Memorial Day 2025
What happens to professional driving networks when America stops working? A full-week study of May 19-26, 2025, bracketing the federal holiday on Monday, May 26, measured against a four-week same-day-of-week baseline.
·
The deepest single-day fleet shutdown of the year.
Memorial Day 2025 (Monday, May 26) produced the most dramatic traffic signal of the study period, with Nexar fleet activity dropping 22.9% below the baseline Monday average, from 248,557 rides to 191,540. The Memorial Day weekend (Saturday May 24 and Sunday May 25) also showed traffic reductions of 4.2% and 4.9% respectively.
Memorial Day 2025 drove a −22.9% drop in US fleet traffic on the holiday Monday, accompanied by a +6.9% increase in average driving speed, reflecting significantly reduced road congestion as commercial vehicles park and leisure travelers take over the road network.
The pre-holiday week showed a clear ramp-down: traffic was near-normal Monday through Wednesday but fell sharply Thursday (−8.6%) and Friday (−11.5%) as operators began winding down ahead of the long weekend. Vermont (+5.3%) and Maine (+1.4%) were the only states to post positive holiday-weekend deltas, driven by rural and leisure-adjacent fleet activity.
A peak consumer weekend, a commercial shutdown.
Memorial Day is observed on the last Monday of May in the United States, honoring military personnel who died in service. In 2025, it fell on May 26. The holiday creates a three-day weekend (May 24-26) that is one of the highest-travel periods of the year for consumer leisure trips.
Historically, AAA reports 38-45 million Americans traveling by car over Memorial Day weekend, making it one of the top three busiest travel weekends annually alongside Thanksgiving and Independence Day. However, commercial and fleet traffic tells a very different story: trucks, taxis, and company vehicles predominantly reduce or cease operations on federal holidays.
This analysis uses Nexar's network of fleet dashcam vehicles as a real-time sensor of commercial fleet activity across the United States. The question is not "did Americans travel more?", that is already known from AAA and tourism data. The question is: by how much do US commercial fleets contract on a federal holiday, and where, and for how many days? The signal is large, sharp, and geographically structured.
Memorial Day produces an intentional, calendar-anchored perturbation in commercial fleet operations. Unlike weather-driven or demand-driven variation, this drop is structurally predictable, making it the ideal test case for fleet utilization models, route scheduling logic, and holiday-period planning across logistics, last-mile, and TaaS networks.
Methodology
Data source. Primary table: nexar-data-warehouse.Intermediate.IT_TRX_RIDE_LOCATIONS_CLASSIFIER, partitioned by ride_start_date. All queries filtered to country = 'United States of America'.
Study period & baseline. Memorial Day weekend May 24, 25, 26; pre-holiday week May 19-23; baseline (same-DOW) the prior 3 weeks (May 3-5, 10-12, 17-19); geographic scope United States. Metrics: total ride count, total miles, total hours, hours/ride & miles/ride, average speed mph, and % change and Z-score vs baseline. National totals are decomposed by ride starting-point state and across 24 major metro areas; assignments use the ride's starting location, not destination.
Z-scores are computed as z = (study − μ_baseline) / σ_baseline with mean and SD drawn from the three same-day-of-week comparators. Three comparators is the minimum viable baseline window for this study, see Section 14 for the power implications and proposed extensions.
Key findings.
Six headline findings: three on volume, one on speed, one on geographic dispersion, one on resilience.
An extreme Monday collapse.
The three-day window: a moderate Saturday softening, a slightly deeper Sunday, and an extreme Monday collapse.
| Date | Day | Fleet Rides | Baseline | % Change | Hrs/Ride | Mi/Ride | Avg Speed |
|---|---|---|---|---|---|---|---|
| May 24 | Saturday | 228,600 | 238,621 | −4.2% | 0.676 | 14.07 | 20.81 mph |
| May 25 | Sunday | 200,338 | 210,580 | −4.9% | 0.673 | 14.12 | 20.98 mph |
| May 26 ★ | Monday · Holiday | 191,540 | 248,557 | −22.9% | 0.660 | 14.28 | 21.65 mph |
57,017 fewer rides than the baseline Monday average. With a Z-score of −36.1, this is an extreme statistical reduction by any standard, driven by federal holiday fleet shutdowns across commercial, logistics, and taxi/rideshare sectors. A drop of this magnitude does not occur from random day-to-day variation; it is the signal of an entire fleet segment ceasing operation in lockstep.
