By Futurist Thomas Frey

The Largest Job Loss in American History

In 1900, 40% of American workers were farmers. By 2000, less than 2% were. That transition took a century and still caused massive social disruption—farms failed, rural towns collapsed, generations struggled to adapt.

We’re about to do something similar in 15 years, not 100.

By my count, approximately 5 million Americans make their living directly from driving or managing drivers. Not building cars or selling cars—actually driving them or enforcing rules about driving.

Here’s the rough breakdown:

• 3.5 million truck drivers
• 500,000+ taxi, Uber, and Lyft drivers
• 300,000+ bus drivers (school, transit, tour)
• 100,000+ delivery drivers (though this overlaps with trucking)
• Tens of thousands of traffic police
• Tens of thousands of parking enforcement officers
• Unknown thousands in DMV operations, driving instruction, traffic courts

These are real jobs. Middle-class jobs. Jobs that support families, pay mortgages, send kids to college.

Between 2030 and 2045, most of these jobs disappear. Not because the work isn’t valuable—it is. But because autonomous vehicles do it better, safer, and far cheaper.

This isn’t speculation. It’s math. And the math is brutal.

Why Every Driving Job Becomes Obsolete

Let’s start with why the technology wins so decisively.

A human truck driver costs roughly $70,000 annually in wages and benefits. They can legally drive 11 hours per day maximum, requiring rest breaks and overnight stops. They make errors—missed exits, poor routing, sometimes crashes. They get tired, distracted, sick.

An autonomous truck costs maybe $10,000 annually in operating costs beyond the vehicle itself—software subscriptions, remote monitoring, maintenance. It can operate 24 hours per day with charging breaks. It follows optimal routes exactly. It doesn’t get tired or distracted. It crashes far less frequently.

Even if autonomous trucks cost twice as much to purchase upfront, the operating cost advantage is so overwhelming that trucking companies will switch as fast as regulations allow.

The same math applies to every driving job. Once the technology works reliably—and it will by the late 2020s—economics forces adoption regardless of social impact.

The Timeline of Job Loss

Not all driving jobs disappear simultaneously. The sequence matters because it determines which workers get hit first.

2028-2033: Urban taxi and ride-share drivers (500,000+ jobs)

This happens first because urban environments are where AVs work best initially—shorter trips, lower speeds, better mapping, more support infrastructure.

Uber and Lyft won’t wait. They’re already testing autonomous vehicles. The moment AVs are approved for passenger service in cities, they deploy them. Why pay human drivers when AVs work 24/7 without wages?

Human ride-share drivers see their income collapse practically overnight. First, trip volume drops—AVs take the most profitable rides. Then, per-ride pay drops—companies lower prices to compete. Then, insurance becomes unaffordable as human drivers are statistically more dangerous than AVs.

By 2033, human ride-share driving is essentially dead in major cities. Some drivers transition to other work. Many don’t.

2033-2038: Local delivery drivers (100,000+ jobs)

Last-mile delivery—packages to your door—is harder to automate than passenger transport because someone needs to carry packages to doorsteps. But this gets solved through combination: AVs drive to neighborhoods, then either robots carry packages or humans meet the AV to grab packages for final delivery.

Amazon, FedEx, UPS switch to AV fleets as fast as vehicles become available. The cost savings are too large to ignore. Human delivery drivers become assistants to AVs, then get eliminated entirely as robot delivery solves the last-mile problem.

2035-2045: Long-haul truck drivers (3.5 million jobs)

This is the big one. Trucking is the largest driver category by far. It’s also the most resistant to automation—not because the technology is harder, but because the political opposition is fierce.

3.5 million truck drivers are overwhelmingly working-class males, ages 35-60, concentrated in politically important Midwest and Southern states. They vote. Their communities depend on them. Politicians listen.

But economics wins eventually. Shipping companies that use AVs have 40-50% lower costs than companies using human drivers. Companies that don’t automate can’t compete. They go bankrupt or automate.

Peak job loss happens 2038-2043. Over five years, 2+ million truckers lose their jobs. Small trucking companies fail. Independent owner-operators can’t compete. Only large fleets survive, and they’re fully automated.

The social and political disruption is massive. But it happens anyway because the economics are inexorable.

