What a decade of ridesharing actually did to American cities

Back in 2013, if I’d grab a cab outside Baltimore Penn Station for work, I’d insist the driver verbally promise me he’d take a credit card. They’d often say the card readers were broken. The windows didn’t always work, the fleet seemed run down.

Such is the feel of monopoly: In 2013, Baltimore’s Yellow Cab held half of that city’s 1,100 taxi permits, as Technical.ly reported back then. Many wouldn’t even reliably come to neighborhoods outside the downtown Inner Harbor. Similar taxi fleets, worn down and with limited coverage, served cities across the United States.

Into that gap stepped Uber, Lyft, and a now-forgotten swarm of regional competitors: Hailo, Sidecar, Gett, Curb, Via. I remember Baltimore’s rickety taxis well; their often rotten service pushed me to familiarize myself with the city’s bus system.

These ride-hailing companies were launching as fast-growing startups, injecting competition into a regulated and monopolistic industry, before their widespread success helped fuel a tech-lash. Even then, the Technical.ly newsroom reported on the threat these companies posed to taxi drivers, especially in a poor city like Baltimore — which was one of the first 20 cities in the country where Uber and Lyft launched. 

Across the country, taxi ridership collapsed as ride-hailing scaled. In Chicago, which publishes granular trip data, annual taxi rides fell from 32 million in 2014 to 5.8 million in 2023, an 82% decline that has stabilized at this lower floor even as rideshare trips surged to nearly 70 million. Similar trends happened nationwide.

Almost 15 years later, ride-hailing has established itself in local economies, for better or worse. 

This early “sharing economy” poster child has plenty to teach about the complexity of predicting how new tech-enabled logistics will affect local communities.

This year, California’s Assembly Bill 1340 took effect, granting rideshare drivers the right to unionize — the first such law in the nation. It’s a milestone that invites reflection on what this industry actually delivered, and what it cost. Meanwhile, a wave of delivery robots are appearing in cities across the country.

Policymakers, economic development leaders and entrepreneurs ought to consider the lessons.

The consumer surplus was real

Economists argue riders got genuine value.

A widely-cited economic study using Uber’s surge pricing data found that every $1 spent on UberX generated approximately $1.60 in consumer surplus — the difference between what riders were willing to pay and what they actually paid. Extrapolated nationally, that’s roughly $6.8 billion in consumer value in 2015 alone.

Most of that long-term value has come in more transparent and efficient rides from better-cared for automobiles. More distributed drivers managed by algorithms now effectively cover service areas comprehensively. A 2016 Pew study found that residents of majority-minority neighborhoods were five times more likely than others to say ridesharing serves areas traditional taxis won’t.

A Los Angeles field audit of 1,680 trips found ride-hail fares were about 40% cheaper than taxis on matched routes, with wait times roughly one-quarter as long. Perhaps most telling: about 1 in 5 taxi riders in the study were never picked up at all. Ride-hail “nearly guaranteed” pickup.

That reliability premium, not just price, is what disrupted the taxi industry. The product wasn’t “a car ride.” It was a bundled reduction in search costs, uncertainty and the very real possibility of being stranded.

‘Disrupting taxis’ doesn’t describe what actually happened

Counter-intuitively, ride-hailing didn’t primarily replace taxi trips. It replaced everything else.

Across multiple city surveys synthesized in transportation research, the breakdown of “what would you have done without this Uber/Lyft trip?” looks roughly like this:

• ~60%: Would have taken transit, walked, biked or not traveled at all
• ~20%: Would have used a personal vehicle
• ~20%: Would have taken a taxi

A Denver-region study found only 9.6% of surveyed riders would have taken a taxi. Meanwhile, 22.2% would have used public transit and 12.2% wouldn’t have traveled at all.

More Builders by Chris Wink

The best-connected person in your ecosystem might be its biggest weakness

Finding the signal in AI’s noise

How would an ecosystem-led investment fund perform?

