By on June 23, 2020

When Sir Thomas More coined the term “utopia,” he lifted two words from Ancient Greek that roughly translate into “not a place.” Turns out people from the 16th century still understood satire, perhaps better than we do today. After all, we are the ones operating under the assumption that we can remap society in order to build consequence-free transportation network without a shred of humor to keep us grounded.

We may not need satire in this instance, however. A new study published in the American Journal of Public Health asks questions about how just effectively the shift to autonomy will benefit society as a whole. Industry leaders have broadly framed the shift toward self-driving as kicking down the door to an idyllic universe where no one wants for transportation, with autonomous taxis serving as the first wave of this planned paradise. The reality may be vastly different that what’s being sold, however. 

The study essentially asserts that the entire concept of robotic cabs doesn’t actually serve poor communities any better than just buying one’s own automobile. Researchers compared the costs of a robo-taxi trip with those of owning a conventional used vehicle in an urban environment. Tabulating the combined costs of vehicle financing, licensing, insurance, routine maintenance, fuel/electricity and everything else they could account for, the team estimated that self-driving taxis would cost a minimum of $1.58 per mile. By contrast, the total cost associated with traditional vehicle ownership (assuming one is trying to be thrifty) ended up being 52 cents per mile. At least, that was the case for their model in San Francisco.

While your author has long suspected that unsupervised robotic taxis might outpace the subway as one of the dirtiest ways to get around (and become potential liabilities for whoever operates them), the general assumption has been that they’ll offer societal and health benefits that vastly outperform private vehicle ownership — almost as if the people making these assessments have never taken a regular cab or piloted an inner-city ZipCar. Other presumed benefits involve improved air quality by making it easier for people to get by without an automobile of their own.

But this thinking comes with some problems. Studies have already shown that ride-hailing firms exacerbate congestion by having a fleet of cars constantly scouring the streets in search of fares. That interim period between riders wastes energy and will be broadly similar when/if autonomous vehicles arrive. Why should we believe they’ll be any different when they’ll be similarly competing for riders and milling around neighborhoods? Even if they’re entirely electric, that energy has to be sourced from somewhere, and much of it will be in service of nothing.

Meanwhile, that thing you have parked in the garage isn’t hurting anybody until you start it up. Yet you can still rationalize autonomous taxis if you think a little outside the box. Consider how frequently poorer people find themselves skipping a chance to seek medical attention or taking a trip to do something that might better their overall situation. Computer-controlled cabs were supposed to help with that, as well. But the study is only getting warmed up with its bubble bursting.

“Even with universal health care, poor people are disproportionately less likely to access health care, because they can’t get there,” Ashley Nunes, one of the study’s authors, told Automotive News in an interview. “There’s been hope that this technology can be used to narrow the gap in health disparity. We find it can’t.”

Modern-day autonomous shuttles have proven themselves similarly problematic, including those trying to accomplish exactly what Nunes is talking about. Most of those efforts were supported by the cities in which they operate, propped up by government grants. Self-driving cabs probably won’t be, and that means companies may have to charge unpleasant prices for the service.

“The real promise of AVs is safe, affordable mobility on demand,” Nunes continued. “That’s the true promise. But is it safe? Safe for whom? Affordable for whom? That was the goal of this particular study. If we give poor people this shuttle, will it be OK? What’s equitable about pooling their rides? Nobody wants to pool a ride, let’s be upfront about that.”

The team actually has a secondary paper that asks these questions with highly similar conclusions. However, it’s less concerned with the social justice aspects of ride-sharing and more of the practicalities of keeping those programs widely available. Nunes and partner Kristen D. Hernandez suggest that capacity utilization rates for AVs would be extremely low at the price levels necessary to turn a profit. The only way to get ridership anywhere near the maximum (thereby wasting less energy and ensuring the public is better served) and make some money is to see the whole program subsidized by the government.

[Image: Sundry Photography/Shutterstock]

Get the latest TTAC e-Newsletter!

24 Comments on “Study Says Autonomous Taxis Will Cost Users More than Car Ownership...”

  • avatar

    I ll skip these stories and the BEV stories .

  • avatar

    More from the No Kidding, Sherlock files?

  • avatar

    Read the report and you figure out pretty quickly it’s flawed.

    “We used a bottom-up approach, using publicly available data (e.g., vehicle financing, licensing, insurance, maintenance,cleaning, fuel, and safety oversight expenditures) in a select city, namely San Francisco, California, to estimate autonomous taxi consumer costs (see the Appendix, available as a supplement to the online version of this article at We estimated costs at varying levels of technological maturity. We found that if commercial autonomous taxi services were offered today, fares would be — on a per mile basis — significantly costlier than continued ownership of older vehicles.”

