top of page

Technology Deep Dive: Carpool Comeback Takes a New Route

2 days ago
9 min read

In 1980, nearly one in five American workers got to work in a carpool. Today it is fewer than one in ten. For decades, transportation agencies have tried to reverse that slide with ride-matching databases, carpool lanes and parking perks, with mixed results. Most of those programs shared one assumption: that the job of a carpool program is to introduce commuters who live and work near each other.


On Houston’s ConnectSmart platform, a different pattern is emerging. Carpool trips are up roughly 29-fold since 2023, and most of that growth is not coming from strangers meeting through an app. It is coming from family members, friends, and coworkers who already trust each other, using technology to make sharing a ride easy enough to become a habit. Carpooling is coming back, but not along the route it left by.


How Carpool Faded


Carpooling’s high point grew out of necessity. The oil shocks of the 1970s brought soaring gas prices and lines at the pump. Many factories and offices ran fixed shifts, so neighbors could leave and come home together. Employers and newly formed ridesharing agencies promoted carpools and vanpools to save fuel. In 1980, 19.7% of US workers carpooled to work, according to the U.S. Census Bureau.


Graph of carpool decreasing from 1980 to 2022

The line shows how steady the decline has been. By 2013, the share was down to about 9%, even though the workforce grew by roughly 40 million people over the same period. Analysts point to a handful of forces that quietly made sharing a ride harder (Planetizen):


  • More cars per household. As incomes rose and cars became more reliable, more households could put every adult behind their own wheel. Once a second or third car sits in the driveway, sharing becomes a choice rather than a necessity.

  • Jobs and homes spread out. Suburban growth scattered both homes and workplaces. A carpool works best when several people travel the same corridor to the same place, and dispersed office parks made those matches rarer.

  • Errands joined the commute. In two-earner households, the drive to work now carries daycare drop-offs, grocery runs, and gym stops. A fixed carpool cannot wait at every stop, so people chose the flexibility of their own car.

  • Schedules loosened. Fewer people work factory shifts, and flextime spread. When coworkers no longer start and finish at the same minute, lining up a ride becomes a daily chore.

  • Coordination stayed manual. Traditional ride-matching programs introduced people and then stepped back. Commuters were left to sort out pickups, delays, cost-sharing, and what happens when someone cancels.


Then COVID hit. Sharing a car with someone outside your household suddenly felt like a health risk, and millions of jobs moved home. Carpooling’s share fell from 8.9% in 2019 to 7.8% in 2021, while the share of people working from home tripled, from 5.7% to 17.9% (TOMNET/TBD brief, ACS data). Carpooling recovered to 8.6% in 2022, but hybrid work left a lasting problem. A partner who comes in on Tuesdays and Thursdays is hard to pair with one who comes in on Mondays and Wednesdays, so the classic fixed-schedule carpool got even harder to keep.


The Carpool People Will Accept


When commuters look past driving alone, they usually choose someone they already know. They choose a ride with someone they already know, live in the same neighborhood, or work in the same building. That preference shows up in commute data, national travel surveys, recent carpool research, and in what ConnectSmart users actually do.


  • Next after driving alone. About 11% of Texas commuters carpool, against 1.3% who ride transit, so carpoolers outnumber transit riders about 8 to 1 (Texas Demographic Center, ACS 2016–2020).

  • Mostly household rides. About 70% of US carpools are household carpools, with everyone in the car living together (Blumenberg & Smart, 2010, 2001 National Household Travel Survey). Analysis of 1990 national data found a similar share.

  • Known partners. In a 2021 survey of 752 carpoolers in the Washington, DC and Charlotte regions, 91% rode with family (36%), friends (30%) or coworkers (25%). Only 9% rode with strangers (Verma et al., UT Austin).

  • Strangers are a harder sell. In a survey of 3,377 adults about future driverless ride-hailing, 48% were willing to ride alone or with family and friends, but only 20% to share with strangers (Batur, Pendyala & Magassy, 2023).

