It's Not Just the Tech: Driverless Cars Inevitably Force Us to Make Political Choices About Regulation and Funding Different Types of Transportation
Will we ever reach the point where cars are fully autonomous (that is, driverless)? If so, how soon? At that point, will the American political system create -- and enforce -- safety regulations that are not simply blank checks for the tech/auto companies? And most importantly, when we pour money into driverless cars and the infrastructure that they will require, what are the tradeoffs? That is, what are the possibly better alternatives that we will leave behind?
Last Friday, I returned to the topic of driverless cars, about which I had previously written only twice (in 2016 and 2023). Having written Part 1 of "Now We'll See Whether the Democratic Establishment Truly Cares About Democracy" just the day before, I was burned out and thus chose to delay writing a very heavy Part 2 until Tuesday of this week, intending to write a relatively light Friday column on a relatively apolitical topic. So much for that. In closing that column, I conceded that "in its way today's discussion sets the table for the conclusion of my more overtly political 'Now We'll See Whether the Democratic Establishment Truly Cares About Democracy' column." Hence the title of that column: "About Those Driverless Cars and the Political Power Play Behind Them."
I suppose it is neither here nor there that this overlap was unplanned, but it certainly was, which I think makes my conclusion possibly more interesting, to wit: the Democratic Party's establishment players will almost certainly give away the store to the tech bros on driverless cars.
The fundamental reason for the overlap is that the safety of driverless cars is not truly a technological issue. How could it not be, one might reasonably ask? After all, driverless cars seems almost technologically inevitable, such that my statement last Friday "that there truly might be insuperable barriers that will make fully person-independent driving impossible" seems almost weird. What was I thinking?
In a trivial sense, I suppose that I might merely have been saying the equivalent of "allegedly" after having written earlier in that column that "[t]here is a very straightforward story in which driverless cars become feasible -- under assumptions that certainly do assume technological advancements, but not of the sort that are imagined with 'Terminator 2'-like melting and reshaping robots." It is still not a certainty, and until I see it, it is still in the category of hoverboards.
I might also point out that we have had "autopilots" in airplanes for decades -- including Otto! -- but so far as I know, we are nowhere near having pilot-less landings. Again, until I see it ... hoverboards.
But that narrow interpretation ultimately misses the point, because even today the technology exists to create a driverless car that would never kill anyone. Think about when automobiles themselves were first invented. I recall reading years ago that, when the first cars were introduced, some jurisdictions required every car to be preceded by people on foot waving red flags. I am too lazy to search for that anecdote, but even if it turns out that it was apocryphal (or a product of my faulty memory), the point is that cars with no bumpers, no collapsing steering columns, and no seat belts or air bags could be -- using then-available technology -- completely safe to the people both inside and outside of the cars.
This means that safety is ultimately not a matter of "getting the tech right and then getting out of the way," because no matter what technology exists, economic incentives will exist to push boundaries and take risks that would not be allowed under more careful regulation. I should add that I am not at all saying that I would have kept the rule requiring human automotive escorts, especially because such a rule made the car almost entirely useless -- except as a delivery vehicle, which not-fer-nuthin' would have been a great thing for animals.
But the point, again, is that the question is not whether the "safest tech" exists. We could, after all, allow driverless cars to operate everywhere today, limiting them to speeds of 10mph or less and requiring them to be swaddled in 24-inch-thick bubble wrap. That would be expensive, however, which brings us right back to the question of what public policy will allow in terms of costs versus safety. The current versions of these cars are surely losing money because the companies are loading them up with features that they will soon want to abandon.
For what it might be worth, however, the evidence on the safety improvements in current-generation driverless cars is arguably encouraging. The Insurance Institute for Highway Safety (IIHS), for example, recently reported that, "[p]er vehicle mile traveled (VMT), Waymo’s driverless vehicles deployed in San Francisco, Phoenix, Los Angeles and Austin were involved in 68% fewer crashes than human drivers, researchers found." There are good reasons to be skeptical about how honest or meaningful those numbers are, but even if they are on the up-and-up, the IIHS added: "'The results show that, on a limited scale, these driverless cars are safer than human drivers — who can be impaired or drowsy or suffer lapses in attention,' IIHS President David Harkey said. 'However, the present data collection system isn't good enough to allow continuous monitoring of a large-scale expansion.'"
In any event, the question again is about the interaction between any particular tech, the related economic incentives, and the regulations that are put in place to determine how big the risks can be. At a minimum, we ought to be able to expect that the current environment, in which the public is still skeptical/uncomfortable about the very idea of driverless cars, would be the high-water mark of careful regulation -- and for that matter, self-regulation by companies that surely know how bad the optics of "driverless car kills dozens" would be. Right?
Not so much, it turns out. A San Francisco-based writer at Business Insider, Lloyd Lee, began a piece that ran last week ("Robotaxis are scaling now, fixing problems later") by noting that he has ridden in Waymos and felt reasonably safe (but still vigilant), but he then went on to argue that the tech companies are succeeding in pushing ahead in the US and the UK in ways that even their own data do not seem to support:
In Florida and Texas, Tesla added new service areas for its Robotaxi. In some regions, the company has started offering a limited number of unsupervised rides.