    In other words, comparing the cost of a taxi service buying and running newer vehicles in San Francisco to me buying five to ten year old vehicles.

    • 0 avatar

      Since the average age of a vehicle on the road in the US is 11 years old that seems a fair comparison. My cost with a older car is about 48 cents a mile. My wife who drives only about 7500 miles a year has a fairly new car thats financed, with every thing (insurance reg, property tax payments maintenance) still only hits 1.20 a mile.

    • 0 avatar

      It is not flawed it perfectly shows that those people who claim AVs will bring affordable transportation to the masses is wrong.

      If anything they have grossly underestimated the cost of AVs because for some reason everyone seems to think these will be earning revenue all day long. If you only have enough vehicles to meet demand in slow times then you’ll have demand pricing for peak times that is significantly higher. Or you’ll have enough vehicles to meet peak demand and you’ll just have to charge more across the board to make those cars that only work 3-4 hrs per day.

    • 0 avatar

      So then, where would older cars come from? The study compares a company suffering the depreciation of their assets against a population with a pre-depreciated asset. It’s not like your 5-10 year old car will still be a 5-10 year old car ten years from now. It only works if you can find enough schmucks to buy a brand new car every five years and pass them on to us ‘smart’ people who are willing to invest more time in vehicle maintenance. Where’s the comparison between an autonomous taxi service and the new car buyer?

  • avatar

    The problem of having “self-driving/autonomous vehicles” is the age old economic one. After all the billions going in to the development of these things they are not going to be sold at a low price. Expect that they will cost at least double the MSRP of the typical hybrids used for Uber and Lyft now. And that is after a long time of gearing up production to minimize costs, and buying them in fleet quantities.
    Yeah, human drivers might be eliminated, but is that going to cover the cost of purchasing and maintaining the self driving vehicles? There will have to be specialized services to retrieve broken vehicles and repair them. None of the auto shops around now will have the capability to do anymore than replace tires and maybe brakes/suspension. All that drive train and autonomous controls will need a new group of techs. They are not going to work for cheap.
    Maybe in some long term, many decades after the vehicles are functional, there will be a breakeven point.
    Sure it might turn into another Tesla thing where people put lots of money into something with a possible long term payoff or a big stock market gain. We’ll see, I’m not going to hold my breath.

    • 0 avatar

      People adapt, well some do. Sure you had the mechanics that didn’t learn about EFI, ABS, ect when those technologies came along, but most did. Those guys eventually retired and the new younger technicians replaced them over time. Now try and find a guy that can tune the idle mixture on a carb with nothing more than a screw driver and their ear, or someone who knows what a dwell meter is, let alone know how to use one.

      That brings up a point from an article I read in an industry publication yesterday from a guy who is a little older than me, that pointed out something I had forgot about. Back in the early days of computer controls when everyone still had dwell meters they were used to test computer outputs like the duty cycle of a torque converter solenoid. There was a chart in the manuals where you would convert the (8cyl) dwell reading to a % duty cycle.

      So yeah there will be a lot of current techs that won’t bother but many in the early to middle of their career will continue and learn how to fix the newer technology just as those who went before them in the last 50 years.

      In fact the repair industry is already there since cars have had adaptive cruise control, lane keeping/centering and automatic emergency braking for several years now. So there are already technicians out there familiar with the procedures and tools to test and calibrate the types of sensors and actuators used in AVs.

    • 0 avatar
      Ol Shel

      True. Their goal is not to lower costs for users. The strategy of ‘disrupters’ is to destroy entire career categories, taking the wages that would have gone to many, and funneling it to a few tech ‘geniuses’ and their investors.

      They’re disrupting societal economic order, with absolutely no plan for the damage they are causing. Because tech bro, bro.

    • 0 avatar

      I would counter that these things will be insanely cheap and easy to service. For the most part, car buyers are emotional sheep. We want our cars to be an expression of ourselves. The effect of marketing and branding diminishes when the purchase is being made by an accountant who looks at reliability. If the car manufacturers themselves become ‘mobility services’ instead of car builders who pass the product on, you will begin to see cars built with serviceability and reliability in mind. No more parts changed every five years with one different bolt hole location, just to thwart the aftermarket vendors. No more crank bolts snapping because an engineer thought he could get a bonus for saving ten cents per bolt. “People will trash the interiors! It’ll cost a fortune!” Nah, look at subway interiors. Thick fiberglass with thin, washable pads. It’ll get power-washed every 24 hours or if a customer refuses it for ‘sanitary reasons’.

