#33: Cost of car ownership



David Keith, assistant professor, System Dynamics Group, MIT Sloan School of Management

Joanna Moody, research program manager, Mobility Systems Center



We’re trying to understand what people value about cars so that we can understand how to build that in to alternatives to private car ownership and use, and how we might then encourage people to switch to a service if we can provide that same level of service, or at least a similar or competitive level of service with other alternatives.

Joanna Moody: Hello, my name is Joanna Moody. I am the research program manager for the MIT Energy Initiative Mobility Systems Center.

David Keith: Hi, and I’m David Keith. I’m a professor in the System Dynamics Group at the MIT Sloan School of Management.

JM: Thank you, David, for joining me here today on the podcast. I’m really excited to be here with you on the podcast to talk about U.S. car ownership and why or why not the U.S. may have reached “peak car”.

Why are we talking about car ownership and use in the United States? I want to just give a couple of high-level statistics to motivate why we might talk about this topic. The transportation sector is now the leading contributor to U.S. carbon emissions, global emissions, passing the energy sector in 2018. And 58% of the transport sector emissions come from cars: the small sedans up to the trucks that individual consumers buy and use on a day-to-day basis. Car ownership and car use is very persistent in the U.S. but there’s been a lot of hope recently. Some positive momentum towards some demographic/generational changes, or even technological changes, that people are hoping might debunk this persistent car ownership and use.

David, can you tell us a little bit about some of those demographic, generational, and also technological changes that people are looking towards to maybe change the dependence that the U.S. has on the car?

DK: Thanks, Joanna. Those societal impacts are increasingly clear. The environmental impact of driving, in terms of greenhouse gas emissions and urban air pollution, but also just the traffic congestion that’s in our cities and [has] really come to dominate many cities around the world. But at the individual level as well, it’s intriguing that we have so many cars. On one hand, cars are really useful to us. They’re enormously flexible and convenient, but they’re expensive and they sit idle most of the time (and rough estimates suggest most cars sit idle about 90% of the time). This question of this paradox that you and I have been interested in is, why is it that on one hand, we have so many cars if they’re expensive and inconvenient and bad for the environment and all of these things.

As you said, there’s secular changes happening in the automotive industry. We know there’s electric vehicles coming to market over the last 20 years. There’s self-driving cars, there’s on-demand mobility platforms. All of these are interesting and influence how we can own and use cars into the future but let’s start with demographics, as you said. One thing is that as generational change occurs that our relationship to car ownership might be changing as well. All the generations have coming of age experiences about rebuilding cars and getting their driver’s license the day they become eligible. Yet younger generationals—millennials—don’t seem to have that same attachment, that having a smartphone that I can contact my friends and hail an Uber is as or more important than having the lump of metal and the driver’s license to drive myself around. So that Millennial’s effect is one thing that’s been front of mind.

There’s some great research by colleagues of ours, Chris Knittel and Elizabeth Murphy here at MIT, who sought to unpack that millennial’s effect. They haven’t found that that effect is real despite the headlines that we see around that. Yes, fewer people are rushing out and getting their driver’s license, but what that research shows is that there’s interesting things happening: that when you’re in your 20s and you’re living in a city and a student or a young professional, yes, you’re less likely to have a car than you might have been previously. But when that generation or those age brackets get a little older and they get married and they move to the suburbs, it turns out that those people are just as car-dependent as previous generations have been. I think one thing that you and I have been really interested in is about the impact of urban form as—the suburbs versus the inner urban areas and rural areas—as to how that conditions the mobility choices that we make.

Another trend that I think we don’t notice in the short term, but it’s going to be really important in the long term, is population growth. The United States is growing at something like 1% per year, which over time when we look out to 2050 or 2060, that leads us to a population of 400 million people and ever more mobility being demanded and ever more cars on our roads. All the indications we have today are that people are demanding just as much driving or just as much mobility as they have previously. Indeed, whilst there’s some forces that suggest that could go down—the “peak car” arguments—there are forces that are pushing upwards on our demand for car ownership and use also. So it’s really not as clear as we might initially think about the direction that this industry is going to go into the future.

