Preschool is NOT a Public Safety Issue

The Linn County Sheriff  and Cedar Rapids police chief have jumped into the public preschool debate.  From

The Linn County Sheriff and Cedar Rapids police chief Wednesday pushed to get more kids enrolled. They point to a study showing high-quality preschool education lowers the crime rate.

“When you work in law enforcement, it’s easy to see that helping low-income, at-risk, 3 and 4 year-olds get a jump start in their education can make a critical difference in reducing the likelihood they will become criminals,” Cedar Rapids Police Chief Greg Graham said.

[Note to Mark Geary: get the title of the study when government officials claim a study supports their request for more public spending.]

It appears that they are relying on the Perry Preschool Study.  The study assigned 58 low-income African-American children into a preschool program group and another 65 low-income African-American children into a no preschool program group.

The study draws these conclusions about a 2-year preschool education program for 3- and 4-year-olds living in low-income families. Teachers had bachelor’s degrees and certification in education, and each served 5–6 children. They used the High/Scope educational model in daily 2 [and] 1/2-hour classes and visited families weekly [for 1 and 1/2 hours].

The study that was available online is the 1984 study reflecting the children at age 19 rather than 40.  The 1984 study does not appear to have been peer-reviewed.  It has a small sample size (compare this total sample size of 121 to the Head Start study that included nearly 5000 children).  There are problems with the presentation of the data regarding crime statistics.  That is, they report being authorized to do criminal records checks on only 20 of the 58 adults who had participated in preschool (plus 4 more arrest records found by a blind search), and on 35 of the 63 adults who remained in the study and had not had preschool (plus five more arrest records found by a blind search).  The arrests are reported as percentages of the entire groups but is it reasonable to assume they have enough data to show the other group members had no criminal records?  Even if you accept their numbers they still report 36% of the preschool program participants as having been arrested five or more times by age 40 compared to 55% without preschool.  Finally, given the small sample size, and the researchers failure to control for parental/family attitudes about education and criminal behavior, is it likely that preschool attendance is the most significant factor in determining later criminal behavior?  It just doesn’t seem like enough evidence to base millions of dollars of spending on, especially since we already have evidence that academic gains from preschool attendance fade away ( see this post for more).