratios have continued to fall, dropping from 17.7 in 1992 to 15.1 by the 2000-2001 academic year.

What do they conclude?

Decades of research... have failed to establish that smaller classes have any measurable impact on student achievement. If this seems counterintuitive, it is not. If districts have to hire a great many teachers in order to reduce average class size, they are forced to be less selective in picking those teachers, with a decline in quality as a consequence. The key question, therefore, is teacher quality, not class size.

In California, where ms-teacher teaches, average class size was reduced from 29 to 19 between the 1995 and 1999 school years. The results were both predictable and disappointing, causing the California Department of Education to acknowledge that the program "may have increased underlying inequities in the state's education system. The Thernstroms explain:

This costly effort... may have done more harm than good to some categories of students... [Many of those hired] were inexperienced and underqualfied, and they tended to end up in schools serving low-income and non-Asian minority children.

The end result was a law suit, filed by the ACLU, specifically citing the "underprepared and inexperienced" teachers hired to bring about the class size reductions.

Class size reduction is one of those reforms that is easy to support. It just seems to make so much sense. But experience has taught us that these reforms yield poor results, while being very costly to implement (teacher salaries/benefits and classroom construction/maintenance are the top two items in a school system's budget). Once the scarcity of well-qualified teachers is taken into account, the equation changes, and it's easy to see why such reforms are ill-advised. Hopefully, we can put this down as a lesson well learned.

No comments: