For what it's worth, the 2022 CTA homicides were a huge outlier. The years 2001-2019 had 0-2 homicides each (filtering "CTA" in "Location"), then 4 in 2020 and 4 in 2021. Meanwhile, leading up to the pandemic, the National Transit Database says Passenger Miles Traveled was approaching 2 billion (https://www.transit.dot.gov/ntd/transit-agency-profiles/chicago-transit-authority), was down to 800 million for 2020 and 2021, and presumably came back up a lot in 2022 (no profile available yet).
I agree it's not infinitely safe, but I do suspect transit comes out safer by most analyses.
Good point! I see:
2001 0
2002 0
2003 1
2004 2
2005 0
2006 0
2007 1
2008 2
2009 1
2010 0
2011 2
2012 0
2013 1
2014 0
2015 0
2016 1
2017 2
2018 2
2019 2
2020 4
2021 4
2022 9
2023 2 (partial; annualizes to ~5)
Looks like some of 2022 homicides being high is something we should expect to continue (higher rates over several years) and some is it being unusual?
For the denominator in my division I compared 2022 trips (which did come back up a bunch and is final) with the 2019 numbers for average trip length (2.5mi). Looking at the data you found for 2021 (which I didn't find; thanks for digging it up!) average trip length is 4mi (798,583,310 /195,980,563). I'd expect 2022 average trip length to be somewhere in between?
Actually, the problem is I misread the 2019 report. The number I pulled out was only for buses, which is not surprisingly lower than overall. Instead of 2.5mi (613M passenger miles divided by 249M trips) it should have been 4.1mi (613M + 1359M passenger miles divided by 249M + 230M trips). I'll update the post.
Yeah, we'll see [how transient the higher rates are]. It looks like NYC also saw a spike 2020-2022 (though I think rates per passenger mile are several times smaller) and this year isn't looking much better (going by https://www.nyc.gov/site/nypd/stats/reports-analysis/transit-bus.page and the NTD pages for the MTA).
I recall hearing that a large percentage of homicides are done by someone the victim is close with. And then I presume a big chunk of the remaining homicides are gang or drug related. After that, I suspect that common sense and being the bigger person in an argument goes a very long way. I suspect that once factored in, this stuff would reduce the rate by 3-5 orders of magnitude.
For cars, I'm not sure how much we can cut the death rate by not doing dumb things like drinking and driving. Given that there's about a 2.5-to-1 ratio of alcohol to non-alcohol-related deaths, I feel like being a good driver probably only cuts your risk by, at most, a factor of 2 or 3. Not an order of magnitude. It's hard because you can do everything right and end up dead because of someone else. There's only so far that defensive driving takes you.
With these two assumptions, public transit starts to look a lot safer.
I recall hearing that a large percentage of homicides are done by someone the victim is close with.
But probably less so on public transit? There are lots of muggings in the data, which often involve threats of violence; maybe some are escalated robberies?
Oh that's a good point. I agree that of the homicides committed on public transit it probably isn't much from people you know. So maybe 2-3 orders of magnitude would be better than what I originally wrote with 3-5.
No, I meant orders of magnitude. If you're not involved with drugs or gangs and you aren't the type of person to severely escalate arguments, I suspect that'd cut the risk by 2-3 orders of magnitude.
I share the belief that there's variance in risk by at least 2 orders of magnitude for being murdered. I suspect you get more than a factor of 3 for driving, probably an order of magnitude, if you include route and vehicle selection in the controllable options.
But that's just for violent deaths. It's much less clear which is safer in terms of rapid response if you have a heart attack. And it's probably a big win for driving if you worry about verbal harassment or uncomfortable situations that feel dangerous (but almost never end in actual physical harm).
It's much less clear which is safer in terms of rapid response if you have a heart attack.
Interesting, I never thought about that before. Do you think that would sway things much?
I'd think that public transit would be much better in the event you do have a heart attack though. If you're behind the wheel you won't be able to control the car and would probably get into an accident, whereas if you're on public transit other people would be around to perhaps help.
And it's probably a big win for driving if you worry about verbal harassment or uncomfortable situations that feel dangerous (but almost never end in actual physical harm).
Yeah I think that's probably true.
I doubt medical response is a big factor in this, for most readers of LW. But I don't actually know - heart attack is up there in terms of likely cause of death for many of us (though not on our next commute, one hopes). On the other side, I'm surprised nobody's mentioned exposure to disease as a risk factor on public transit - only a few years ago it was an overwhelming safety factor, by many orders of magnitude, and pretty much all anyone would consider in this decision. It's hard to believe that it's now irrelevant, though it's hard to figure how to weight it.