JM: One of the areas that people are particularly interested in, in the transportation space, is can we translate the car trips that are being made—or that will be made as populations grow and demand for mobility grows—with a different paradigm than the traditional car ownership, private car model. That’s looking at these new on-demand mobility services, or as we might call them “ride hailing”, where I open up my phone and I say, “I need to make this trip.” I give you my origin, where I am now, and where I want to go, my destination, and you match me with a vehicle that takes me there when I need it. Those systems or services exist now, as Uber or as Lyft or other platforms globally. And there’s hope that they may be even more responsive as autonomous vehicles enter the market and become part of these fleets.

There’s this hope that this would enable more multimodal lifestyles, in the sense that some of my trips may continue to be car-based, but I’d have them on demand, but otherwise, I would walk, I would bike. It might connect me to public transit and then I would take a train or a bus, and that this might dismantle this sort of dependence on the private automobile, in particular, in cities. This was one of the motivations behind our study, was looking at can on-demand, car-based services really compete with the private car ownership and use in the U.S.?

We looked specifically at the value that car ownership and car use bring to individuals. One of the reasons that we differentiate between car ownership and car use is that most people talk about these services as, “I’m needing a trip. I need to go somewhere that I would otherwise take my private car, but instead, I call this car on a mobility demand platform.” That is, sort of, a one-to-one replacement. But what about that car ownership aspect? Can you tell the listeners here a little bit more about what is the value of owning a car even when you’re not using it? What makes up that ownership value?

DK: That’s a great question. The hope with these mobility services is that we can still enjoy the same amount of mobility, and indeed, automobility, but that we can do it with fewer cars. Instead of driving a car for one or two hours a day that I own privately, if the same car drove me to work and then it came and picked you up and took you to the supermarket or out for dinner, that  car might be able to drive 10 hours a day, for example, which would be a 5x increase in utilization. At a first-order approximation would mean we need one-fifth the number of cars in cities. That sounds great. Even bigger opportunities if we start to think about pooling, for example, where multiple riders sit in the same vehicle.

And so from an efficiency or resource use perspective, that all sounds fantastic, but it doesn’t place as much emphasis on why it is that we own cars in the first place. We live in a world where—particularly in the United States, but in many other countries, including Australia, where I’m from—that we have a nearly one-to-one correspondence between car ownership and adult age or driving age individuals in our countries. That means that wherever I am, there’s either a car sitting outside my house, or if I’m not at my house somewhere else potentially that my own private car. What does owning that 2,000 kilograms of metal, what does that afford me that I don’t get if I use Uber or Zipcar or some other car sharing? One thing that we’ve studied at length and Joanna, you’re an expert at this, is the status that I derive from owning that car. It says something about who I am, whether it’s a sports car or a luxury car or a rugged Jeep, that reflects something about my personality.

Another thing that you and I have been thinking about recently is reframing this low utilization. Yes, we’re only driving cars 10% of the time, one or two hours a day. But as a friend of ours, Nick Pudar from GM, is fond of saying, “The rest of that time, the car’s sitting there ready to be driven at a moment’s notice.” If it’s the day before Thanksgiving and I want to drive and see my family, or I want to go out and pick up some ice cream at night, or there’s a hurricane coming and I want the possibility of moving to higher ground. I don’t need to wait for a car or hope that that car is going to come and get me and make sure that my phone’s charged so that I can order that ride. I’ve got that car there. I’ve got that option value or that flexibility.

When we were thinking about designing this research that we’ve done recently, there’s the cost side of the equation and much of the existing literature is focused on the cost of car ownership. But we really wanted to flesh out and quantify the value of owning the car. Not just the use-value of getting me from A to B (certainly that’s very important), but also some of this non-use value—that the flexibility and the status and the symbolism and all the other more tacit things that we might derive from having our own car at hand.

JM: Yes. This is really motivated by the fact that there’s been a lot of headline news about how expensive car ownership is to the average U.S. household. And AAA estimates or Edmunds and other groups that does this has estimated that it’s around $9,200 or $9,300 a year for the average U.S. household to own and use a vehicle, the total cost of that vehicle. That is very expensive, but also 91% of U.S. households make this purchase, and own a car or own at least one car.

There’s this real question, this begging question of, okay, it’s really expensive, but most people still choose to make that purchase. And so our hypothesis was that the car provides value greater than that cost. That yes, it’s expensive, but the flexibility, the convenience from the ownership, and then also the mobility that its use provides will offset or at least equal that cost.

