Student Assessment results for vermont students were recently released (report here), and with several years worth of data, we can start to look at trends. They look good. As education Commisioner Richard Cates said in an inteview with Vermont Public Radio that he is generally pleased with the results.

VPR reports students are scoring higher in virtually every category at each grade level.

Despite the good news, there is a disturbing trend in the data.

Test scores for students coming from lower socioeconomic status families score consistently lower. Using eligiblity for free school lunch as an indicator of family income, scores for students from low income families are between 30 and 50 percent lower than students from familes with higher incomes.

These are the kids that aren’t going to college. The lack of family financial resources, as evidenced by eligiblilty for free lunches, and less ability to achieve in an academic context, as evidenced by these test scores, will, in most cases, mean that graduating high school will be as far as they will go.

Looking at this picture through a slightly different lens, we find that academic achiverment and income are related. Data from the 2004 March Supplement to the Current Population Survey put the median income for persons with a high school diploma at $25,935, compared to $41,800 for those with a Batchleor’s degree.

The social stratification is pretty clear.

It is difficult to repair the lack of skills and background necessary to perform well academically demonstrated by low test scores in early grades. As Cate puts it “By the time they get to high school, I don’t care how good the teachers are or how much money the school has, it is very difficult by the time they get to high school to reverse that.” And low achivement in high school, with the attendant lack of motivation to continue on, generally lead to lower paying jobs. And the test data are showing us the relation between family income and children’s test scores.

America is supposedly a classless society. But looking at these data, we see a much different picture.