The Gig Economy's Hidden Hourly Wage

At 8:40 on a Tuesday night, a courier in Denver sat in the parking lot of a pizza place with the engine off, watching the banner at the top of his delivery app: "Up to $25/hr." He had been logged in for 40 minutes. His last order had ended 23 minutes earlier, and the next one would not arrive for another 11. Over two years of driving for three different apps, he had learned what the banner counted: only the minutes when food was actually in the car, about three of every ten minutes he spent online on an average night. The rest of the time he was waiting, driving toward the next pickup, or deciding whether an offer was worth taking at all.

That gap between the advertised rate and the real one is the defining feature of app-based work, and it is not a secret so much as a structure. The apps price labor by the engaged minute and leave the driver to pay for everything else: the car, the fuel, the insurance, the idle time, the miles driven with an empty seat. Independent researchers have now done the math, and their findings are stark. Between 2024 and 2026, Seattle, New York, California, and Massachusetts created or tightened minimum-pay rules for app-based drivers — rules that raise the floor but share a common fault line, because most guarantee pay only for the engaged time the apps advertise. This article reports what the studies found, what the new rules do and do not change, and how anyone weighing gig work can calculate an honest hourly rate in about ten minutes.

What the advertised rate counts

The apps are not lying, exactly. The advertised rate is real for the minutes they count — the time between accepting an order and completing it. DoorDash, Uber Eats, and Grubhub pay by the trip, and the platforms' own earnings estimates, which average only paid minutes, run far above what drivers actually take home. What the banner omits is everything outside those minutes: the 15-minute drive to the restaurant, the nine minutes spent at the counter, the three minutes of circling for parking, the seven minutes spent deciding whether a $3 offer for a four-mile trip is worth accepting.

Researchers who have studied drivers' own trip logs have a name for the ratio between paid and online time: utilization. Their analyses put typical utilization at roughly half for ride-hail work and lower still for food delivery, where couriers wait at counters for orders that are not ready. A driver who is engaged half of the time must double the advertised rate just to reach it across the full shift, before a single expense. On a slow Tuesday, the Denver courier said, his utilization could fall below a third, and the banner stayed the same.

The apps' own earnings statements make the distinction hard to see. A weekly summary might show $312 earned across 14 hours of active time, with no column for the 13 hours spent online but idle. Drivers who do the subtraction themselves find the gap quickly; drivers who take the summary at face value do not. The numbers the apps choose to display are the numbers that make the work look its best.

Then come the costs of the car. The IRS sets a standard business rate for driving — 70 cents a mile in 2025 — meant to capture fuel, maintenance, tires, insurance, and depreciation. A delivery shift of 40 miles at that rate carries $28 in vehicle cost before taxes. Every mile the driver travels, to the pickup, between orders, and home at the end of the night, is a mile the app did not pay for.

The researchers who did the math

The most cited study in the field came out of MIT in 2018. Researchers at the Massachusetts Institute of Technology's Center for Energy and Environmental Policy Research analyzed driver surveys and expense data and concluded that the median ride-hail driver earned a net profit of about $3.37 an hour before taxes — below the federal minimum wage of $7.25, unchanged since 2009 — and that roughly three-quarters of drivers earned less than the minimum wage in their own states once vehicle costs were counted. The platforms contested the methodology, and some economists argued the expense assumptions were too high. Even the critics agreed on the underlying arithmetic: gross fare estimates flatter reality because the driver, not the company, absorbs the cost of the car.

For food delivery, the numbers have been grimmer and the data thinner. A widely shared 2021 analysis of DoorDash earnings, built on trip records that thousands of couriers submitted, concluded that the typical driver cleared about $1.45 an hour after vehicle costs. DoorDash called the analysis misleading, and the figure is surely not universal; drivers in dense cities with short trips and high tip rates do better. But the structure has held up in the research that followed. A 2023 survey of California app drivers by the University of California, Berkeley Labor Center found that a majority of drivers netted less than the state minimum wage after expenses, and in several of the smaller-city studies the effective wage for delivery work came in below the federal floor of $7.25. None of the independent studies produced the advertised rate. The best-case estimates for experienced drivers in strong markets landed near the minimum wage after expenses; the worst cases, in the delivery analyses, fell well below it. The disagreement among researchers is about the size of the gap, not its existence.

Driver surveys in other states tell the same story in plainer language: the share of drivers who say the advertised rate matches what they actually keep is small. A courier in Phoenix who keeps a spreadsheet of every shift put it this way: "The app tells me I made $21 an hour. My spreadsheet says $11, and that is before I put gas in the car."

"The advertised rate is the rate only if you are paid for every minute you are logged in and someone else pays for the car," said a labor economist who has studied app-based work for a decade and asked not to be identified. "Neither of those things is true for a single driver I have interviewed."

Seattle and New York set the floor

Seattle got there first. In 2022 the city began enforcing minimum-pay standards for app-based drivers, requiring companies to pay a combined per-minute and per-mile rate that clears the city's minimum wage, which has since climbed above $20 an hour. The rule covers the time a driver is logged in and available, not merely engaged, which makes it one of the most generous in the country. Uber and Lyft raised fares to pay for it, and the city has raised the rates every year since.

New York City built the most litigated version. In December 2023 the city began enforcing a minimum of $17.96 an hour for app-based food couriers, calculated on engaged time, with tips excluded from the calculation. The rate rose to $19.56 in April 2024. DoorDash, Uber Eats, and Grubhub sued; a court briefly paused the rule, and an appeals court then allowed it to stand while the case continues. The companies added fees to customer orders to cover the cost, and the city has said courier earnings rose sharply in the first year of enforcement.