We actually went about surveying in the U.S., and we chose four cities: Chicago, Washington D.C., Seattle, and Dallas, Texas in the United States. We asked individuals a series of survey questions to be able to quantify the value that they give to car ownership and use. Then we also were able to use some scenarios around these questions to actually isolate the value of ownership—capturing this idea of option value or the fact that the car is there whenever I need it, that status—and then also the use-value. That use-value that might be able to be replaced by an on-demand service or even other modes of travel: walking, biking, public transit.

We did this by asking individuals a series of two choices. You, as a survey taker, get a question and it says, “Would you give up your car for a year and receive an amount of compensation,” and we can vary that level of compensation that we show the respondents. Let’s say, “Would you give up your car for a year and receive $20,000?” or, “Would you keep your car and forgo that compensation?” By varying that amount, we can actually estimate across individuals what the average value of car ownership and use is.

Then to be able to understand the value of ownership separately from use. We gave individuals another scenario, we said, “Now, we’re not asking you for your current travel behavior. We’re saying, imagine a world in which you have a ubiquitous, completely free ride hailing service that can replace all of your current trips that you make by your private car.” We’re going to take all of the trips that you make by private car, and we’re going to say, “You can make them without any additional inconvenience by just calling another car on a fleet.” So we’ve taken use out of the equation, and then we ask, “Okay, now how much would we need to compensate you to give up your car for a year?” We vary that amount, and we can get an estimate of what is that value of the ownership, even after we’ve replaced all of the use of that car, and it’s really just sitting there for you in your driveway and you’re not using it.

Walk us through a little bit of the high-level findings from this survey and what they mean. Do we confirm our hypothesis that the value of car ownership and use is greater than the cost?

DK: Once we did all those online choice experiments, the key result we came to was that at an average across the population that we sampled from, we found a total value of car ownership and use of over $11,200 a year. So, compared to estimates of the total cost, or the true cost of owning a car, in the order of a little more than $9,000. Our findings suggest that what’s happening, or how we might explain why so many people choose to own a car, is that it’s rational for them to do that. That they’re enjoying more benefit [i.e., value] from having that car than the cost of owning it.

Now, that might sound kind of obvious. Like, yes, we all enjoy cars and if it’s got to be worthwhile for me to do it, but that actually flies counter to most of the existing literature on this topic to date. The dominant theory to date about why we have so many cars was actually something different. It was related to cost, which is that people systematically underestimate the cost of car ownership. Because on a day-to-day basis, I put gas in my car every couple of weeks, and there are some other costs, but that many of them are less salient to us. Whether it’s insurance or maintenance or changing the tires or some of these things that only happen infrequently. That when we think about the cost of taking a trip in a car on a day-to-day basis, the marginal cost is not that great. That may well be the case and our findings don’t actually nullify that at all. It may well be that people underestimate the cost of car ownership, but we really want to emphasize that our indications are that people are really deriving a lot of value from owning and using that car. And that that should really reframe how we think about the issue of car dependence more broadly, but also what characteristics we are going to need to design into systems that are going to be a viable alternative to widespread car use. We’ve deliberately used these phrases—car ownership and use—because I think one of the most interesting things once we unpack that result is that all of that $11,000, just over half of that value we estimate as being the value of owning the car itself, not using it. Even at an order of magnitude level, we get just as much value out of having the car there and having the possibility of driving it as we do from actually getting from A to B itself.

This really challenges us to think about what the attributes of an effective urban mobility system are. Yes, there are things we wanted at society level, that it’s non-polluting and maybe it doesn’t make too much noise and different things like that. But what we’re seeing in this work is that from the consumer perspective, car ownership provides a lot of this flexibility. Flexibility over my schedule, about being in my own space at the moment. We have this other layer over the top, which is that we’ve got the Covid pandemic to factor in as well, and our results do touch on that. So, Joanna, I’m going to have you tell us about that in just a second, but it really speaks to some of the behavioral explanations for why we are so attached to car ownership—something that’s evolved in our country and in the Western world over the last hundred years. That some of these qualities are going to be actually quite difficult to design into alternatives like on-demand mobility systems.

