Article | Connecticut’s schools are worse than they look

September 9, 2004

By Jay P. Greene and Greg Forster

Connecticut's students rank 19th in the nation for academic achievement. At first glance that might seem acceptable, if not great. But a new study by the Manhattan Institute finds that Connecticut ranks that high only because its students face fewer life challenges than students in other states. Remove the effects of student characteristics, and Connecticut schools rank a much less comforting 29th in the nation.

The best comparison we have for states' academic outcomes is eighth-grade scores on the National Assessment of Educational Progress, ``the nation's report card,'' a test administered by the U.S. Department of Education. Scores for eighth grade are preferable because they're the oldest grade for which state-by-state comparisons are available.

Connecticut schools look passably good by this yardstick. In reading, 77 percent of Connecticut students reach the ``basic'' level of achievement, the lowest level recognized by the test. In math, 73 percent of students reach the basic level. These scores rank Connecticut 19th in the nation -- not stellar, but in the top half.

However, there's more to the story than meets the eye. Some students come to school more ready to learn than others. Connecticut's students are better off than the national average in most areas affecting students' ability to learn -- they're healthier and wealthier, they're more likely to come from intact families, they're more likely to attend preschool and so on. We should take this reality into account when we evaluate Connecticut's schools in comparison with those in the rest of the nation.

A new study by the Manhattan Institute allows us to do just that. We systematically measured the levels of 16 social factors that researchers agree affect student outcomes. By comparing the actual academic performance of students in each state to the level of performance we would statistically expect those students to achieve given their characteristics, we were able to measure which states' schools were truly performing well and which just looked good because of their students.

Connecticut's schools did only just about as well as their student demographics would predict. Levels of the student advantages we measured were 10.4 percent above the national average in Connecticut. Taking these advantageous circumstances into account, we find that Connecticut's student achievement is 101 percent of what we would expect it to be. This ranks Connecticut 29th among the states in academic outcomes adjusted for student characteristics -- well into the bottom half of the nation.

Some other states with highly advantaged students do a much better job of teaching them. Massachusetts' students have advantages 11 percent higher than the national average but perform at 106 percent of expectations, ranking them 15th in the nation for academic performance adjusted for student characteristics. And some states with very disadvantaged students do even better -- Texas' students have characteristics 19.8 percent worse than average but perform at 110 percent of expectations, ranking them fourth in the nation.

The usual complaints about low funding can't explain Connecticut's performance. Connecticut spends $12,166 per student, more than any other state except New York, to produce its lackluster educational results. So when we adjust academic outcomes for spending levels as well as for student disadvantages, Connecticut falls even further behind. Its students' academic performance is only 73 percent of what we would expect it to be given its student population and spending level, ranking it 49th in the nation for education efficiency.

What could be done to improve Connecticut schools' performance? We found that states with stronger accountability testing or more school choice had significantly better academic outcomes after student characteristics were taken into account. This adds to a solid body of academic research finding that these reforms improve the academic achievement of disadvantaged students by giving schools a powerful incentive to serve them better.

Connecticut's schools shouldn't be allowed to get away with providing mediocre educational services simply because good student demographics keep their test scores from falling too low. Given what it spends on education, Connecticut has a right to expect its schools to do a good job with whatever students it sends them. When the students are particularly advantaged, we should raise our expectations accordingly.