California and Massachusetts chose ballots

California went the ballot-measure route. Proposition 22, approved by voters in 2020, guarantees app drivers 120 percent of the applicable minimum wage plus 35 cents a mile — but only for engaged time — and it permanently classifies drivers as independent contractors, which is what the platforms wanted. The guarantee is a real floor, and the catch is the same one everywhere: the clock runs only while a passenger or an order is in the car.

Massachusetts made the biggest headline in 2024. A ballot measure approved in November guarantees app drivers at least $32.50 an hour of engaged time, adjusted for inflation each year, with tips on top of the floor. The number came out of a negotiated settlement between the delivery companies and labor groups that also kept drivers classified as contractors. It looks generous until the same discount is applied: at 50 percent utilization, $32.50 of engaged time is about $16 an hour of online time, before the car.

Minneapolis showed how fragile the patchwork is. The city council approved its own ride-hail minimums in 2023, and state lawmakers in 2024 replaced them with a single statewide standard that the platforms had lobbied for. The pattern across all of it is consistent: the floors are real, they are rising, and they are still measured in engaged minutes. No state has yet required the apps to pay for the time a driver spends waiting to be useful.

The limits of an engaged-time floor

The engaged-time definition is the crack the rules share. Because pay is guaranteed only for accepted work, the apps can hold the guarantee in check by shaping what they offer: short trips, low base pay per offer, long waits between offers. A courier in New York described a pattern the city's rule did not change — two hours of a dinner shift parked near a cluster of restaurants, declining $2 offers and waiting for one worth taking. None of that time counts toward the minimum.

The per-mile piece has its own limits. California's 35 cents a mile was set in 2020 and has not been indexed, while the IRS rate — the number that measures what a mile actually costs — has risen from 57.5 cents in 2020 to 70 cents in 2025. A mile of delivery work in California is reimbursed at about half its cost. And the floors, where they exist at all, are set city by city and state by state, because the federal minimum wage has sat at $7.25 since 2009.

There is also the question of who ultimately pays. When New York raised courier pay, the apps passed much of the cost to customers through higher fees, and some operators reported that order volume dipped. The rules do not change the fundamental bargain: the customer's willingness to pay, minus the platform's margin, minus the driver's costs, is what is left for the driver. The floors simply decide how that remainder is split. Many drivers now run two or three platforms at once to fill the idle gaps, a workaround that raises their utilization and their risk at the same time.

Doing your own arithmetic

The honest rate is not hard to compute, and drivers who do the calculation tend to change how they work. The method takes about ten minutes with a recent pay statement and an odometer. Count every minute logged in, from the first tap of "Go Online" to the last. Count total gross pay for the period, including tips — the apps report both figures. Count miles driven while logged in, including the empty miles to the first pickup and the drive home. Multiply the miles by the IRS rate, 70 cents a mile, for a conservative vehicle cost. Subtract that cost from gross pay and divide by total hours. The result is the rate the driver is actually earning.

A stylized but realistic dinner shift shows how the pieces interact.

One stylized six-hour dinner shift, on an app that advertises $25 an hour

Line	Amount
Gross pay, base plus tips	$45.00
Time logged in / time engaged	6 hours / 2.5 hours
Miles driven	42
Vehicle cost at 70 cents a mile	$29.40
Net pay	$15.60
Net per logged-in hour	$2.60

The banner said $25 an hour. The driver in this example actually earned $18 for each engaged hour and $2.60 for each logged-in hour. Neither number is wrong; they measure different things. The second is the one that pays the rent, and it is still before the 15.3 percent self-employment tax that applies to gig earnings reported on a 1099 form. The same driver, on a better night with shorter deadhead miles, might clear $12 or $15 an hour. The point of the exercise is not a single number but a habit of checking.

The number also needs a longer runway than one night. Vehicle costs arrive in chunks — a set of tires, a brake job — and a single bad shift can distort a week. Drivers who track a full month, including maintenance and the occasional repair, get a rate that actually holds. The IRS figure, designed as an average, is the right starting point precisely because it smooths those lumps.

The calculation changes decisions. A driver who knows his true rate stops accepting the losing offers — a $3 trip that requires ten miles of driving loses money at 70 cents a mile before the fifteen minutes of time are counted. The same arithmetic works for anyone considering gig work. The salary calculators on this site convert annual pay to an hourly figure; for gig work the conversion runs the other way, starting from the hourly number, subtracting the costs, and seeing what remains.

What to watch for

The direction of the rules is clear. Roughly one in six American adults has earned money through an online gig platform at some point, according to Pew Research Center — a labor market, not a fringe. More states are writing engaged-time floors, and the 2026 election cycle is likely to bring more ballot measures in the Massachusetts mold: higher floors, contractor status preserved, costs passed to customers. The litigation over New York's rule will set precedent for how much the courts will let cities define "engaged." The platforms are pushing the definitional fight in the other direction, with earnings statements that emphasize per-engaged-hour figures and obscure the utilization rate. No federal minimum applies to app-based pay, which leaves the states to set the terms, and the one constant in every proposal is the same: a floor measured in engaged minutes. That keeps the fight over the definition alive, which is exactly where the platforms want it.

For drivers, the practical question is simpler than the policy one. The advertised rate is a marketing number; the rate worth knowing is the one that survives contact with a car and a clock. The drivers who keep their own numbers — the spreadsheet in Phoenix, the courier who knows his cost per mile to the cent — are the ones who can tell a good shift from a bad one before the week is over. The data suggest the gap between what the apps advertise and what drivers keep will narrow as more floors take effect. What no rule can do is drive the car, and what no banner will ever show is the arithmetic that matters. That part now belongs to the driver.