American car owners spend thousands of dollars each year on payments, insurance, and maintenance for vehicles that spend the vast majority of their existence doing nothing. Federal travel data collected by the U.S. Department of Transportation show that the typical driver logs roughly one hour of driving per day, leaving the car stationary for more than 23 hours out of every 24. That ratio, once converted, produces the widely cited figure: the average car sits parked about 95 percent of the time it is owned. The statistic has shaped debates over urban parking policy, curbside pricing, and how cities plan charging infrastructure for electric vehicles, yet the number itself rests on survey methods that have not kept pace with changes in how Americans actually use their cars.
Why a 95 percent parked share changes the cost equation
The practical consequence of the 95 percent figure is straightforward: a depreciating asset worth tens of thousands of dollars occupies valuable land almost every hour of the day. Cities devote enormous amounts of space to surface lots and garages sized for peak demand, even though most of those spaces sit empty overnight and on weekends. UCLA researcher Donald Shoup converted the government driving-time data into a parking-share calculation and argued that drivers should pay for parking in proportion to the real cost of that space. His analysis reframed the question: the problem is not that Americans drive too much, but that they store cars on public land for free during the 95 percent of the day those cars are not moving.
That reframing carries direct financial weight for households and municipalities alike. If a vehicle is parked 95 percent of its life, then the land under it is, in effect, subsidized storage. Pricing that storage at market rates, Shoup’s core proposal, would push some owners to reconsider second or third household vehicles. For cities evaluating EV charger placement, the ratio also matters: a car that sits in one spot for 23 hours a day needs access to a charger at that spot, not at a gas station visited for five minutes. This logic underpins arguments for residential and workplace charging, where long dwell times align with slow, lower-cost charging equipment rather than fast chargers designed for brief stops.
A reasonable hypothesis follows from the data structure itself. The National Household Travel Survey records travel at the person level, not the vehicle level, and households with multiple cars tend to leave at least one parked all day. If NHTS vehicle-level GPS subsets were reweighted by household fleet size, the measured parked share would likely rise above 96 percent in metropolitan areas with more than one million residents, where multi-car households are common and commute distances are shorter relative to suburban sprawl zones. No published federal table confirms that specific threshold, but the survey architecture points in that direction and suggests that the 95 percent figure is probably conservative for dense urban regions.
Federal survey data behind the parking calculation
The 95 percent figure traces back to a simple conversion of minutes driven per day, a metric the federal government has tracked for decades. The daily travel summaries from the National Household Travel Survey, published by the Bureau of Transportation Statistics, report average driving time per person and per driver. The survey collects household trip logs that record trip purpose, travel time, time of day, and vehicle attributes across a nationally representative sample, providing the raw material for estimates of how much time cars spend moving versus sitting still.
Earlier waves of the same program, then called the Nationwide Personal Transportation Survey, produced broadly similar results. The 1990 NPTS documentation from the Federal Highway Administration’s Office of Highway Information Management describes how surveyors gathered daily travel diaries and summarized average minutes per day of driving. The 1995 NPTS dataset, archived at the Bureau of Transportation Statistics’ repositories, contains household, person, vehicle, and travel-day trip files that allow independent replication of the driving-time calculation. Later NHTS microdata, made available through federal research partners, extend those patterns into the 2000s and 2010s.
The arithmetic is not complicated. If the average driver spends about 60 minutes behind the wheel on a given day, that leaves 1,380 minutes, or 23 hours, when the vehicle is not in motion. Divide 1,380 by 1,440 total minutes in a day and the result is just over 95 percent. The calculation assumes a single-driver vehicle and treats all non-driving time as parked time, which is a simplification but not an unreasonable one for most privately owned cars that are rarely used by more than one household member at the same moment.
What makes the 95 percent statistic compelling is its intuitive match with lived experience. In most neighborhoods, a drive down the block at midday reveals rows of parked cars, many of which will not move until the workday ends. Office garages fill in the morning and sit largely unchanged until evening. Even vehicles used for commuting typically make only two trips per day, leaving long stretches of idle time at home and at work. The federal survey numbers translate that visual impression into a measurable ratio.
Gaps in the evidence and what they mean for drivers
The 95 percent figure is durable partly because it is hard to disprove with the same data that produced it. No primary FHWA or BTS table directly publishes annual hours parked per vehicle. The number is derived from minutes-driven summaries, and the underlying surveys rely on self-reported travel days rather than continuous vehicle telemetry. Respondents record trips on a single assigned travel day, which may not capture weekend errands, late-night drives, or days when a car is lent to another household member. Any unusual day can skew reported averages for that household, and the survey design smooths those irregularities only at the national level.
The NHTS microdata downloads and the archived NPTS files also lack odometer-based validation of total parked time across multi-vehicle households. A family with three cars might report one adult’s travel day in detail while the other two vehicles sit in the driveway unreported. That gap almost certainly pushes the true fleet-wide parked share higher than 95 percent, not lower, because the survey counts the movement of the most-used vehicle while ignoring the complete idleness of the least-used one. Occasional under-reporting of short trips, such as quick errands or school drop-offs, would add only a few minutes of driving time and is unlikely to overturn the basic parked-to-moving ratio.
There are other blind spots. Commercial fleets, car-share services, and taxis do not fit neatly into the household travel framework, yet they represent some of the most intensively used vehicles on the road. A taxi or delivery van may be in motion for eight or ten hours per day, dramatically reducing its parked share. Because NHTS focuses on personal travel, these high-utilization vehicles are largely outside its scope. The 95 percent statistic should therefore be understood as a statement about privately owned passenger cars, not about all vehicles in the economy.
For individual drivers, the methodological caveats do not erase the central lesson: most privately owned cars are idle almost all the time. That reality has practical implications for how households think about ownership. A car that sits for 23 hours a day behaves more like a subscription to access mobility on demand than like a constantly used appliance. Some households may decide that one shared vehicle, supplemented by car-share or ride-hailing, provides sufficient coverage. Others may see the sunk cost of a second car and look for ways to put it to work through peer-to-peer rental platforms or informal carpooling.
For cities, the evidence gaps point toward a need for better data rather than a rejection of the 95 percent benchmark. Continuous GPS logging, anonymized and aggregated, could provide a more precise picture of how long vehicles remain parked on different types of streets and in different neighborhoods. Until such datasets are widely available, planners will continue to rely on survey-based estimates and simple arithmetic conversions. Those tools are blunt, but they are consistent: whether the true number is 95, 96, or 97 percent, the conclusion is the same. The American car is, above all else, a parked object, and policies that ignore that fact risk misallocating both money and urban space.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
