Coming Out of the Dark on Dropouts

A feeling of relief about recently reported Haverhill dropout numbers needs to be tempered with continuing concerns about how those numbers are reported. Recently released data on dropouts may reveal less about students leaving school than they do about shortcomings of the Haverhill school administration in reporting and using school performance data. Following the recent misreporting of data on dropouts, Haverhill Public Schools will need to rebuild trust with the School Committee and the public and work toward building its capability to use accountability data to drive improvements in school performance.

A recent (February 26, 2018) press release from the Department of Elementary and Secondary Education (DESE) highlights a decline in dropout rates over the past five years in five urban school districts, including Haverhill. However, with the Haverhill Superintendent himself invalidating past dropout figures in his report last October, it is hard to know what to make of the recent data. The chart below shows Haverhill’s dropout rate as reported by DESE based on district-submitted data. Starting in 2009 Haverhill’s reported dropout rates ran well above the trend of other cities; only in 2017 did reported rates return to more typical levels by means of one dramatic drop (for the most recent reported school year 2016-17).


It is important to recognize that this chart may not accurately represent the students who actually dropped out. Rather it represents what the district reported to DESE. The Superintendent has stated that data on dropouts and transfers were incorrectly reported to DESE for some unspecified number of years. He publicly provided revised data from one year (2015-16) indicating that more than half the students that were reported as dropouts (59 students for that school year) should have been classified as transfers.  (This was discussed in a Benchmark Blog post of October 29, 2017.) We have no data from earlier years to substantiate just when a meaningful decrease in dropouts may have occurred. This information is lost to history. The decline in reported dropouts for the 2016-17 school year is based on data Haverhill submitted in the fall of 2017 after revising its procedures for identifying transfers. The sudden decline in the reported dropout rate is thus a result of the change in reporting procedures, made at least in part to address long-standing errors in Haverhill reporting, not of a sudden change in the number of students dropping out.

For eight years Haverhill was an outlier with its high dropout numbers. If these figures were not correct, it seems that district leadership was flying blind for years on the dropout situation. Now we are asked to trust that the dropout rate really did decline. We are told that more than 50 students annually were reported as dropouts who had actually transferred to other schools, mostly out of state and outside the DESE student tracking system. The district is now reporting more students as transfers and fewer as dropouts. If previously too few were identified as transfers, one might reasonably ask whether too many are now being classified as transfers – that is, labeled transfers without proper documentation that they have reenrolled elsewhere.

DESE expects school districts to identify transfers when they receive a transcript request from another school or communication from a parent. Non-returning students not identified as transfers are to be reported to DESE as dropouts. While DESE is able to identify in-state transfers to public schools through its tracking system, it must rely on the district to accurately report out-of-state transfers and to count only those for whom reenrollment elsewhere can be appropriately documented. See DESE Dropout Reporting Guidelines.

A remaining problem for Haverhill is a lack of trust in the capabilities of the current school administration to accurately report and effectively use school performance data. Whether dropout reporting problems arose due to inattention, lack of effort, or just poor communications and lax oversight of reporting, Haverhill citizens do not now know what to believe. The dropout rate, a useful performance indicator in many cities, has been reduced to an unanswered question in Haverhill. We do not know to what extent good work in the schools went unrecognized and what opportunities were lost to better understand student needs and craft initiatives to address them. That situation needs to be avoided going forward. In today’s performance-oriented world of data-driven management, urban schools will fall behind if they are not capable of effectively using the best evidence to guide improvement.

Trust and transparency are important if data are to be used to improve Haverhill schools. A new superintendent will need to reestablish trust that Haverhill is accurately reporting performance statistics. That will require careful review of procedures, close supervision, and direct administrative oversight of data submission to ensure that Haverhill is neither over-reporting nor under-reporting dropouts. A new superintendent deserves to start with a verified baseline of key performance measures. He or she will need to ensure that DESE guidelines are being followed with clear and transparent procedures for distinguishing between transfers and dropouts. The School Committee should expect a candid report on procedures, past and present, to ensure they are now in good order to support accurate reporting.

There is much work ahead if we are to achieve the School Committee’s goal to “Make the district more data driven and performance oriented.” Accurate and openly reported data practices would be a good start. Using the data effectively to guide improvements would be a next step.