Robotaxi companies are also expanding beyond the US. Uber and UK-based Wayve said on Wednesday that they received regulatory approval to start offering supervised autonomous rides in London this summer. Waymo also plans to launch a commercial service in London by the end of the year.
Lee immediately added that "[t]he rapid pace of scaling can give the impression that robotaxi companies are confident that autonomy has largely been solved." Is it?
The problems have been persistent enough for SF's pro-tech mayor, Daniel Lurie, to call for stricter statewide regulations last month.
Those calls have extended to the federal level, where regulators from the National Highway Traffic Safety Administration warned of the pitfalls autonomous vehicles still face in active emergency scenes.
NHTSA Administrator Jonathan Morrison wrote in a July 8 letter to AV developers that the federal agency has documented a "clear pattern of driverless AVs interfering with law enforcement and other first responders."
"The question is, as they expand, can they fix [the problems] as fast as they find them?" Koopman said. "And with floods and construction zones and a bunch of other things, the answer is they haven't been able to fix them."
And this is where the ugly politics of the Democratic Party's establishment become the most worrisome part of the story. In my "Now We'll See Whether the Democratic Establishment Truly Cares About Democracy" piece, I focused on Third Way, which is the most powerful group of connected political insiders who constantly push the party to reject anything progressive, from labor rights to environmental limits to consumer safety, being especially enraged right now about some Democratic candidates' support for Medicare for All.
Why do these nominal Democrats reject all of that? It turns out that they are called "the corporate wing" of the party for a reason. Why run the risk of annoying your richest backers by regulating them? Last week, a writer named Thom Hartmann wrote an excellent short history on Third Way and its Clinton-era predecessors, "Who Is the 'Third Way' and Why Are They Fighting Progressives Instead of Republicans?" It is very much worth reading in full, but this passage captures the key point here:
Describing themselves as committed to "democratic capitalism," they explicitly recommend Democratic politicians maintain relationships with wealthy donors, use capitalism as the framework for Democratic governance, create "partnerships" with business, and push "market-oriented solutions."
...
The simple reality that Third Way understands is that on-the-take Republicans on the Supreme Court have — in 5-4 all-Republican-appointee decisions — legalized a corrupt oligarchic system where massive corporations, foreign governments, and morbidly rich individuals can essentially buy elections for their favored candidates.
....
Instead of joining progressives in working to tear it down, they're choosing to work to keep the corporate-owned Democrats in power and keep those unwilling to take PAC money out.
In the context of driverless cars, as in every other area, "partnerships" with business and "market-oriented solutions" will amount to blocking any attempt to regulate business meaningfully and, if that fails, to water the regulations down and under-enforce them. As a related example, the current political battle over regulating social media reportedly resulted in an interim bipartisan agreement that "does not include a 'duty of care' provision, [which] would require companies to
design social media platforms with children's safety in mind." As always, the Democrats overall have a better record on this, but the pressure not to regulate the party's donors is intense, and the establishment always seems to take the side of businesses donors.
Finally, there is an entirely different set of political decisions that must be made when dealing with driverless cars, involving what economists call opportunity costs. I quoted above part of a sentence from my column last Friday regarding the tech issues that need to be worked out, but the full sentence is relevant here:
There is a very straightforward story in which driverless cars become feasible -- under assumptions that certainly do assume technological advancements, but not of the sort that are imagined with "Terminator 2"-like melting and reshaping robots -- after which it could be a relatively straightforward matter of redeploying federal transportation funding to allow driverless cars to interact with street signs and tech embedded in roadways.
What I airily described there as "a relatively straightforward matter" is, however, the whole enchilada. When I wrote those words, I was imagining having federal highway funds taken away from building more roads (and adding more lanes) and instead appropriating the money toward making existing highway infrastructure tech-friendly in ways that could enhance safety. That, however, is almost certainly hopelessly naive.
After all, one of the most politically useful arguments for funding public transportation is that it is very, very safe. It is also much better when it comes to land-use, pollution, resource depletion, and so on. But telling people that "cars are now 100 percent safe" -- even if we were also to use technology and resources to make public transportation equally safe (which I doubt the politicians would approve) -- will only make the demand for private transport rise. Indeed, the tech companies have been claiming that driverless cars will clear out space on urban streets by eliminating parking spaces and shortening following distances, but at least so far, the cities that have approved wide adoption of driverless cars have seen more traffic, not less.
The larger point, however, is that even safe cars are wasteful. Unfortunately, Republicans hate public transportation, and corporate Democrats will invoke their favorite "fiscal responsibility" dodge to say that we should not push for more public support of even non-wasteful transportation within and between cities.
This, in turn, means that it makes no sense to limit ourselves to thinking only about an idyllic world in which cars are suddenly safe. Even if that were possible, there are good reasons to suspect that the laws and regulations that make them safe would be attacked and diluted. Meanwhile, some guy once famously said that there is no free lunch. Every dollar spent on one thing cannot be spent on anything else, which means that it is not a matter of "driverless cars or not" but "driverless cars instead of what?" On both regulation and spending, the Democratic Party's establishment is unlikely to be on the right side of the debate.
- Neil H. Buchanan