Delta Cost Project: 20% of Higher Education Spending Results from Attrition

Many institutional research offices track attrition and engage in enrollment modeling. Their work informs conversations among campus decision makers about whether retention enhancement programs provide positive returns on investment. A recent report by the Delta Cost Project highlights the findings from a study of college spending and attrition. Of particular interest to IR professionals is the methodology offered to assign “institutional production costs to attrition” and related benchmarks.

The Delta Cost study, entitled The Costs of Student Attrition
utilized a definition of attrition as “departure from all forms of higher education prior to completion of a degree or other credential” and found that “roughly 20 percent of education and related spending in higher education results from attrition. Reducing attrition costs is both educationally effective (more students obtain degrees) and cost effective (due to efficiency gains resulting from reduced attrition)." Further findings include:
  • Attrition costs are much lower than the overall attrition rate.
  • Most attrition is not caused by academic failure.
  • Attrition costs should not be measured at the institutional level, because most students graduate from institutions other than those in which they first enrolled.
  • Attrition costs are highest for students who leave after several years.
  • Attrition levels in higher education are similar to attrition levels in other industries.

For more information, visit the Delta Cost Project website.

Join the conversation: How is your campus exploring issues related to cost and attrition?



To add a comment, Sign In
Total Comments: 1
Harry posted on 10/18/2012 2:50 PM
This is a great study. To be able to predict the rate of attrition before it happens is better. It helps institutions to lower the rate and increase their operational efficiency. There should be some form of statistical analyses which will give the early alert or warning transmitted to the decision makers of which individual student is in a high risk of dropping out, before the student actually stops going to school. The statistical model can be different for dropping out student group and transferred student group. This study can be sharpened by differentiating these two groups. To learn more about the potential statistical analyses which can be applied to address the attrition issue, see my presentation at and find the abstract.
Harjanto Harry Djunaidi

*note from editor, this comment was edited to remove a URL that linked to a non-working website.