This is the eighth in a series of newsletter abstracts of audits to bring you up to date in what is happening in the Federal audit arena. Please note there is a considerable lag between when an audit occurred and when the audit report is published.
Duke University Audit – January 21, 2009
Audit Agency: DHHS-OIG-Region IV
Title: Review of Administrative and clerical Costs at Duke University for the Period of October 1, 2002 through September 30, 2004
Objective: To determine whether the university claimed reimbursement for administrative and clerical expenses as direct costs to grants, contracts, and other agreements with Department of Health and Human Service (HHS) components in accordance with applicable Federal Regulations.
Awards: 2,566 grants, contracts, and other agreements with HHS components from October 1, 2002 through September 30, 2004.
Findings: The auditors statistically selected 114 charges for administrative and clerical salaries and 120 charges for administrative costs from the overall population. After the audit and response by duke University, 35 items of the former totaling $23,995 and 46 items of the latter totaling $14,899 were disallowed. This $38,894 was then extrapolated to the population as a whole, resulting in a finding requiring Duke to refund $1,661,011. Based on the 81 items disallowed, the auditors concluded that Duke needed to revise its policies to comply with the requirements of OMB circular A-21 to ensure consistent treatment of administrative and clerical costs.
Significance of Audit: This audit is probably the first where clerical and administrative costs were statistically tested and the result extrapolated to the population as a whole. In the past, audits were of specific awards and were based on a judgment sample that limited any disallowed costs to those specific costs disallowed. In these cases, Universities often charged the disallowed costs back to the department responsible for the audited award. Statistical audits audit the university not a specific award, PI or department, and, as in this case, the findings can be much more significant.
Lessons for Emory: Since the items for testing were selected statistically, there is no place to hide poor compliance. The auditors used a 90% confidence interval in their work. that means that they were 90% certain that the results of any other similar statistical audit would fall between $1,661,011 (the lower limit) and $3,650,986 (the upper limit). Most likely the result of a similar audit would fall closer to the midpoint of $2,655,999. Now that the door is open on statistical audits, the Feds could use them on effort reporting, travel, or any other category of expense or compliance item.