2030-2045: Traffic enforcement gradually eliminated (100,000+ jobs)

This happens slower because it’s government jobs, not private sector, but it’s equally inevitable.

When human drivers are rare, traffic police become unnecessary. No speeding tickets—AVs obey speed limits. No DUIs—AVs don’t drink. No reckless driving citations. No running red lights. No expired registration stops.

What’s left? Investigating crashes involving the remaining human drivers. Responding to disabled vehicles. Some specialized law enforcement.

Cities and states quietly reduce traffic police through attrition—retiring officers aren’t replaced. Traffic court systems shrink. DMV offices close or consolidate. Parking enforcement becomes pointless when parking disappears.

Secondary Job Losses

The direct job losses are just the beginning. Entire industries built around human driving collapse.

Auto insurance (300,000+ jobs): When human-driven cars become rare, personal auto insurance nearly disappears. Liability shifts to manufacturers. Claims adjusters, underwriters, agents—most become unnecessary. Some insurers adapt by insuring AV fleets. Most shrink dramatically or exit the business.

Personal injury law (tens of thousands of jobs): Traffic accidents are the bread and butter of personal injury lawyers. Accidents drop 90%+. So does the litigation. Some lawyers pivot to product liability cases against AV manufacturers. Many leave the field entirely.

Driving schools (50,000+ jobs): When learning to drive becomes optional, driving instructors lose students. Some states might still require driver’s licenses as backup skills. Most people won’t bother learning. The industry collapses.

DMV operations (tens of thousands of jobs): Fewer drivers means fewer licenses, fewer registrations, fewer tests, fewer renewals. DMV offices close. Staff get laid off.

Traffic court personnel (thousands of jobs): No traffic violations means no traffic court. Judges, clerks, prosecutors handling traffic cases need new jobs.

Roadside motels (unknown thousands of jobs): Long-haul truckers stop overnight at roadside motels. When trucks drive 24/7 without stopping, those motels lose their customer base. Many close. Hospitality workers in truck-stop towns lose jobs.

Add it all up: the 5 million direct driving jobs likely trigger another 1-2 million secondary job losses. Total: 6-7 million Americans lose employment because of autonomous vehicles.

Who Gets Hit Hardest

Job loss isn’t evenly distributed. Certain demographics get devastated.

Working-class males, ages 35-60: Trucking is overwhelmingly male (95%+) and middle-aged. These workers have limited formal education—many didn’t finish college. Their primary skill is driving, which becomes obsolete.

Retraining a 50-year-old trucker with a high school diploma to become a software developer is unrealistic. Historical data shows older workers with limited education struggle to transition to new careers.

Geographically concentrated: Trucking employment is concentrated in Midwest and Southern states—Missouri, Arkansas, Oklahoma, Kentucky, Indiana. These regions have limited alternative employment. When trucking disappears, entire communities lack jobs.

High income loss: Median trucker income is $47,000 annually—solidly middle-class. Losing that income drops families into poverty or near-poverty. Retail spending in affected communities collapses.

Compare this to manufacturing job losses from automation and globalization. Similar demographics, similar geographic concentration, similar political anger.

Except this is larger and faster.

Economic Impact on Communities

What happens to a town where 20% of the workforce drives trucks?

Look at coal mining towns for the answer. When the industry collapsed, towns collapsed. Young people left. Property values crashed. Retail businesses closed. Tax revenues disappeared. Schools and services deteriorated.

Multiply that across trucking-dependent towns throughout the Midwest and South.

Places like Joplin, Missouri, or Fort Smith, Arkansas—where trucking is a major employer—face devastation. These communities don’t have diverse economies that can absorb thousands of unemployed workers.

The opioid crisis, the deaths of despair, the political anger we saw from deindustrialization—expect all of that again, but larger.

The Retraining Fantasy

Every discussion of automation-driven job loss includes optimistic talk about retraining programs. “These workers will learn new skills and transition to better jobs!”

History suggests otherwise. Retraining programs sound good. They mostly fail.

Trade Adjustment Assistance programs for manufacturing workers showed limited success. Most participants either didn’t complete training, or completed training but didn’t find employment in their new field, or found employment at significantly lower wages.

Why do retraining programs fail?

Age matters: Learning new skills gets harder as you age. A 50-year-old who’s driven trucks for 25 years can’t easily become a coder or nurse.