Speed Insight · Memorial Day Monday
Despite far fewer rides, average speed on Memorial Day Monday reached 21.65 mph, the highest single-day average of the entire study period. With commercial freight largely off the road, the remaining Nexar-equipped vehicles experienced notably less congestion.
A characteristic ramp-down.
The week leading into Memorial Day shows a characteristic ramp-down across the second half. Monday through Wednesday rode close to baseline (±1%); the divergence opened Thursday and deepened Friday.
| Date | Day | Rides | Baseline Avg | % Change | Pattern |
|---|---|---|---|---|---|
| May 19 | Monday | 249,979 | 245,017 | +2.0% | Normal week start |
| May 20 | Tuesday | 255,446 | 252,552 | +1.1% | Slightly elevated |
| May 21 | Wednesday | 254,436 | 255,466 | −0.4% | Flat / baseline |
| May 22 | Thursday | 240,891 | 263,441 | −8.6% | Ramp-down begins |
| May 23 | Friday | 233,871 | 264,185 | −11.5% | Pre-holiday wind-down |
Fleet activity was normal through Wednesday (±1%) but dropped significantly Thursday (−8.6%) and Friday (−11.5%) as fleet operators prepared for the long weekend. Many commercial operators effectively begin their "holiday" on Thursday, creating a 5-day effective shutdown for portions of the fleet.
Plains contract, New England holds.
The Memorial Day weekend (May 24-26) vs same-DOW baseline shows significant geographic variation. Plains states lead the contraction; New England leads the resilience.
| Largest Contractions | % Change | Most Resilient | % Change |
|---|---|---|---|
| Kansas | −25.5% | Vermont | +5.3% |
| Alaska | −24.0% | Maine | +1.4% |
| Nebraska | −23.6% | Delaware | −4.5% |
| Iowa | −21.4% | Rhode Island | −5.0% |
| Louisiana | −20.0% | New York | −6.3% |
| South Carolina | −19.6% | ||
| Mississippi | −18.8% | ||
| Indiana | −17.8% | ||
| Colorado | −17.5% | ||
| Tennessee | −15.9% |
Indiana saw a −17.8% fleet traffic drop over the Memorial Day weekend. The Indianapolis 500 (held on May 25) is a major local event that typically draws 300,000+ spectators, but Nexar's fleet-camera data shows the commercial side of Indiana's road network actually contracted significantly, consistent with the city's event logistics being front-loaded in the days prior to race day.
Major states by volume. The nine highest-volume state fleets, ranked by absolute baseline volume.
| State | Study Rides | Baseline | % Change | Assessment |
|---|---|---|---|---|
| New York | 134,857 | 143,857 | −6.3% | Moderate |
| California | 94,518 | 104,924 | −9.9% | Moderate |
| Texas | 49,425 | 57,096 | −13.4% | Elevated |
| Florida | 38,349 | 45,140 | −15.0% | Elevated |
| New Jersey | 32,539 | 36,454 | −10.7% | Moderate |
| Pennsylvania | 20,009 | 22,840 | −12.4% | Elevated |
| Illinois | 18,756 | 21,344 | −12.1% | Elevated |
| Georgia | 18,466 | 20,689 | −10.7% | Moderate |
| Nevada | 12,019 | 13,200 | −8.9% | Moderate |
The three largest-volume markets (New York, California, Texas) all moderate. Drops sharpen in mid-tier freight-heavy states: Florida (−15.0%), Pennsylvania (−12.4%), Illinois (−12.1%). New York's relative resilience reflects its 24/7 urban fleet composition. The 10 deepest-declining states are all Plains, Mountain, or Southern freight-heavy geographies, the pattern is structural. Source: IT_TRX_RIDE_LOCATIONS_CLASSIFIER, ride starting-point state.
Entertainment hubs lead the drop.
