February 15, 2022
Outdated Policies Hinder FHWA’s Ability To Oversee Unobligated Emergency Relief Funds
What We Looked At
In August and September 2017, three costly hurricanes hit the United States and caused devastation to transportation infrastructure, particularly highways and bridges. The Department of Transportation (DOT) provided emergency aid for infrastructure repairs through the Federal Highway Administration’s (FHWA) emergency relief programs. Due to the size of DOT’s investment and the speed required for emergency relief, we initiated this audit. Specifically, we assessed FHWA’s controls over the use of its emergency relief funds.
What We Found
FHWA does not always deallocate funds that have not been obligated by the end of the fiscal year, as its policy requires. For example, of the $48,022,716 that FHWA allocated to Florida in 2018 and 2019 for hurricane damages, approximately $46 million remained unobligated in March 2020, but had not been deallocated. FHWA’s Division Offices also do not always deallocate unobligated quick release funds—funds that are allocated quickly for emergency repairs—as required. We noted that Florida DOT took over 10 months to obligate $13.4 million of $25 million in quick release funds, FHWA officials stated that the Agency is not statutorily required to deallocate these funds. FHWA also did not document required approvals for $48 million in quick release funds, and has not deobligated almost $2 million in unneeded emergency relief funds allocated to Texas DOT. FHWA is required to maintain adequate internal financial management controls. By not following its own policies, the Agency increases the risk that unused quick release funds and unobligated allocations will not be available for other States in need. Furthermore, the Highways ER Manual, last updated in May 2013, does not reflect current practices or align with other Agency policies. Still, despite these oversight weaknesses, we found a low incidence of improper payments in a sample of Highways disbursements.
FHWA partially concurred with our recommendation 1, did not concur with our recommendation 2, and concurred with our recommendations 3 through 6.