Geography matters: Retraining for jobs that don’t exist locally is pointless. If your town’s economy is based on trucking, retraining to become a software engineer is useless when there are no software jobs nearby.

Motivation matters: Workers displaced by automation are often demoralized. They built careers in fields that disappeared. Starting over at 45 or 55 is psychologically difficult.

Economics matter: Retraining takes time. During that time, bills don’t stop. Families can’t afford to live on hope and retraining stipends.

Retraining will help some people—especially younger workers with education and mobility. It won’t help most middle-aged truckers in rural communities.

Political Responses (And Why They’ll Fail)

5 million unemployed drivers become a powerful political force. Expect attempts to slow or stop autonomous vehicles through regulation.

Delay tactics that will be tried:

• Require human drivers in AVs as backup (defeats the cost advantage)
• Mandate licensing fees so high they neutralize savings (gets challenged as protectionism)
• Ban AVs on certain roads or in certain states (creates patchwork that’s unworkable)
• Require union labor for AV fleet operations (gets challenged, probably fails)
• Implement robot taxes to fund displaced workers (gains traction but insufficient to solve the scale of job loss)

Some states—especially those with powerful trucking unions and concentrated employment—will resist hard. They’ll pass laws protecting driver jobs.

But economics forces change. Companies operating in states that allow AVs have huge cost advantages. Products shipped by AV fleets cost less. Companies that automate thrive. Companies that don’t, fail.

Eventually, even resistant states give in. The competitive pressure is too strong.

What Actually Helps (A Little)

If retraining fails and political resistance fails, what actually helps displaced workers?

Universal Basic Income: Direct cash payments cushion the blow. Won’t replace middle-class incomes, but prevents destitution. Politically controversial but might become necessary when millions need support.

Age-based protections: Some proposals suggest grandfathering older workers—anyone over 55 keeps their job until retirement. This spreads the transition over time but delays inevitable adjustment.

Geographic relocation assistance: Help people move from trucking-dependent towns to cities with diverse economies. Historically unpopular (people don’t want to leave home), but sometimes necessary.

Early retirement incentives: Help workers aged 55-65 exit workforce early with dignified support. Reduces labor force participation but prevents poverty.

Public employment programs: Government creates jobs for displaced workers—infrastructure, care work, community services. Expensive but effective at maintaining income and dignity.

None of these fully solves the problem. They just make it less catastrophic.

The 2045 Landscape

By 2045, the transition is complete. Human professional drivers are rare—mostly specialty operations where automation doesn’t work yet.

The 5 million people who lost driving jobs are scattered across outcomes:

• Some found new careers (younger, educated workers)
• Some retired early (older workers with support)
• Some live on UBI or disability (those who couldn’t transition)
• Some moved to cities with better opportunities (those with mobility and resources)
• Some never recovered economically (those without options)

The communities that depended on driving employment are hollowed out. Some adapted by attracting new industries. Most declined.

The political anger from this disruption reshaped American politics for a generation—similar to how manufacturing job losses reshaped politics in the 2010s.

But it’s done. The jobs are gone. The efficiency gains are enormous. Shipping costs dropped 40%. Product prices fell. The economy overall grew.

Just not equally for everyone.

Winners and Losers

Winners:

• Tech companies building AV systems
• Logistics companies using AVs (lower costs, higher profits)
• Consumers (cheaper goods from lower shipping costs)
• Younger workers who adapt to new economy
• Companies that automated early

Losers:

• 5+ million drivers who lost jobs
• Their families and communities
• Small trucking companies that couldn’t compete
• Workers too old or immobile to transition
• Towns built around trucking industry

The wealth created by automation is real and massive. The pain caused by automation is also real and massive.

The question isn’t whether AVs eliminate these jobs—they will. The question is whether we help the people who lose them, or let them suffer alone while the benefits flow elsewhere.

Based on history, the answer is probably: we let them suffer. Then we act surprised when they’re angry.

But their anger won’t stop the technology. It just determines how painful the transition is.

Next column: what happens to the kids and elderly who gain freedom from this same technology.

Related Articles:

The Future of Trucking Employment – Bureau of Labor Statistics analysis

Retraining Programs: What Works and What Doesn’t – Brookings Institution research

Economic Impact of Automation on Communities – NBER study on job displacement