  • Houston history. Family members have made up about 75% of Houston carpools (Li et al., 2007). Even at carpooling’s peak, most shared rides were household trips.

  • ConnectSmart riders. Survey respondents among our most active carpoolers reported riding with people they already know, and none had yet joined a stranger match on their own; about a third said they would be open to it.

  • Reliance on the routine. Among our most frequent carpoolers, 83% say they would change or stop their routine without the app. Once the habit forms, people depend on it, and on the person beside them.


Why does a familiar face matter so much? A carpool is a small, repeated act of trust. You let someone into your car, or you get into theirs. You count on them to show up on time, to drive safely, and to tell you if plans change. With family or a friend, that trust already exists, and so do easy ways to settle costs, swap driving days or cover for a sick child. With a stranger, every one of those questions has to be answered from scratch, often by text, before the first ride.


Even “slugging,” the casual carpooling often cited as proof that strangers will share a ride, runs on social structure. In Northern Virginia, where slug lines feed Washington, DC, riders line up at the same park-and-ride lots at the same hour, bound for the same job centers, such as the Pentagon, Rosslyn and Crystal City. Unwritten rules govern the line and the car: first come, first served, little talk unless the driver starts it, no eating or fiddling with the radio. Riders learn these norms by watching, and the system polices itself (FHWA Casual Carpooling Focus Group Study, 2013). Over time, the same faces become familiar. One Northern Virginia slug described going “to the same spot at the same time and see familiar faces every day” (Northern Virginia Magazine). Sluggers are not true strangers; they are familiar strangers, held together by a shared place, a shared destination and shared rules. Slugging works because it builds a thin layer of trust, not because it does without it.

Trust is the gate. A carpool program that starts with strangers asks people to clear the highest bar first, and most never try. A program that starts with existing relationships lets people say yes before they have to think about it.


ConnectSmart’s Two Routes to a Shared Ride


ConnectSmart is a Houston-area mobility app managed by TxDOT Houston District. It pairs everyday driving tools, such as navigation, incident alerts, and construction detours, with rewards for smarter trips. Because drivers already use it daily, it can nudge them toward carpooling at the moment they plan a trip. It offers carpooling in two distinct ways, built for two different kinds of commuters.


Route 1: match with strangers. This is the classic model, rebuilt for hybrid work. A commuter posts a trip as a driver or rider, with route, time, seats, and preferences such as women-only matching. The app lists compatible carpools, and the commuter invites one. From there, the app carries the trip through the steps where legacy programs left people on their own: confirming and scheduling, coordinating pickup and drop-off, tracking arrival in real time, and rating the trip afterward. Commuters can carpool trip by trip, with no group to join, or set up scheduled HOV3 groups through their employer or with friends. Route 1 serves the third of commuters who say they are open to riding with someone new.


Posting a driving request & viewing available carpools
Route 1 in the app: post a trip (left), then invite a compatible match from the list of available carpools (right)

Route 2: match within your social network. This route starts where trust already exists and widens outward. Many families already share rides every day without calling it carpooling. Route 2 recognizes and rewards those trips, so the habit and the app become familiar. From there, the same tools make it easy to add a friend or neighbor, and then a coworker who lives nearby:


  1. Fampool: household and family, the closest ties.

  2. Friendpool: friends and the wider social circle.

  3. Coworker pool: colleagues and workplace ties.


carpool matching rings
The rings show social distance: each carpool habit forms in the inner circle, then spills outward to people we know less well.