Twenty US metropolitan areas with notable Memorial Day traditions, ranked by % change. Entertainment hubs (Nashville, Miami, Indianapolis) lead the drop; tourism-adjacent cities (Vegas) and 24/7 metros (New York) hold steady; Minneapolis posts a marginal gain.
| City | ST | Study | Baseline | % Change | Context |
|---|---|---|---|---|---|
| Nashville | TN | 944 | 1,218 | −22.5% | Major entertainment hub · fleet parks |
| Indianapolis | IN | 968 | 1,225 | −21.0% | Indy 500 · logistics pre-positioned |
| Miami | FL | 2,091 | 2,587 | −19.2% | South Florida holiday effect |
| Dallas | TX | 2,086 | 2,563 | −18.6% | Commercial fleet holiday shutdown |
| Houston | TX | 5,749 | 6,918 | −16.9% | Energy / industrial fleet shutdown |
| Denver | CO | 1,124 | 1,353 | −16.9% | Outdoor recreation · fleet drops |
| Washington DC | DC | 876 | 1,037 | −15.5% | Federal holiday · gov't fleet parks |
| Seattle | WA | 1,209 | 1,425 | −15.2% | Pacific NW holiday pattern |
| Orlando | FL | 2,164 | 2,520 | −14.2% | Tourism up · fleet contracts |
| Austin | TX | 1,765 | 2,055 | −14.1% | Reduced commercial · leisure up |
| Los Angeles | CA | 10,178 | 11,632 | −12.5% | Leisure traffic displaces fleet |
| Atlanta | GA | 1,408 | 1,602 | −12.1% | Moderate · major hub city |
| Philadelphia | PA | 3,085 | 3,496 | −11.8% | Mid-Atlantic corridor slowdown |
| New Orleans | LA | 383 | 432 | −11.5% | Entertainment hub · fleet affected |
| San Diego | CA | 2,648 | 2,961 | −10.6% | Beach destination consumer surge |
| Boston | MA | 1,144 | 1,277 | −10.4% | Moderate impact |
| Chicago | IL | 4,290 | 4,766 | −10.0% | Major leisure destination |
| Las Vegas | NV | 6,307 | 6,815 | −7.5% | Tourism surge · 24/7 fleet ops |
| New York | NY | 87,468 | 93,289 | −6.2% | Large fleet · moderate impact |
| Minneapolis | MN | 518 | 516 | +0.4% | Marginal · resilient composition |
Source: IT_TRX_RIDE_LOCATIONS_CLASSIFIER, ride starting-point city.
The fastest day of the window.
Average speed, hours per ride, and miles per ride across the study window. Memorial Day Monday is the standout, the fastest single-day average of any day in the study.
| Date | Day | Avg Speed (mph) | Baseline Speed | Speed Δ | Hrs/Ride | Mi/Ride |
|---|---|---|---|---|---|---|
| May 19 | Monday | 20.26 | - | - | 0.719 | 14.57 |
| May 20 | Tuesday | 20.24 | - | - | 0.723 | 14.63 |
| May 21 | Wednesday | 20.18 | - | - | 0.746 | 15.06 |
| May 22 | Thursday | 19.82 | - | - | 0.750 | 14.86 |
| May 23 | Friday | 20.21 | - | - | 0.756 | 15.27 |
| May 24 | Saturday | 20.81 | 20.14 | +3.3% | 0.676 | 14.07 |
| May 25 | Sunday | 20.98 | 20.66 | +1.5% | 0.673 | 14.12 |
| May 26 ★ | Monday · Holiday | 21.65 | 20.26 | +6.9% | 0.660 | 14.28 |
Despite the large drop in ride counts, hours-per-ride and miles-per-ride remained stable or slightly compressed on Memorial Day Monday (0.660 hrs vs 0.712 baseline), suggesting the remaining fleet trips were shorter urban runs rather than long-haul journeys. Essential services (delivery, emergency, utilities) maintain activity while long-haul freight parks.
Far beyond the noise envelope.
