Snow Days and Student Learning
Benjamin Hansen is a research associate at IMPAQ International, LLC.
Routinely around the country, children and parents gather around radios, news stations, and the internet waiting for one announcement: SNOW DAY!!!!
After all of the excitement (and perhaps some early morning pancakes), they hop back in bed to sleep for a few more hours before braving the winter wonderland for a day of igloos, sledding and hot cocoa.
Sorry, just having some flashbacks from growing up in Colorado. However, snow days provide a useful source of variation to researchers. The number of cancellations affecting any school varies year-to-year, and the variation in snow days should be uncorrelated with variation in any other other schooling inputs. Also, lost days aren’t made up until June, which can allow for many cancellations (sometimes dozens) prior to the administration of state assessments. While, ironically, Bart Simpson once spent one such snow day studying for an exam, for most students a snow day equates to one less day of school, allowing this interruption to provide a truly natural experiment for analyzing the effects that instructional days have on student performance.
In independent research, Dave Marcotte and I found that weather-induced cancellations can play a sizable role in student achievement, which we summarize in our new Education Next article, Time for School? On average, an additional instructional day increases test scores by .015 standard deviations. I also found further evidence of the impact of additional school days in Minnesota, a state which altered the day it took its state tests six times in six years.
With our estimates in mind, it is entirely reasonable that more instructional days could serve as a substitute for other education policy interventions. But longer school years for the most part have not been used. It might be worth asking why, as school year length in the United States largely has remained unaltered since the 1960’s.
The real answer is that schools have been increasing the number of instructional days recently, but only the number of days prior to the administration of assessments. Over the past several years, anecdotes of the school calendar creeping earlier into the summer have circulated. In consequence, several states (WI, FL, TX, and NC) have passed laws mandating a minimum start date in late August or early September. David Sims, in Strategic responses to school accountability measures: It’s all in the timing, finds some evidence that districts making such changes were able to raise their performance on assessments. In our Education Next article, Dave Marcotte and I summarize the evidence on instructional days and highlight the many interactions instructional days may have with accountability regimes.
Before we eradicate summer vacation, I would like to bring up a few things. The variation in instructional days we have studied affects the number of days prior to the exam. If we extended the school year, that would likely mean more days after the test. We still don’t know a whole lot about what happens in the classroom after state tests are given. Extending the school year after the exams would feed into the next year’s performance, or affect a teacher’s course planning for the year. If school heads into a semi-retired state after students take the accountability tests, then perhaps schools should increase the number of accountable days. In other words, they should consider making the test date later when increasing the length of the school year. Also, of course there are many costs to extending the school year, and one would need to consider the effects on teacher labor supply, energy costs, and scheduling with parents. However, now that the benefits have been initially established, it is worth investigating the costs to see if this policy change can be a key to improving schooling in current and future generations.