Two features run across every stage. Easy Quick Match removes the setup: the driver shows a QR code, the rider scans it to check in, and the shared trip is recorded, with no searching, posting, or scheduling. Instant random post-trip rewards make each shared ride a small, slightly unpredictable win. That surprise keeps the experience fresh, and the steady record of shared trips helps a new routine stick.


scan driver's QR code & see pick-ups on route
Route 2 in the app: riders scan the driver’s QR code to check in (left), then the trip runs in carpool navigation with pickups (right)

Signs the New Route is Working


graph of carpool increases from 2023-2026

Each bar represents a year of carpool trips on ConnectSmart, and the blue label shows the growth over the previous year. Growth did not just continue; it accelerated: 2×, then 3×, then nearly 5×, with December still to come. Growth that compounds like this usually means something is spreading through people’s networks rather than through advertising: each new carpooler brings partners, and those partners bring others. Today, 23% of ConnectSmart users have carpooled at least once.


The circle is widening. In the network map below, each dot represents a carpooler and each line a carpool partnership. Two kinds of carpoolers stand out, and each plays a different role in the spread.


Socialites are the connectors. They ride with three to six or more partners and keep flexible schedules. They are also the platform’s best recruiters: 83% personally brought in three or more partners, and about one in ten brought in ten or more.


Super Novas are the anchors. They take 11 or more carpool trips a month with the same one or two partners. Half ride with the same partners every time; the other half rotate within a small circle of three to five. They are driven by cost and time savings, and they give the network its steady daily volume.


carpool hubs in Houston's network
Houston’s carpool network: each dot is a carpooler, each line a partnership. Hubs with many lines are Socialites; tight pairs and closed rings are Super Novas

Who people ride with shows the spillover from family outward. Household members account for 20% of carpool ties, coworkers 26%, and the wider social circle 62% (a carpooler can fit more than one group). Against Houston’s historical 75% family share, that is a striking shift. It means the platform is not just relabeling trips families were already taking together; it is creating new shared rides that would not have happened otherwise.

The trips also reach well beyond the 9-to-5. Among Socialites, 89% carpool for social outings and 83% for errands, compared with 67% for the work commute. Super Novas look similar: two-thirds share rides to social outings, half for kids and daycare runs, half for community events and the gym, and a third for work. The errands that once broke up the fixed commuter carpool are now part of what holds the new one together.


The habit holds. This fall, ConnectSmart cut rewards by more than half for one month, from about $0.70 to $0.30 per trip. Across all rewarded trips, activity stayed 4–8% above its trend for the first three weeks, dipped in weeks four and five, then reached record highs once rewards returned to about 85–90% of their earlier level. Rewards help start the behavior; routine and relationships keep it going.


It is cost-effective. Each dollar of ConnectSmart incentives removes about 6 vehicle miles of travel. A comparable app-based carpool incentive in the Seattle area removed about 2.5 miles per incentive dollar (Shen, Wang & Gifford, 2021), so ConnectSmart gets roughly 2.4 times as much from each dollar.


A Carpool Comeback Built on Trust


The old carpool asked commuters to share a car with a neighbor on a fixed schedule, five days a week. Modern life pulled that model apart: more cars, scattered jobs, busier households, looser schedules and, finally, a pandemic. The new carpool works with how people live now. It starts with family, spreads to friends and coworkers, covers errands and evenings as well as the office, and lets an app handle the coordination that used to make carpools fall apart.


For agencies and employers, the lessons are practical:


  • Start with the rides already happening. Recognize and reward family and friend trips instead of treating them as not counting.

  • Remove the friction. Matching, check-in, tracking, and rewards should take seconds, not a round of texts.

  • Find and support the connectors. A small group of Socialites recruits much of the network; give them easy ways to invite others.

  • Use incentives to start habits, then taper. Our test suggests routine carries trips for weeks once it forms.


Stranger matching still has a role. One in three ConnectSmart carpoolers say they are open to riding with someone outside their circle, and vetted friend-of-a-friend matching is a natural next step. But the fastest growth is coming from the people we already trust. The carpool comeback is upon us.


“Carpooling starts with someone we have some relationship with, and it lasts with someone we can count on.” — Carina Boston Pinales, Executive Director, Movability

Sources

Subscribe to Our Blog

Thanks for subscribing!

bottom of page