Z-score analysis across the five most informative comparison days. Memorial Day Monday is not merely outside the noise envelope, it is far beyond it.
| Day | Study | μ Baseline | σ Baseline | % Change | Z-Score | Significance |
|---|---|---|---|---|---|---|
| Sat May 24 | 228,600 | 238,621 | 3,553 | −4.2% | −2.82 | Significant (p<0.01) |
| Sun May 25 | 200,338 | 210,580 | 5,389 | −4.9% | −1.90 | Marginal (p≈0.06) |
| Mon May 26 ★ | 191,540 | 248,557 | 1,580 | −22.9% | −36.1 | Extreme (p<0.0001) |
| Fri May 23 | 233,871 | 264,185 | 7,480 | −11.5% | −4.05 | Highly Sig. (p<0.001) |
| Thu May 22 | 240,891 | 263,441 | 2,471 | −8.6% | −9.13 | Highly Sig. (p<0.001) |
A Z-score of −36 lies far outside any reasonable noise envelope, orders of magnitude beyond conventional significance thresholds (Z=−1.96 for p<0.05, Z=−3.29 for p<0.001). The result is not "statistically significant" in the marginal sense; it is a structural reorganization of the fleet's operating state, observed for a single day, then released. The pre-holiday Thursday and Friday drops (Z=−9.1 and Z=−4.1) are themselves highly significant, confirming the multi-day ramp-down. Saturday and Sunday also clear conventional significance. Z-score = (study_value − baseline_mean) / baseline_SD, baseline SD across 3 comparison dates per DOW.
The commercial counterpart to AAA.
How Nexar's measurements compare against published commercial-fleet and consumer-traffic benchmarks. All four metrics are either within range of expectation or directly explainable by the fleet-vs-consumer composition gap.
| Metric | Nexar Finding | External Benchmark | Assessment |
|---|---|---|---|
| Holiday Monday fleet drop | −22.9% | Commercial fleet: −20% to −40% | Validated |
| Weekend consumer traffic | −4.6% (fleet) | Consumer leisure: +5% to +15% | Opposite (fleet vs consumer) |
| Pre-holiday Friday drop | −11.5% | Commuter / fleet: −10% to −20% | Validated |
| Speed increase (holiday) | +6.9% | Less congestion, faster speeds | Consistent |
Three of four findings sit comfortably inside the published commercial-fleet range. The fourth, the apparent contradiction with consumer leisure traffic, is the expected outcome of measuring a different population (commercial fleets) than AAA measures (consumer vehicles). Far from contradicting the consumer narrative, the Nexar data completes it: it provides the commercial counterpart to AAA's leisure measurement.
A fleet sensor, not a total-traffic sensor.
Why Nexar's data shows the opposite of AAA's consumer-traffic narrative.
AAA reports that ~45 million Americans travel by car on Memorial Day weekend, a significant increase over normal weekends. Nexar data shows the opposite trend. This is not a contradiction, it reflects Nexar's fleet composition.
Nexar cameras are predominantly installed in commercial fleets: trucks, taxis / rideshare, company vehicles, utility vans. Consumer personal vehicles, the leisure Memorial Day traveler, are largely not Nexar-equipped. Commercial fleets park on federal holidays; consumer vehicles flood the highways. Result: Nexar data captures the commercial traffic collapse while AAA measures the consumer surge.
This dual reality is what makes Memorial Day uniquely interesting for Nexar analysis: the roads experience more total vehicles (consumer leisure) but far fewer commercial vehicles (fleet operators). The combination produces faster average speeds for the Nexar fleet remaining on the road, they experience the benefit of reduced commercial congestion even amid higher overall traffic volumes.
Conclusions & recommendations.
Five findings translated into operational recommendations for fleet managers and analysts.
Major fleet shutdown signal
Memorial Day = the deepest single-day fleet drop of the year. A −22.9% drop on the holiday Monday is a strong, statistically ironclad signal that US commercial fleet operations virtually pause on Memorial Day. Fleet managers should build this into route scheduling and utilization models.
Ramp-down begins Thursday
Treat Memorial Day as a 4-5 day reduced-capacity window. Thursday (−8.6%) and Friday (−11.5%) reveal that the effective "holiday period" for fleet operators starts 3-4 days before the holiday. Capacity planning should mirror this longer arc, not the calendar day.
Speed benefit for active fleet
Holiday operation = meaningfully less congestion. Vehicles remaining on the road during Memorial Day Monday experience +6.9% faster average speeds. Last-mile delivery operators who choose to operate on the holiday may find it advantageous from a time-efficiency perspective.