I doubt medical response is a big factor in this, for most readers of LW. But I don't actually know - heart attack is up there in terms of likely cause of death for many of us (though not on our next commute, one hopes).
Gotcha.
On the other side, I'm surprised nobody's mentioned exposure to disease as a risk factor on public transit - only a few years ago it was an overwhelming safety factor, by many orders of magnitude, and pretty much all anyone would consider in this decision. It's hard to believe that it's now irrelevant, though it's hard to figure how to weight it.
Yeah it is tough to weight. My intuition is that it's pretty small though. I don't think a lot of people die of diseases that are transmitted from others, although I could see a decent proportion of those diseases being transmitted on public transit.
All comparisons like this are completely misleading, because they're based on averages across widely variant situations and people. It's a BIG hint that you're talking about transit homicides vs automobile death by collision. The two biggest determinants of safety for this decision are NOT "for the median trip, which mode has fewer deaths?", but "for you personally right now, are you drunk, and/or are you involved with often-violent people?".
Sure, there are many car deaths where the traveler is not drunk, and some transit deaths (unclear if you're counting the walk and wait portions, or only on the transport, but it doesn't matter) of uninvolved bystanders. But they're only a fraction of the reported statistics, small enough that the comparison falls apart.
For this decision, like MOST choices in life, variation is so great that you really can't update on averages.
For this decision, like MOST choices in life, variation is so great that you really can't update on averages.
I like the way you put that. I think I've had some intuition for this but this kind crystalizes it for me. Thanks.
That said, and to steelman a bit, if you want an attempt an analysis, you kinda have to work with what you've got, and I suspect that's the angle that OP is coming at this from. Although I do think that it would be worth caveating more because of the high variance.
I'm not sure how to get to a place where we see fewer bold false claims, but a culture of giving sources, and supporting requests for them, seems like it would help?
A culture of giving sources and supporting requests for them is a culture of "doing your own research". Over the last years, the ideology of the censorship industrial complex is that "doing your own research" is bad and that people should instead believe in the authorities. No mainstream media outlet would run an OPed titled "In defense of doing your own research".
Culturally, many people believe that it's important to trust in authorities instead of make statements in support of the mainstream narrative.
This leads to strange effects such as posts on skeptics.stackexchange that ask whether certain antivaxx claims are true getting downvoted in a way that they wouldn't have five years ago because investigating whether nonmainstream claims are true was part of the point of skeptics.stackexchange.
Actually, advocating for a culture of "doing your own research" is not going to be an easy project.
You could add a community note to that tweet. I think that's a pretty good tool to correct and discourage such claims on Twitter.
I'm not sure how to get to a place where we see fewer bold false claims, but a culture of giving sources, and supporting requests for them, seems like it would help?
Are there any networks on mastodon that are actually capable of maintaining the incentives, the level of accountability, that it takes, to stop spreading misinformation? Because I'm not aware of any, but [looks] a MIT instance would be a very credible candidate? Is it quite good?
I don't really think it's worth trying to save microblogger cultures, the microblogging format inherently encourages oversimplification, rapid spread, and it's profoundly hostile to real conversation. Personally I'd recommend contributing to the construction of better systems like bonfire.networks instead.
there were 0.71 driving deaths per 100M miles travelled
I don't think this is a good basis for comparison. The comparison in the tweet you link to seems to talk about commuting, so it would probably make sense to compare based on the number of trips rather than based on the miles traveled. In particular, in the data you link to, we can see that the number of car accidents from trucks is pretty small, while I would guess that truck drivers represent a bigger share of miles traveled, thus lowering the death rate per mile, without this being relevant information.
[EDIT] I looked it up
For public transit, you write (for a high crime rate city, but not the highest possible, and for an outlier year, as pointed out in another comment)
This is 3.7 homicides per 100M trips.
In the US, there are about 400 billion car trips per year, and about 40k death per year, so this implies an average death rate of 10 death per 100M trips. This is an average.
This makes sense, but going this direction gets pretty tricky. A trip that you could plausibly take public transit for is probably also the kind of relatively slow urban commute that is unlikely to result in a fatal crash. For example, I live in Somerville MA, and checking FARS we have had one car crash that was fatal to someone in the car in the last decade. [1][2]
I can't easily find numbers on how many car trips people in Somerville took over the last decade, but back of the envelope: 80k people * an average of one car trip per day * 365 days * 10 years = 290M trips. So 0.3 per 100M Somerville MA trips? (Which is also perhaps the most literal interpretation of the "where you live" in the original tweet?)
