(Updated with a comment from Naval Ravikant)

What up humans?! Bradley Voytek back in action with a quick #uberdata economics analysis. Recently someone named Paul Faguet sent me a tweet about an article from The Verge in which Naval Ravikant, cofounder of the startup network AngelList, is said to be, “selling his cars” in favor of using Uber and other transportation services.

(Edit: According to Naval Ravikant, he calculated that his 3-series BMW cost him $1200-$1500 per month. $300 garage at work, $300 garage at home, $600 car payment (BMW 335 convertible), $200-$300 registration, fuel, insurance, and maintenance. He uses UberX 3 times per day costing him approximately $54/day (fewer times on weekends). He also adds, “I can relax in the back on phone or iPad, no road rage on arrival, no hunting for parking, door to door service.”)

Paul asked me what the economics of such a decision looks like. While I’ve done this before for taxis, cars are different. More complex. This is a topic that was recently on my mind as well because my wife and I bought a new car a few months ago (our family added a new little human who takes up a lot of space). When we bought the car we opted to sell our old car. We came to this decision because we crunched the numbers and concluded that owning two cars in the Bay Area doesn’t make any financial sense.

In this post I look at how much owning a car in San Francisco really costs and compare that to using Uber instead.


Cars are tricky beasts because, much like housing, the decision to own one is often not one of economics even though housing and transportation are the number 1 and 2 most expensive parts of the average American’s budget, respectively. Clearly there are cheaper options to owning a car, however public transportation, for example, is very limited and, in San Francisco at least, MUNI can be unreliable with only a 52.7% on-time rate.

There is also a strong psychological factor to car ownership as well. It’s no secret that Americans love their cars. There’s the freedom aspect, the “cool factor”, and, of course, supposed sex appeal.

In fact, a 2010 research study out of the UK sought to experimentally test the effects of cars on perceived attractiveness. The study, “Effect of manipulated prestige-car ownership on both sex attractiveness ratings”, published in the British Journal of Psychology had heterosexual men and women rate the perceived attractiveness of the opposite sex. The researchers manipulated the surrounding environment in the photographs of the men and women receiving the ratings such that the people were shown seated in either a Bentley Continental GT or a Ford Fiesta ST (both pictured below).
Bentley Continental GT & Ford Fiesta ST

Okay, here are examples of the actual stimuli that they used:
Uberdata - Dunn & Searle Figure 1

They found that overall, men tended to rate women as more attractive than women rated the men. They also found that car type had no effect on how men rated women’s attractiveness, but women rated men pictured in a Bentley Continental as more attractive than the same men pictured in a Ford Fiesta ST. Here are the results in bar chart form:
Uberdata - Dunn & Searle Figure 2

This is an interesting psychological effect, as it tells us that there is a certain sex appeal to being seen in a nice car. Humans, man… how weird are we?

Anyway, with those psychological effects in mind, let’s step back and look at the economics of car ownership in the Bay Area. The true cost of car ownership has to account for real-world costs including depreciation of car’s value, fuel, insurance, maintenance and repair, taxes and fees, etc.

Owning a car costs somewhere between $5,000 and $15,000 per year

(sources: 1, 2, 3)

Cars depreciate in value. A lot. New cars really are pretty bad financial decisions (but of course, it’s not just about the finances).

Unsurprisingly, this cost scales with income, as people with more money buy more expensive cars that are also more expensive to own and operate. Meaning Uber’s demographic probably tends toward the higher end of the $5,000-$15,000 range. Of course these numbers depend on the type of car, driving habits, etc., so caveat lector.

Interestingly, the above costs-to-own estimates do not account for parking, tickets, and tolls, which also tend to be pretty high in San Francisco, which is one of the most expensive places to park in the US.

As a regular BART and Muni rider, I was surprised to learn that only about 15% of San Franciscans use public transportation to get to work. And for those who drive, their daily commute averages around 50-60 minutes roundtrip. Public transportation users spend almost twice as long each day getting to and from work.

Given all of this information, how does Uber compare? Well by January 2013, in San Francisco the average driver took only 2 minutes 40 seconds to arrive after they agree to pick you up.

Uberdata - Uber pickup times

Honestly (#humblebrag) this can often feel too fast. “Oh no we’re so awesome that we’re too convenient!” #firstworldproblems for real. This means that the average wait time for an Uber is essentially no more than what it would take you to walk to your car and start it up yourself. It’s negligible. And far better than what you’d get compared to calling a cab (good luck) or even with a street hail.

The point of this is that there’s a hugely significant time savings with using Uber. Do you want to sit in traffic or crammed into a standing-room-only BART car for an hour or two a day, or do you want to read or relax in the quietness of an Uber?

The answer to that depends on the cost. So how much does Uber cost? Again, for San Francisco, the median fare is only $22. For UberX it’s $17. That means that 50% of UberX trips cost less than $17 (and 50% more) but you can expect to pay around $17 for any given trip. The math here is simple:

This means that if you ditch your car you’d be saving enough money every year to afford up to 882 UberX rides per year.


That’s enough to use Uber to get to and from work each and every day, with another 382 trips left for whatever else you wanted to do. That’s enough to support an Uber-only transportation model. Even at the lower end of car ownership costs of $5,000/year, which I think is fairly unlikely for San Francisco, that comes out to 292 UberX rides per year.

When I first read about Naval Ravikant’s decision I was skeptical that it would be practical for anyone other than the most wealthy. His is probably not the decision most people make, realistically, but an honest look at the cost of car ownership really puts the finances into perspective. And the math doesn’t lie. In a city with with a fast, practical, (and cool) alternative like Uber, it would very much be worth your time to crunch the numbers and see how much you’re really spending on your gas, parking, maintenance, etc. and how much time you spend doing those things.

Because having your own personal driver pick you up in a slick car each and every day is a hell of an appealing alternative.