By Interestana AI Editorial — AI-drafted, human-overseen. How we report
DOGE Program Cost US Government $6.7 Billion in Paid Leave

The U.S. government incurred a $6.7 billion expense through a deferred resignation program, part of an initiative to reduce the federal workforce, according to a new report from the Government Accountability Office (GAO). This program, spearheaded by the Department of Government Efficiency (DOGE), contributed significantly to a 435% surge in federal agencies' use of administrative leave between 2023 and 2025, totaling $9.5 billion in salaries paid to employees not actively working. Civilian salaries and benefits, which constitute approximately 5.5% of the federal budget, were among the initial targets for cost-cutting measures. In February 2025, DOGE communicated a "Fork in the Road" offer to two million federal workers, providing an option to voluntarily resign while continuing to receive full pay and benefits until the end of September 2025. This voluntary separation was accepted by nearly 140,000 employees, based on federal data.
Office of Personnel Management (OPM) Director Scott Kupor contested the GAO's findings, asserting in a recent Substack post that these personnel reductions are projected to yield long-term savings for taxpayers. Kupor stated to Fortune that the GAO report overlooks the distinction between a one-time expenditure of $9.5 billion for a reduction of 270,000 federal employees and the subsequent annual savings of $40 billion. He characterized this as a "400% return on investment" that benefits taxpayers substantially. The GAO report itself acknowledged potential inaccuracies in the data due to variations in how different agencies report administrative leave, a challenge that has persisted across multiple administrations. A prior GAO report from 2014 identified discrepancies in leave granting practices between the Department of Defense and the now-dissolved U.S. Agency for International Development, prompting recommendations for OPM to establish standardized guidance for leave recording. Elon Musk was involved in assembling this special advisory group in the early stages of its development. The deferred resignation program, implemented by DOGE, aimed to achieve significant workforce reductions. The program offered incentives for employees to leave voluntarily, thereby reducing long-term payroll obligations. The GAO's analysis focused on payroll data to quantify the financial impact of this policy. The total cost of administrative leave, including the DOGE program's contribution, highlights a substantial financial outlay for the government during this period. The debate over the program's efficacy centers on the interpretation of short-term costs versus long-term savings. Kupor's defense emphasizes the projected fiscal benefits, while the GAO's report provides a detailed account of the immediate financial implications. The persistent issue of inconsistent leave reporting across federal agencies underscores the complexity of accurately assessing such programs. The GAO's recommendation for standardized guidance aims to improve data reliability for future analyses of federal workforce policies. The DOGE initiative represents a significant effort to reshape the federal workforce structure and its associated budgetary impact. The $6.7 billion figure specifically relates to the cost of paying employees who were part of the deferred resignation program and were not actively performing duties during their paid leave period. This program was a key component of the broader strategy to downsize the federal bureaucracy and streamline government operations. The substantial increase in administrative leave usage underscores the program's scale and its immediate financial consequences. The GAO's role as a government watchdog is crucial in providing objective assessments of such large-scale policy implementations and their fiscal ramifications.
Original source — read the full reporting at the publisher:
Read on FortuneGet the weekly AI digest
AI news + new model releases, weekly. Drafted by our agents, reviewed by humans.