[1] This excludes car crashes where the only fatalities were pedestrians, which I don't like but I think is the right way to do this analysis?
[2] Going back farther there used to be a lot more. I think this is a combination of safer cars and traffic calming?
I agree with you that this gets pretty tricky.
One trip per day seems very low, don't people usually do at least two trips per day (going to work, going back home), or even 4? Are trips bi-directional (in which case I must apologize, this would be a misunderstanding on my side)?
None of this is to discourage your request that such claims be supported by sources, as a standard.
One trip per day seems very low, don't people usually do at least two trips per day (going to work, going back home), or even 4? Are trips bi-directional?
I was counting a trip as one direction: if I take the subway to work that's one trip, and then another when I come home. But a lot of trips won't be car trips, and everyone has a commute, and not everyone goes out every day? So I was guessing this would average to one car trip per person per day? Though thinking more now that's probably too low?
4 of the 9 homicides occurred on CTA property, but not on trains or buses. Does that mean you should include all homicides that occur on streets, driveways and parking lots?
When you travel by public transit, especially at the late hours we disproportionately see deaths at, you'll spend a lot of time waiting. When you drive, though, you just get in your car and go. So I think the comparison is right as is?
Focusing on only deaths makes some sense since it's the largest likely harm, but I will note that death is not the only outcome some people are afraid of on public transit; if you include lesser harms like being mugged or groped then that is going to tip the scales further towards driving since those things have a ~0% chance of happening while driving, and a small-but-probably-non-trivial chance (I haven't looked it up) on transit.
If you want to bring those in (and I agree a more thorough analysis would) then you'd also want to bring in non-fatal automotive injuries; being in a collision can be pretty bad without killing you, and collisions are rare on public transit.
Oh I entirely agree.
My guess is that a lot of the difference in perception-of-danger comes from how much control people feel they have in each situation. In a car I feel like I am in control, so as long as I don't do stupid stuff I won't get in an accident (fatal or otherwise), even though this is obviously not true as a random drunk driver could always hit me. Whereas on transit I feel less in control and have had multiple brushes with people who were obviously not fully in their right minds, one of whom claimed to have a gun; I may not have actually been in more danger but it sure felt like it.
You shouldn't be comparing only HOMICIDES, the point of this comparison is that you are much more likely to die (from any cause, most likely from an automobile accident) than you are to die while taking public transit (from any cause). Homicide may be the most likely cause of death while taking public transit (or it may not), but I imagine the odds of getting killed in an accident while driving present a much greater risk.
The statistics I'm comparing are public transit homicides vs driving accident deaths. So if you have a heart attack, get shot while driving, or your train disastrously derails that's not included. I think this is fine, though, because including these far less common kinds of death is very unlikely to change the result?
I recently came across this tweet, screenshotted into a popular Facebook group:
This got ~1M views, doesn't cite anything, was given without any research, and, I'm pretty sure, is wrong. While I'm a major fan of public transit, they've stacked this comparison in a way that's really favorable to cars, and it's not surprising that public transit doesn't make it.
Safety is a complicated concept, and risks are situational: in a car you're much more likely to be hurt in a collision, while on public transit you're much more likely to be hurt by another passenger. To get a clear comparison I looked just at deaths, which is also an area where we can get good statistics.
I can't find a listing of public transit agencies by homicide rate, but Chicago is a large city with a lot of homicides and they make their data available so let's look there. In 2022 there were 244M CTA rides. Downloading the Chicago Police Data and filtering to 2022 homicides on public transit, I see nine. This is 3.7 homicides per 100M trips.
(Note that the original claim was for any city, and there are dozens of US cities with homicide rates higher than Chicago's. I think it's pretty likely that at least one of these cities has a public transit system with more homicides than the CTA.)
To fairly compare this to the risk from driving, we need homicides per distance. How long is a trip? I can't find 2022 data, but the CTA's President's 2020 Budget Recommendations gives 4.1mi (1359M + 613M passenger miles divided by 230M + 249M trips). This means 0.9 homicides per 100M miles travelled.
I live in MA, and while 2022 FARS data isn't out yet, in 2021 there were 0.71 driving deaths per 100M miles travelled. While MA does happen to be the safest state for driving, MN, NH, and RI also have lower driving fatality rates than the CTA.
I'm not sure how to get to a place where we see fewer bold false claims, but a culture of giving sources, and supporting requests for them, seems like it would help?
Comment via: facebook, mastodon