FDA’s Inspection Decline Fell on High-Risk Food Facilities; A Scheduling Fix Addresses Only Part of the Gap

Two inspection categories moved in opposite directions

FDA’s domestic food-inspection decline was not evenly spread across the facilities it oversees. Our comparison of the risk-category figures in a federal inspector general’s report shows that the annual number of high-risk facilities inspected fell by 2,616 between the 2017 baseline and the average for 2022 and 2023—a 37.7% decline. Inspections of facilities outside the high-risk category increased by 260, or 4.6%, over that same comparison.

The difference changes the accountability question. Restoring the total number of inspections to a pre-pandemic level would not by itself show that the highest-risk category had recovered. Conversely, the small increase in non-high-risk inspections did not mean that category received enough coverage to meet its longer inspection cycle. The records describe two different problems: lost high-risk coverage and an already inadequate pace in the much larger remaining inventory.

The figures are historical, from the Health and Human Services inspector general’s June 2025 report, not an estimate of inspections completed during 2026. We compared them with FDA’s 2026 plan to shift more routine inspection work to state partners and with the current status of oversight recommendations. The comparison identifies the outcomes that would demonstrate progress; it does not claim the new plan has already failed.

A return to the old total would still leave a capacity problem

Using the report’s rounded category averages, the combined decline was 2,356 facilities a year. The high-risk loss exceeded that net decline because the increase in non-high-risk inspections partly offset it. This is an arithmetic decomposition of the published counts, not proof that FDA intentionally diverted staff away from high-risk facilities. Changes in the inventory, staffing, inspection complexity and pandemic disruption are not separately measured by these figures.

The Food Safety Modernization Act requires high-risk domestic facilities to be inspected at least once every three years and non-high-risk facilities at least once every five years. The inspector general estimated that FDA would need to inspect approximately 7,000 high-risk and 11,000 non-high-risk facilities each year to cover the inventory on those cycles. Its post-pandemic averages of 4,326 and 5,913 were well short of both. The underlying estimates use average 2017–23 inventories and are not a list of precisely that many inspections due in any particular year.

Even the 2017 non-high-risk count of 5,653 was far below the roughly 11,000 annual pace estimated in the report. That is why a promise to restore pre-pandemic volume is insufficient for this category. For high-risk facilities, the earlier count of 6,942 was much closer to the report’s approximate annual requirement. The two baselines therefore imply different recovery tasks.

What FDA’s new division of work is supposed to change

FDA’s 2026 Human Foods Program priorities propose a program called BRIDGE, under which state partners would conduct more routine food-safety-system inspections while FDA concentrates resources on international, high-risk, complex and targeted work. The agency describes a 2026 proof-of-process stage followed by full implementation over the next four years. That is a staged plan, not a statement that the entire transfer of work is complete.

The historical counts already include inspections that state, local, tribal and territorial governments performed on FDA’s behalf. Counting more state activity as newly added federal coverage would therefore require care: the meaningful test is an increase in completed coverage across the combined system, not simply a change in which government supplies the inspector. Separate high-risk and non-high-risk results would show whether the proposed division of labor actually addresses both deficits.

FDA’s May 1, 2025 response to the inspector general explains the constraints. It agreed with all four recommendations, described nearly 90 food-investigator vacancies at that time and cited hiring and retention difficulties. It also described workforce measures and other oversight tools, including remote assessments and partnerships. Those are FDA’s documented explanations from 2025, not verified current vacancy counts or interviews conducted for this report.

One documented fix does not close the entire inspection gap

There is verified progress in a related oversight record. GAO’s January 2025 food-inspection report now marks its recommendation to reduce unsuccessful domestic inspection attempts as implemented. The June 2026 update says FDA created procedures to check that facilities are operating before an unannounced visit and improved the inventory used to schedule inspections. Any account claiming that FDA has done nothing about wasted inspection attempts would omit that finding.

The HHS inspector general’s four recommendations remained listed as open and unimplemented when checked for this report. They cover inspection frequency, identifying inactive facilities, examining the decline in significant violations, and timely follow-up. The apparent difference is not proof that the agencies contradict one another. They are separate recommendations, with different scope and verification. GAO’s closure supports a specific scheduling improvement; it does not establish that FDA now meets the three- and five-year inspection cycles.

GAO’s June 2026 update also leaves its broader performance-management recommendation open. FDA told GAO it could not provide a detailed update and anticipated the work as part of a larger agency-wide effort. That remaining issue is central to judging BRIDGE: the public needs results that show what the changed system accomplished, including inspection coverage by risk and whether serious problems received prompt follow-up.

The inspector general found that 91% of the 769 significant-violation inspections in its follow-up analysis did not receive another inspection within six months, the review’s timeliness benchmark. That is not a finding that 91% received no follow-up at all; the report separately put the no-follow-up share at 8%, with rounded categories. Nor does an inspection lapse establish that a facility’s products caused illness. It identifies the gap in physical verification that the recommendations are meant to close.