Special edition interview with Keith Hennessey — former Assistant to the President for Economic Policy
When it comes to AVs and the future of transportation, the media tends to focus on the tech or business sides of the industry. However, as the technology gets closer to market, policy and government will have enormous influence on how things unfold. After hearing from many CEOs and entrepreneurs, I was excited to speak to a policy expert. I was fortunate enough to sit down with Keith Hennessey and hear his perspective. Professor Hennessey is the former senior White House economic advisor to George W. Bush. As the Director of the White House National Economic Council, he helped the President on policy design, including the response to the ‘07-‘08 financial crisis. He teaches courses at the Stanford Graduate School of Business on Fiscal and Economic Policy and recently started teaching a new course, “The Open Road: Innovation in Cars, Driving, and Mobility”.
How do you think the current administration will affect the future of transportation / self-driving cars?
“I think there are two parts to this answer. First, I think about how a traditional Republican administration might think about this issue. Second, I ask myself if there is any reason that a Trump administration would think differently than a traditional Republican administration.
Per the first point, a traditional Republican administration would tend to have a lighter hand in regulation and at the same time be more concerned about the proliferation of State regulation. I would guess that the experience with California and emissions standards would incentivize a Republican administration to take a federal preemption approach. Having 50+ different regulatory regimes is something they would want to avoid. We already saw the Obama administration leaning toward the preemption but didn’t follow through because they were unsure they had the legal authority to do so. It wouldn’t surprise me if this administration would preempt but do so with a lighter hand.
Per the second point, there is no real reason to think that the Trump administration would behave differently from a traditional Republican administration. The DOT now has small government conservatives in its ranks. Elaine Chao, Jeffrey Rosen, and Derek Kan are all economic conservatives. I would expect the administration to be innovation- and transportation-friendly and defer to experts.”
How about with regard to impacts on the labor force and lost jobs?
“If we look at Trump’s rhetoric, it’s been focused on job loss due to trade and not really on automation. He hasn’t focused much on robots taking jobs away. However, we should keep in mind the broader policy around job loss. A lot of people are employed in driving (taxis, truck drivers, etc) but there is also a lot of turnover in the economy generally. Of course, changes in this industry will have a big effect on people working in the industry, but the change on a macro level won’t be that big relative to our workforce. Our economy is dynamic and people are changing jobs all the time. We need to remember that mobility has created a lot of jobs too.
When it comes to long-haul trucking, there are so many aspects: loading/unloading, maintenance, security, etc. I am not sure that all those jobs will be solved by an autonomous system. I can see a world in which the human no longer drives but is in the vehicle, loads and unloads cargo, and is responsible for security. Perhaps the human does other things — maybe they are doing two things, two different jobs, at once. Let’s not underestimate the ability for our economy and system to adapt.”
What is the biggest obstacle to AV adoption from a policy perspective?
“In my opinion, it’s activist local and state governments. If they start crafting individual legislation, this would require innovators to segment their products. You can imagine a scenario where each State insists on a different feature, such as a steering wheel or rear view mirrors. If there are too many different standards, how do you develop a product?
In particular, I would worry about policy makers trying to preemptively solve problems. The early stages of the World Wide Web (before it was called the Internet) were a complete Wild West. There was no regulation of any sort. Now, we have come to a place where we do have regulations in place on the Internet, related to privacy, data collection, cybersecurity, etc. The tremendous set of technologies during the World Wide Web era benefited from not having regulation even though some people getting were negatively affected during that period of time. We have seen that we need time to figure out what the problems are. I am convinced it is better for policy to catch up to technology than vice versa.”
What needs to be done now to ensure policy is right for when the technology is ready?
“In general, I believe it is better to react as opposed to preempt. We first need to see what bad things could happen and then figure out what should be done. In terms of the changes on labor, crime, and other effects, we don’t know what will happen yet. The American approach to policy is one where we allow technologies to flourish and evolve first before regulating and fixing issues. The European approach, relying on the precautionary principle, is the opposite. The European approach can stifle innovation and can be leveraged to protect from competition. I want innovation and experimentation and do not want vested interests protected from innovators. We have to let things happen and see what problems emerge — especially in this scenario where we have no idea what is going to happen. There are perfectly plausible hypotheses that may turn out completely wrong”.