JM: I think that I’m hearing two main takeaway points from you. One of those is that the value of car ownership and use total, the value that the private car provides far outweighs the costs, even if individuals are underestimating the true cost of car ownership. Yes, that may be true, but they’re still making rational decisions about owning and using a car because the value they derive from it—the benefits they derive from car ownership and use—outweigh even the “true” estimates of the cost of car ownership and use.

Then the second piece of this is breaking that value down by ownership and use separately. To put some numbers to that 50/50, we’re looking at $5,000-$6,000 of value placed on having a car sitting there even when you’re not using it. How can we build that same level of certainty, flexibility, reliability, control, into an on-demand mobility service, or a package of urban mobility services to rival that value of the car? I think that’s really difficult to do. That’s asking for a level of service guarantee on some of these systems that it’s really hard to even understand how we might get there technologically or from an infrastructure or a policy perspective.

And then you mentioned this extra layer of the Covid-19 pandemic. We did a survey during the Covid-19 pandemic over the summer in 2020. We did additionally ask our survey respondents how much compensation they would need for a month during the pandemic as opposed to a typical travel year in 2019, for example. What we found was that there was a threefold increase in the value placed on private cars and that that threefold increase came almost exclusively from ownership. From the fact that that car was there for me potentially in an emergency, in an anomalous situation, where it’s not my day-to-day travel needs that I’m worried about. It’s about what if I need to get to a health clinic? What if I’m not comfortable getting on the alternative today because I’m afraid of viral transmission?

It’s about that, again, control, certainty, reliability, which we often term as option value. The option to take that car when I need it, that was really where people have the value and this value increased during Covid. I think that really suggests that car ownership, in particular, is a security blanket for people. It’s providing that level of certainty in any type of emergency or global pandemic situation.

They don’t necessarily trust that an on-demand mobility service, or a package of mobility services, would be able to meet those, what we would call, “tail-end situations”— meaning they’re not your normal behavior, but they’re the extremes of, “I need to take that once in a year trip.” Whether it’s to go see family long-distance or to drive to the hospital in an emergency, to evacuate during a hurricane as David mentioned. And so that is where I think the Covid pandemic has even highlighted more this value of car ownership and how much it matters to us that that car is there for us whenever we need it, even if we’re not using it.

DK: At this point, the listeners might be thinking that we’re mad king, pro-car advocates. That’s not actually the case, Joanna. I think what I found interesting in doing this piece of work is that rather than this being normative about us trying to prescribe what the solution is, this is descriptive work: we’re trying to understand what people’s motivations are and how they behave currently.

You and I both are big advocates of transit and micro-mobility and other solutions as well, and we live in a city of a major metro area so we have a lot of those options to us. I think what’s interesting though, is that if you think about this bundle of mobility needs that we have throughout the year—whether it’s wanting to go for a weekend away or an annual vacation, or because I have to drive as part of my job—what’s revealed to us, as we try to value not just car ownership but these other mobility services (from transit to micro-mobility), [is] that there’s something that cars can provide uniquely and that most other mobility alternatives really struggle to provide.

So it’s great to be able to take a scooter in the inner city or to take a car sharing, take a Zipcar for a day trip to the beach. All of those things are fantastic. But that there’s this long tail that we haven’t really yet come up with effective product offerings that people are really grabbing hold of. The data reveals that most ride hailing, car sharing, and alternatives only really make up single-digit percentage points of our total mobility portfolio as a society.

That makes me look ahead and ask, ”What can we take out of this?” Based on what we’ve learned in this study, if I really wanted to get a lot of people out of private car ownership at scale—tens, hundreds of millions of people—what might we need to do that we’re not doing currently?

JM: This is an area where we’re both actively thinking about what are the next steps? How do we take this understanding of the value of a car and translate that into recommendations or understanding of what we would need to provide in a package or an urban transportation system that would really provide a real alternative to private car ownership and use in U.S. cities?