      • 0 avatar

        OK, Let’s take an example: Personal Computers. When they became available, 40 years ago, they were $2,000-5,000 in 2020 dollars. And those computers did not do much compared to today’s machines. Over time the cost came down. Now most people pay $400-500 for a computer unless they want or need the capabilities of something that is $1,500 or more. Of course many today don’t have a “computer”, they get along with a smart phone and or a tablet.
        But suppose the cost of a self-driving taxi in 2025 comes in at the same price as a Prius. That’s a pretty big ‘if’.
        That is still a large chunk of change to come up with, if you need thousands of them, either from investors or financed. I’m betting that the board of directors would be looking at the numbers and deciding if the potential revenue from riders will come close to the monthly expense. We’ll see.

  • avatar

    Shouldn’t they be compared to car leasership?

  • avatar

    Johnny Cab in Total Recall taught me what I need to know about autonomous taxis.

  • avatar

    “Turns out people from the 14th century still understood satire, perhaps better than we do today.”

    Quote of the Year (with correction to the 16th century).

  • avatar

    Shocker – not. These things will be expensive like four-wheeled fighter jets (at least at first), so the owners have to recoup the cost.

  • avatar

    SATIRE, n. An obsolete kind of literary composition in which the vices and follies of the author’s enemies were expounded with imperfect tenderness. In this country satire never had more than a sickly and uncertain existence, for the soul of it is wit, wherein we are dolefully deficient, the humor that we mistake for it, like all humor, being tolerant and sympathetic. Moreover, although Americans are “endowed by their Creator” with abundant vice and folly, it is not generally known that these are reprehensible qualities, wherefore the satirist is popularly regarded as a sour-spirited knave, and his ever victim’s outcry for codefendants evokes a national assent.

    -Ambrose Bierce

  • avatar

    I noted a couple statements whose assumptions are mostly false, as seen from personal experience.

    • Consider how frequently poorer people find themselves skipping a chance to seek medical attention or taking a trip to do something that might better their overall situation. Computer-controlled cabs were supposed to help with that, as well.
    — My late mother was by no means poor, living in her own home until the day she passed away. Yes, she still had har car, too. However, she refused to use any kind of taxi service, including ride-sharing, to go to medical appointments and if she couldn’t get a neighbor to drive her there in her own car, she simply skipped the appointment. What she really wanted–and hoped would be available before her passing–was a car that would drive her there by itself, giving her absolute freedom to go where she wanted, when she wanted, without needing to depend on anybody else. The autonomous car would have met that need but she would still have wanted to own her own than use some sort of ride-sharing apparatus.

    • One thing the article did point out is that ride-sharing in a form similar to Zip-Car will be far more a liability to the owning company than a benefit. Why? Because so much of our population has become a throw-away society–if it ain’t yours, trash it. Oh, maybe not in a literal sense for most of them but they’ll still not be extra cautious about how they use the vehicle as compared to one they own. Worse, in dense, urban environments, “trash it” could well be literal; turning them into rolling latrines and simply vandalizing the vehicles for the fun of it. Either way, such an autonomous vehicle is not likely to survive one year on the roads as compared to the eleven-plus years most cars achieve today.

    No, autonomous taxis are likely to be an overt failure unless and until those vehicles prove capable of resisting such vandalizing and actively deliver their passengers to a police location and charge said passengers with criminal complaints–including video of the crime or attempted crime. Even then, that won’t fully prevent vandalism to the exterior of the vehicle, which could also render it immobile and potentially irrecoverable at a complete loss to the owner.

  • avatar

    I own a 2002 eclipse spyder with close to 200,000 mi on it. The car runs great! I have owned eclipses from every generation. Bring back the fun to drive cars and sales will go up!

  • avatar

    Autonomous vehicles would be great for providing,on-demand mobility for the elderly, injured, or disabled as a superior and safer option than public transportation.

    For the rest of us? Not so much. The idea of summoning a driverless vehicle and waiting even 2 minutes for your hopefully right sized and potentially germ filled conveyance to arrive and take you where you need to go at the posted speed limit is more than many of us would tolerate on a day-by-day basis.

Read all comments

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • dal20402: “Best” implies that getting flung over the handlebars headfirst isn’t a design objective...
  • dal20402: I wasn’t talking about the LY7, but the LFX and LGX. There are a million Traverses and Impalas on the...
  • 28-Cars-Later: I feel the same way Ajla, GM products aren’t even in my consideration save a few and I’m...
  • 28-Cars-Later: No, they really don’t for a variety of reasons. But I have to agree with the overall point,...
  • ajla: I really don’t want to sound like an insufferable fanboy but GM doesn’t have anything that would...

New Car Research

Get a Free Dealer Quote

Who We Are

  • Adam Tonge
  • Bozi Tatarevic
  • Corey Lewis
  • Mark Baruth
  • Ronnie Schreiber