What about policy surrounding safety?
“Safety is different. We have to be preemptive on safety. At an absolute minimum, people need to be convinced that the new vehicles on the road are at least as safe or safer than what they are used to.
I am guessing that there will be a lot of attention paid to accidents and deaths of autonomously-driven vehicles. How do policy makers and innovators think about policy in the transition period, where there is still a majority of human-driven vehicles? We are moving toward a future that is likely safer but in the meantime we will have new types of accidents, injuries, and deaths. People’s initial reactions will be to come up with rules and laws to correct problems — but that may slow progress towards getting to this better future.
At the same time, we need to be careful that safety is not used as a rationalization for other interests. A lot of European countries have done this with regards to GMOs when they are actually protecting their domestic agriculture. The challenge is how to distinguish between legitimate safety goals and safety being used as a Trojan horse for vested economic interests.
Another important thing to consider is what the policy framework will look like for AVs. Safety in automotive is under NHTSA while in aviation it is under the FAA. There are considerable differences between the two. Is an AV more like an airplane or a car when it comes to policy? It’s not clear which policy framework should be employed.”
What do you think are the most important negative effects of AVs?
“I would be careful to distinguish between negative effects and negative externalities. Congestion and pollution are not negative externalities particular to AVs. They are negative externalities of vehicles. It doesn’t matter whether a human or a robot is driving — the tailpipe is the source of emissions. Therefore, it’s important, when designing policies to address externalities, to address the source of the externality. We need to address vehicles, not the driver of vehicles.
The negative externality particular to autonomy is the potential for crime and illegal/bad behavior in those vehicles. Drug deals, terrorism, carjacking, getaway cars, human trafficking, etc. can all take place even easier in an AV. It’s hard to know what will happen but this is an area where bad people will use great technology to do bad things in new and innovative ways.”
How do you see protection of privacy playing into this?
“I have no idea. I don’t know if anyone does… There are not well worn pathways or models on how to think about privacy with respect to our mobile devices. We only have opt in and opt out at this point — that’s as sophisticated as it gets. So there’s no way we can know what will happen for transportation when it’s even more far out and complex.”
What stakeholders / groups do you think will be most obstructive to adoption of AVs?
“I worry most about the trial attorneys suing firms producing autonomous vehicles for the initial round of accidents and deaths. The question of legal liability when an AV crashes is critical. If the liability isn’t right, then you can get entrepreneurial trial lawyers taking down innovative companies.”
How do you think the dynamic between Federal and State regulation plays out on AVs?
“This is quite uncertain. It depends on many things, but is initially a legal question. Do the lawyers agree that NHTSA would need new legislative authority to be able to preempt. If so, then does administration propose the law or does someone on Capitol Hill propose it?
Legal questions aside, this strikes me as an issue where there is a significant opportunity for things not to become partisan. This does not strike me as an issue where there is a clear partisan line. I can imagine both parties trying to work something out, even though some small side-issues will need to be debated and agreed upon.
The Obama administration did not enact AV policy but sent a clear signal with their guidance.
It was not legally binding but gave a clear signal and provided States with a lot of information. Everybody listened and many were likely influenced. If Secretary Chao were to give a speech, every person in the industry would understand the direction the administration is intending to go and would plan accordingly. This hasn’t happened yet because the administration is still too new and they are not ready to make a big policy play.”
Are AVs likely to be deployed the areas with the most sympathetic policy, or the great possible business potential regardless of policy?
“I think it will be both. It’s most likely to be developed where there is an obvious business case and where regulators create a friendly environment. Arizona is a good example: Governor Ducey sent a very clear signal, with flat and dry policy. They have good cities there and good tech in the State and have attracted some good companies to work on their projects there. Nevada was out there very early as well but they don’t have the same economic engines. They tried to lean on policy but it’s hard to make case that companies should go there.
There is also a question for me around what happens to the big states. California is clearly well positioned with a huge population, big markets, and a center of technology and innovation but the policy is not encouraging. Texas and Florida are the other two that I think about. They are strong economic centers and have regulatory environments that are pro-business, pro-innovation, and pro-growth. I really think the Texans and Floridians could take a lead in this industry. Across all states, this is an opportunity for governors in particular to distinguish themselves. If they follow the Arizona model and at least get policy friendly environment, they could lay an attractive foundation for business to flourish.”