One of the major findings here is that the system attributes need to give people or mimic the value of car ownership. Things like flexibility, certainty, and control are what really matter to people. And so how might we think about designing what we often in our field talk about as mobility-as-a-service package—this is the idea of, instead of buying a car, you would purchase a monthly subscription to different transportation alternatives. This is sort of like how you would shop for a cable package for TV channels, but instead, you would pay some monthly amount and then you would be able to take transit trips, biking trips, walking trips, and some trips by ride hailing or calling vehicles. Here, I think that this emphasizes the need to think a little bit more about some of the attributes that provide that certainty, reliability, and convenience that might rival the private car. Things like guaranteed wait times, guaranteed travel times, making sure that the hours of operation cover any of these sort of tail end extreme scenarios: what if I really want to go get ice cream in the middle of the night while I’m watching the election results come in. Can I go out on my street, walk even just a couple of blocks, and get a Zipcar that I can take there or call an Uber or Lyft and know that it will be there within a reasonable amount of time.

We’re trying to understand what people value about cars so that we can understand how to build that in to alternatives to private car ownership and use, and how we might then encourage people to switch to a service if we can provide that same level of service, or at least a similar or competitive level of service with other alternatives. And so we’re in the process of designing a second survey, a follow-up to our survey where we valued car ownership, to try and better understand how we can design a more competitive mobility-as-a-service option to combat private car ownership and use in the U.S.

DK: I think that’s exciting because there’s unanswered research questions there, but there’s an enormous opportunity for innovation and entrepreneurship and having a real impact in our cities to go and launch new offerings that could combine multiple modes in novel ways. Certainly, you and I think that involves access to using a vehicle at some times, sometimes during a week or a month or a year, but it doesn’t necessarily mean that we have to own it in the same way that we do today.

We’ve spoken about a number of different possibilities in terms of ride hailing and car sharing like Zipcar, which is round trip car sharing—which is that I pick up the car at a particular neighborhood location and I bring it back to the same place—but it’s really much more of a continuum like that. One possibility that I really am interested in is a peer-to-peer car sharing where a neighbor might rent me their car, mediated through a platform. A company like Turo comes to mind in this case. Between us, we can own one car that two families could use, for example. I think the abstracting away from the specific offerings we have now to try and get at what it is about those different alternatives that make them more or less attractive has really practical applications, as well as the theoretical contribution that we’re obviously trying to do as academics.

JM: Yes, absolutely. Just to summarize some of the main findings from our study now, and then what we’re looking at going forward, is that we’ve really found that the value of private car ownership and use in the U.S. today makes it well worth the high costs of car ownership and use for most individuals. That most of that value comes from the ownership, not the use itself. What that means is that there really are very few alternatives, even in dense urban areas in the U.S. A reminder that we surveyed in Chicago, in Washington D.C., and in Seattle, where they do have fairly robust public transit systems and alternatives to private car ownership and use. But even in those areas, the value of car ownership, the value of that flexibility and control and reliability that it provides was very high.

What that means is that there really is a challenge for us to identify and design urban transportation alternatives. We feel they have to probably include some form of automobility in order to compete with that same level of service, flexibility that owning and using your own car provides. There are opportunities here for not just us as researchers, but also for entrepreneurs entering into this space to really try and find ways to mimic or provide that same level of service and value that the current private car ownership and use in the U.S. does.

Overall, the value of the private car really explains why car ownership and use has been so persistent in the U.S. despite the high cost of car ownership, despite the fact that those cars sit 95% of the time unused. That even with new technologies emerging, there’s still a lot of space and increases in population and demand for mobility, suggesting that we haven’t yet reached peak car. That we are going to continue to see the private car owned and used in our cities until we can identify these alternatives or package of alternatives that can really compete with it.

This is, I think, a really important area for us and for others to continue to think about. We can’t just hope that generational changes or new technologies are going to somehow supplant the dependence on cars. We need to be actively thinking about and designing alternatives that can provide people the same level of mobility and value that the private car currently does.

With that, I want to ask, David, where can people that are interested in learning more about this work go to find out more information?

DK: You and I, Joanna, we can be found in the MIT Energy Initiative in the Mobility Systems Center. For my research on this and related topics on Google Scholar and on Twitter and LinkedIn. You can find me in all the usual places. We’d certainly love to hear from any listeners who have follow-up questions or related ideas.

JM: Likewise, you can find me, probably the best place is Twitter, I’m @moodymobility—that’s m-o-o-d-y, my last name, mobility. We really want to thank you for listening, and David, I want to thank you for the conversation and for joining me here on the podcast today.

DK: Thank you, Joanna.


Press inquiries: miteimedia@mit.edu

We're hiring! Learn more and apply