This matters enormously for cities. Ride-hailing didn’t just shift riders between for-hire vehicles. It induced new trips and pulled riders off buses, bikes, and sidewalks onto single-occupant car travel.

Those consequences showed up in traffic data. A metro-area panel study published in Nature Sustainability found that ride-hailing entry was associated with:

• +0.9% increase in congestion intensity
• +4.5% increase in congestion duration
• -8.9% decline in transit ridership

In San Francisco, daily vehicle miles traveled rose from 4.9 million in 2010 to 5.6 million in 2016, a 13% increase. A detailed analysis attributed “transportation network companies” (TNCs — what researchers call Uber and Lyft) as the single largest contributor to the city’s congestion growth in that period.

The mechanism is straightforward: “deadheading.” Drivers cruise empty between pickups. Estimates put out-of-service travel at roughly 20% of ride-hail vehicle miles traveled in San Francisco and as high as 50% in New York.

For drivers, the picture is genuinely mixed — and sensitive to how you count.

Early Uber data from 2015 boasted gross hourly earnings of $18-24 per hour in major markets, which looked competitive with traditional taxi and chauffeur wages. But that Denver-region study that fully accounted for deadheading, downtime and vehicle expenses found much lower net figures: gross wages averaged $15.57/hour, but net hourly wages ranged from just $5.72 to $10.46/hour depending on expense assumptions.

What’s unambiguous is that drivers value flexibility. Structural economic estimates find drivers derive substantial “surplus” from choosing their own hours in real time, and would reduce their hours by more than two-thirds if forced into inflexible schedules.

This is the core tension California’s new union law attempts to address: flexibility is real compensation, but it doesn’t replace health insurance or retirement savings for drivers who work full-time on the platform.

What economic developers should learn

The ridesharing story offers hard-won lessons for anyone navigating emerging technology’s promises and externalities:

1. Consumer surplus doesn’t equal social surplus. Riders benefited. But when you add congestion, reduced transit ridership and shifted labor risk, the net social value becomes “context-dependent” — positive in transit-poor areas, likely negative in dense urban cores without countervailing policy.
2. Disruption often means cost-shifting, not cost-elimination. The taxi medallion system had real problems — like those poorly maintained Baltimore cabs. But ride-hailing didn’t eliminate those costs. It shifted them to drivers (who now bear vehicle expenses and lack benefits), to transit systems (which lost riders and revenue), and to all road users (who sit in more traffic).
3. The regulatory window matters. Once riders and drivers coordinate on a platform ecosystem, the politics and economics of changing course become much harder. Cities that set congestion pricing, driver minimums, and data transparency requirements early, like New York, have more leverage than those playing catch-up.
4. Watch the mode substitution, not just the disrupted incumbent. If your city celebrates ride-hailing for “disrupting taxis” while transit ridership falls 9%, you may have traded a small regulated monopoly for a larger unpriced externality.

The regulatory landscape has solidified. California’s Proposition 22 was upheld by the state supreme court, keeping most drivers as independent contractors with union rights layered on top. Massachusetts created a minimum earnings standard through settlement. New York City’s Taxi and Limousine Commission continues refining per-trip pay formulas explicitly designed to prevent platforms from“manipulating driver availability.”

The Uber-Lyft duopoly is entrenched. The dozens of regional alternatives from 2013 are gone. What remains is a permanent feature of urban transportation, one that created real value for riders, restructured labor markets in ways we’re still debating, and reminded every city that private benefits and social costs don’t automatically align.

For economic developers watching the next wave of platform-mediated disruption — whether in AI, delivery logistics, or something not yet named — the ridesharing decade offers a simple framework: Large private benefits plus meaningful external costs equals a strong case for pricing, transparency and labor standards, rather than binary “ban versus free-reign” debates.

Last fall, I was back in a Baltimore taxi for the first time in years. The credit card machine worked, and I didn’t even have to ask.