Which gov worldwide are the most aggressive in leading policy for AVs?
“Singapore is the clear leader. They are really forward-thinking. Beyond Singapore and the USA, there are six other countries I will be watching: Israel, Germany, the UK, Japan, Korea, and China. Israel is interesting because of how innovative they are. Germany, Japan, and Korea have big automotive manufacturers — the future of this industry will have a major impact on them and they have significant interests at stake. The UK is interesting because of their recent decision to leave Europe and because they have been innovative in the past in London with congestion pricing. They have been thoughtful about transportation policy in the past and I wouldn’t be surprised if they are in the future.”
How do you think cities will engage?
“It wouldn’t surprise me to see different cities taking different approaches. If I were working for a mayor, I am not sure that I would think only about autonomy. I would be thinking of transportation and mobility of people and stuff in and around my city. AV, parking, ridesharing, public transportation, ride-hailing, etc. are all relevant. It might not make sense to just focus on AV policy. I would want to focus on how all the innovations in mobility will affect demand for public transportation, infrastructure needs, and congestion on city streets. It needs to be a more holistic approach to figure out what policies should be changed. All these pieces are changing over same time frame and it’s really hard to see how they will converge and impact the overall picture. Uber and Lyft have already changed patterns in cities. With so many things happening at once, you cannot consider any one piece in isolation.
The second order effects are huge. We are basically getting at the question of how people and stuff move. When we boil it down even further it all it’s all about time and place: where you live, work, where you play, where you shop, and how you get to all these places. When those fundamental elements change, everything changes.
There’s one question that has been on my mind: how will congestion change with more autonomous vehicles on the road. The answer is far from clear. If you can’t answer that question, let alone know the magnitude of the change, you can’t know how to plan. Of course upfront policy may affect this outcome but we don’t know exactly how that will occur either. A lot of the outcomes depend on human behavior. If your car can drive itself, there might be more vehicles on the road (zombie vehicles). Suppose this means you can work on the drive to work. Are you more willing to spend more time in the vehicle because you can be productive? Does this mean you’re more or less willing to live farther away? There’s an effect on human preferences on how humans spend their time in the vehicle. Should the employer pay you for work in the car? There are so many questions that emerge from the changes in this industry.”
How do you think startups should engage with lawmakers?
“My recommendation for startups: spend your time and energy developing your product and trying to understand the current and future regulatory environments. Don’t waste your time trying to significantly influence the environment unless you have a lot of resources — it’s crazy for startups to try and influence legislation when it’s so difficult to do. You need to make a great product; the return on your time and money will be so much better. I don’t think it’s a good thing that big firms have an advantage but it’s reality. You have to watch out because some big firms will use their policy influence advantage to protect their business. I believe that if you’re a small innovative company involved in the future of transportation and spending a ton of money in Washington, you’re not doing it right.”
What do you think is the industry’s current biggest blind spot when it comes to policy?
“I am not sure but my concern with the new entrants in the automotive space (everything from Zoox to Tesla and Waymo) is not understanding that a lot of policy makers think incrementally to models they already understand. If you want to go to Washington and say you have a new way to think, it’s going to be hard to get them on board. If you can provide a comparison (the FAA policies for example) then they can think about it. A completely new way to think about it is going to be really hard. They want to apply old mindsets to new set of products.
Automotive safety up until now has been done incrementally. A major set of new rules took years if not a decade. When regulators have been focused on issues such as seat-belt tension, telling them that the pedals and steering wheel are gone is orders of magnitude more complicated. There are different speeds of thinking and changing between Silicon Valley and Washington. Changes in the tech industry are in weeks and months. In Washington, a year or two is a modest amount of time. We are going to see how two timescales bumping into each other.”
What do you think is a big misconception/inaccuracy about the industry right now?
“A lot of people seem to be convinced that increased autonomy will inevitably lead to reduced ownership of individual vehicles. I can understand the economics behind this and it is a reasonable argument, but i’m less certain of this outcome.”
