Education: Class Size- the Unions Have a Partial Point

Teachers and parents love them and unions want them.  Small class size is, at times, all the political rage.  And there are certainly advantages to smaller class sizes.  But although we can generally get behind the idea, there is one major obstacle standing in the way- financial reality.  For example, the NEA has proposed that class sizes not exceed 15 students.  Sometimes it appears as if they pull these numbers out of their butts.  There is no study out there declaring that 15 students in the optimal amount in any classroom.   In fact, today there are 18 states with average class sizes appreciably below that optimum number suggested by the NEA.

The conversation comes down to whether this country is willing to spend the necessary money for  smaller class sizes.  It is a classic case of cost/benefit analysis.  Assuming in those states wishing to lower their teacher to student ratio to 15:1 that we hire the requisite number of teachers and assuming we pay them the average starting salary (not overall average salary- thus, a lower salary), it would cost states over $8.5 billion a year.  California alone would have to spend more than half that amount.  Can anyone really expect a cash-strapped state like California to expend an additional $4 billion on hiring teachers?

The premier study in this area was the STAR study out of Tennessee that showed a one time improvement in test scores when class sizes were reduced.  A 1996 study out of California had proponents all excited when that state spent over $1 billion to reduce class sizes.  When test scores improved, they believed there was hope.  A RAND Corporation analysis indicated that the increase in test scores was due to other reforms initiated at the time as test scores at schools with larger classes also increased.

There is also the problem and associated costs of infrastructure investment.  Besides spending over $8 billion annually on hiring and retaining teachers, classrooms would have to be built, books bought, computers and transportation provided.  These items will clearly push those costs way beyond the $8 billion.

Where does the statistical evidence take us?  There are nine states plus DC where the average class size is considerably below the national average of 15.2 students to 1 teacher.  In only one area- NAEP performance at the 4th grade (elementary) level- does class size seem to have any effect on educational outcomes.  When it gets to the 8th grade level, this group actually has the lowest performance and they have the worst average SAT scores.  This idea of smaller class sizes at the lower grade levels is further supported by looking at the performance of the states with the largest class sizes.  At this level of education, they perform the worst on the 4th grade NAEP.  Thus, smaller class sizes at a grade levels below the fourth would likely improve academic performance, but as we go up the grade levels, class size makes less of a difference in academic performance.

And this makes intuitive sense.  It is in the early grades that good study habits and classroom discipline coupled with an imparting of what is expected of a student is the most important.  These grade levels lay the groundwork for future educational success.  The demographics differ from state to state and the costs of decreasing class sizes in states at the lower grade levels would obviously vary.  Suffice to say, the task would be considerably less costly than $8 billion per year and could even be achieved by reshuffling the chairs on the deck by moving teachers from upper grades to lower grades.

Perhaps states should meet the unions half way here through a compromise. By lowering class sizes in the lower grades (those 5th grade and lower), it would likely improve academic performance.  Increases in average class sizes in grades 6-8 would have no discernible deleterious changes in academic performance.  The statistics show that on the 8th grade NAEP, the large class size states exceed the averages of all the other categories- including the small class size states.

And if we look at individual performances, we can see that class size exclusively does not achieve the task of improving academic performance.  For example, New York and DC rank 2nd and 3rd lowest in terms of class size, but neither are considered way above average in terms of academic metrics.  Conversely, large class size states like Washington, Oregon, and Colorado DO perform considerably above average.  The likely culprit is demographics- namely, the low class size states that do perform below average have a higher population of minority students while the high class size states that perform above average have a lower population of minority students.  Making class size smaller in targeted districts with high minority populations may make some sense up to a certain point in order to overcome the disadvantages with which they enter the educational process.  But, it makes sense- as DC proves- only up to a certain point.  As DC proves, even if the class size was less than its current figure, it would  have absolutely no effect on academic performance.