Big School, Small School: A look at the achievement gap in four states -- Rural Policy Matters, Volume 1, Number 2 (April 1999)

Many states, by encouraging larger schools in hope of lowering per pupil spending, may be unwittingly contributing to the widening gap between the academic achievement of the rich and the poor. That is one of the implications of the preliminary findings of a study on how school size interacts with poverty to affect student achievement in four states -- Georgia, Montana, Ohio, and Texas.

"Consistently, I'm finding that when I divide a state's schools into two groups, smaller half and larger half, the smaller half shows a much weaker relationship between poverty and achievement than the larger half," says independent researcher Craig Howley, who formerly directed the U.S. Department of Education's ERIC Clearinghouse for Rural and Small Schools at the Appalachia Educational Laboratory. Howley, and Marshall University education researcher Robert Bickel, are principal investigators for the research, which is supported by a grant from the Rural Challenge Policy Program.

The study replicates and expands previous work by Howley and others in West Virginia, Alaska, and California. The additional four states -- in four separate regions of the nation -- are testing whether the earlier findings prevail nationally.

Preliminary results seem to indicate that, overall, achievement does not differ much between large and small schools in these states. However, school size makes a difference when you look at communities of different levels of affluence and poverty.

In general, while big schools slightly benefit the achievement of students from wealthier communities, they actually harm the achievement of students from poorer communities. And the negative affects of big schools in poor communities are much greater than their positive effect in affluent communities. This is a somewhat surprising finding, since overall, school size doesn't show much of a relationship with student achievement. Only when you look, as in Howley's research, at the interaction between school size and poverty does the relationship of size to achievement become clear.

In practical terms, it seems that school size can make several years' difference in learning over a school career. The achievement gap between kids from affluent and kids from poor communities can be widened (as happens in some states) or narrowed (as happens in other states). Montana seems to be an example of a state where small schools help close the gap between affluent and poor areas.

And the results are the same whether you measure school size or district size, says Howley.

"Massification is a century-long trend to big schools and districts," says Howley, "and it hasn't served students, communities, or democracy well."

One crucial issue: Is the better performance of small schools in poor communities because the school is small or only because small schools typically have fewer kids in each classroom? Could we get the same result from big schools with smaller classrooms? Howley says that statistically controlling for the small classroom variable does not much change the results of the analysis. The influence of school and district size is different from -- and not weakened by -- the effect of class size.

This makes sense because factors like school climate and culture, community, and extracurricular participation rates are features of schools that powerfully, but indirectly, influence student performance. Small schools probably help attach kids who, if they attended large schools, would be regarded at best as unimportant to the success of the school, and as worst as a hindrance to increasing school test scores, an increasingly important factor in school funding in many states.

It might be that all kids are needed and valued in small schools, because small schools have to do all the things large schools do, but with fewer people. Everyone has to pitch in, and students are actively recruited for that purpose, not just by educators, but by friends, neighbors, and relatives.