The Value of Government Services: Evidence From Flight Delays During the 2025 Shutdown
Key Takeaways
- The value of government services is difficult to determine, but when they are disrupted, the value they provide can be inferred from the resulting losses.
- The flight delays occurring during the government shutdown provides an opportunity to place a value on government services in aviation.
- An analysis shows that an estimated 1 million extra work-hours of delay were incurred during the 43 days of shutdown beginning Oct. 1, costing $66 million.
Government services often have more value in preventing disruptions than producing directly observable outputs. As a result, their contributions are difficult to measure in normal times. When these services are disrupted, however, their value can be inferred from the resulting losses. In this article, I quantify the economic value of federal government services in aviation by examining flight delays during the 2025 government shutdown, using a difference-in-differences research design and large-scale administrative data.
Analyzing the Shutdown's Impact on Air Travel
This analysis integrates 76 million flight operations since 2022 across 82 major U.S. airports, drawn from the Bureau of Transportation Statistics (BTS).1 In 2025 prior the October shutdown, 22.6 percent of U.S. flights were delayed. Roughly one-third of these delays were attributed to airline operations, one-third to late arrival of aircraft and the remaining one-third to Federal Aviation Administration (FAA) Traffic Flow Management, which responds to weather conditions, traffic volume, runway constraints, equipment outages and related factors. This analysis focuses on delays initiated by the FAA, as these were directly affected by the federal government shutdown.
Flight-level delay records are merged with daily airport-level weather data from the National Oceanic and Atmospheric Administration (NOAA), allowing the analysis to net out the effects of snow, wind speed, thunderstorms, temperature and other meteorological conditions. (See the appendix for a full list of weather conditions examined in the analysis.) To value every delay minute, each airport is linked to metropolitan-area wage data from the Bureau of Labor Statistics.
The last point reflects a key idea that air travelers are not representative of the average household. Air passengers (particularly those flying through large hub airports) tend to have higher incomes. To measure the value of delay time, the Department of Transportation's (DOT's) 2016 guidance suggested using hourly wage as the unit of the per-hour income loss due to the time wasted by the delay. However, the hourly wage in the guidance used the national median wages, which are also nearly a decade old (leading to $47 per hour of delay).2
Rather than applying a uniform national value of time, I measure the value of passengers' time using current wages at the Core-Based Statistical Area (CBSA) level, adjusted to reflect the income distribution of air travelers across metropolitan areas. The adjustment is quantitatively important for airports serving high-income metropolitan areas and business travelers.
Identification relies on a difference-in-differences design, comparing changes in flight delays at the same airports during the shutdown period to changes outside the shutdown window. The specification controls for detailed weather conditions and airport-specific seasonality. By exploiting variation across airports and over time, the design isolates excess delays attributable to the temporary disruption of federal services — such as air traffic control and system coordination — rather than to adverse weather or month-to-month seasonality in air traffic. To remain conservative, I report only shutdown effects that are statistically significant at the 1 percent level.
The Costs of the Shutdown
Applying this conservative framework, I find six airports with statistically significant shutdown effects, which results in an estimate of 1 million extra work-hours of delay at a total cost of $66 million incurred during the 43 days of shutdown beginning Oct. 1. The costs are highly concentrated in high-income airports:
- Boston's Logan International Airport alone accounts for $21 million, reflecting both large increases in delays (more than five times) and a high value of passenger time ($67 per hour).
- New York's three major airports — John F. Kennedy International Airport, Newark Liberty International Airport and LaGuardia Airport — together account for $34 million, more than half of the national total.
- San Francisco International Airport incurs approximately $7 million, largely due to passengers earning very high income and thus having a high value of time ($76 per hour).
- Ronald Reagan Washington National Airport bears about $3 million.
These losses are associated with large increases in operational disruptions. Relative to baseline periods, the total delay time increased by 527 percent in Boston, 467 percent in New York, 209 percent in San Francisco and 546 percent in Washington, D.C. Full results for these airports are shown in Table 1.
| Select Airports | Value of Passenger Time (per Hour) | Number of Delays | Estimated Number of Extra Delay Flights During Shutdown | Estimated Total Work-Hours of Extra Delay During Shutdown (in Millions) | Percentage Increase From Normal | Estimated Cost of Extra Delay During Shutdown (in Millions) |
|---|---|---|---|---|---|---|
| BOS | $66.70 | 4203 | 3448.13 | 0.3151 | 527.24% | $21.02 |
| DCA | $68.86 | 1443 | 1027.67 | 0.0461 | 546.63% | $3.18 |
| EWR | $62.81 | 3496 | 2854.37 | 0.2308 | 437.79% | $14.49 |
| JFK | $62.81 | 2085 | 1630.33 | 0.1603 | 409.63% | $10.07 |
| LGA | $62.81 | 4395 | 3210.37 | 0.1547 | 535.96% | $9.72 |
| SFO | $76.21 | 2097 | 1375.04 | 0.0951 | 209.93% | $7.25 |
| Total | 17719 | 13545.90 | 1.0022 | $65.73 | ||
| all significant at 0.1% | all significant at 1% | all significant at 1% |
The difference-in-differences estimates show that these increases persist after conditioning on detailed NOAA weather controls and airport-specific seasonality. This indicates that the increases cannot be explained by unusually severe meteorological conditions or recurring delay due to air traffic seasonality and personnel shortages prior to the shutdown.
Conclusion
Taken together, the evidence illustrates the substantial but often unmeasured value of government services in important economic activities such as aviation. The estimated delay costs — adjusted for the higher incomes of air travelers — represent a conservative lower bound on the true value of these services, excluding additional losses from canceled flights and broader productivity spillovers.
Russell Wong is a senior economist in the Research Department. Comments from Alex Wolman are much appreciated.
The system-impact delay data are maintained for 82 ASPM airports in the Operations Network. See the appendix for a full list of airports.
The DOT's guidance finds that 59.6 percent of air travel is for personal reasons and 40.4 percent is for business reasons. To measure the value of time, the DOT applies a factor of 1.9 to the median income in 2016 for personal air travelers and a factor of 2.5 for business air travelers. Instead of assuming uniform value, I update the value of time by using CBSA-level median income from the BLS's 2023 Occupational Employment and Wage Statistics.
To cite this Economic Brief, please use the following format: Wong, Russell. (December 2025) "The Value of Government Services: Evidence from Flight Delays During the 2025 Shutdown." Federal Reserve Bank of Richmond Economic Brief, No. 25-45.
This article may be photocopied or reprinted in its entirety. Please credit the author, source, and the Federal Reserve Bank of Richmond and include the italicized statement below.
Views expressed in this article are those of the author and not necessarily those of the Federal Reserve Bank of Richmond or the Federal Reserve System.
Receive a notification when Economic Brief is posted online.