Geographic variation is significant
Plains states drop deepest; coastal leisure resilient. Plains states (KS, NE, IA) show the deepest fleet drops (−21% to −26%), likely driven by high agricultural / freight fleet concentration. Coastal leisure states (VT, ME) show resilience or marginal gains. Fleet operators should model regional holiday effects separately.
Key city insight
Miami & Dallas lead the drop; NYC & Vegas hold steady. Miami (−19.2%) and Dallas (−18.6%) show the steepest drops among major cities, while New York (−6.2%) and Las Vegas (−7.5%) show more resilience, due to essential services, tourism-adjacent fleets, and 24/7 city operations.
Limitations.
Limitations are presented as a first-class section, not a footnote. These are not excuses for null results; they define the scope within which any finding should be interpreted.
Fleet composition bias
Nexar data captures commercial / fleet vehicles, not the total vehicle population. Consumer leisure traffic, the bulk of Memorial Day road use, is not directly observed by this dataset.
Geographic representation
Nexar density varies by state. Small-state findings (VT, ND, WY) are based on smaller sample sizes and should be interpreted with caution; their deltas are more sensitive to single-day noise.
Baseline period
Only 3 comparable weeks are available per day-of-week (prior to the holiday). Extended baselines would improve Z-score precision and reduce variance estimation error.
Starting-point geography
State and city assignment is based on the ride's starting location, not destination. Long-haul trips are attributed to the origin state, overstating origin-state contraction for fleets that primarily depart on holidays.
Year-over-year comparison unavailable
This analysis uses 2025 data only; year-over-year trends require 2024 comparison data which is outside the study scope. Multi-year context is therefore not addressed.
Appendix · raw data tables.
The full study-period daily series and the 4-week same-DOW baseline summary, for reproducibility.
Full Study Period · Daily Data
| Date | Day | Rides | Miles | Hours | Hrs/Ride | Mi/Ride | Avg MPH |
|---|---|---|---|---|---|---|---|
| 2025-05-19 | Monday | 249,979 | 3,641,845 | 179,775 | 0.7192 | 14.569 | 20.26 |
| 2025-05-20 | Tuesday | 255,446 | 3,736,029 | 184,570 | 0.7225 | 14.626 | 20.24 |
| 2025-05-21 | Wednesday | 254,436 | 3,830,421 | 189,782 | 0.7459 | 15.055 | 20.18 |
| 2025-05-22 | Thursday | 240,891 | 3,578,322 | 180,574 | 0.7496 | 14.855 | 19.82 |
| 2025-05-23 | Friday | 233,871 | 3,571,251 | 176,720 | 0.7556 | 15.270 | 20.21 |
| 2025-05-24 | Saturday | 228,600 | 3,215,552 | 154,510 | 0.6759 | 14.066 | 20.81 |
| 2025-05-25 | Sunday | 200,338 | 2,828,743 | 134,839 | 0.6731 | 14.120 | 20.98 |
| 2025-05-26 ★ | Memorial Day | 191,540 | 2,735,053 | 126,317 | 0.6595 | 14.279 | 21.65 |
Baseline Summary · 4-Week Average by Day-of-Week
| Day of Week | Avg Daily Rides | Avg Miles | Avg Hours | Hrs/Ride | Mi/Ride |
|---|---|---|---|---|---|
| Sunday | 210,580 | 3,020,882 | 146,229 | 0.6945 | 14.347 |
| Monday | 245,017 | 3,485,674 | 174,479 | 0.7121 | 14.222 |
| Tuesday | 252,552 | 3,582,197 | 179,516 | 0.7108 | 14.181 |
| Wednesday | 255,466 | 3,681,853 | 184,899 | 0.7238 | 14.409 |
| Thursday | 263,441 | 3,881,474 | 194,772 | 0.7394 | 14.734 |
| Friday | 264,185 | 3,846,905 | 193,753 | 0.7334 | 14.562 |
| Saturday | 238,621 | 3,347,257 | 166,678 | 0.6986 | 14.030 |
Analysis conducted by Rui Carneiro · nexar-ai.com. Data source · Nexar IT_TRX_RIDE_LOCATIONS_CLASSIFIER · US fleet vehicle network. Memorial Day 2025 · Generated June 2025 · Data period May 19-26, 2025 · Baseline April 28 - May 23, 2025.