Hacker News new | past | comments | ask | show | jobs | submit login

The CDC has keeps statistics on how people in the USA die http://www.cdc.gov/nchs/fastats/lcod.htm

Looking at the 2011 prelims "Accidents (unintentional injuries)" comes in at #5 with 122,777 deaths and Intentional self-harm (suicide) is at #10 with 38,285. Assault (homicide) is no longer in the top 15.

Also, Assault (homicide) by discharge of firearms is 11,101 with all other Assault (homicide) totaling 4,852. To give some context to the Assault (homicide) numbers, "Accidental poisoning and exposure to noxious substances" totals 33,554 deaths or about 3x the Assault (homicide) firearm number or 2x the total.

Looking at the stats and what kills us, we spend a lot of time looking at the stuff that is actually going down versus the stuff that is increasing.




It's important to remember that causes of death are often also activities which carry utility and even outright life-saving capability.

Accidental poisoning, for example, is often due to exposure to cleaning chemicals and drugs which have massively reduced illnesses and deaths. Spending more to make them less accessible may actually be self-defeating, in that it could well cause more deaths than it prevents, by making it harder to get and use those things for their intended purpose.

Further, many of the more prominent killers in society have already been the subject of safety campaigns. Cars are massively safer today than they were in the 70s and earlier, as a direct result of safety research and societal effort. As returns on such efforts have been diminishing for some time, the next dollar of safety research or societal effort is quite likely to impact more net lives when directed at a "lesser" killer that hasn't been the subject of as much study.

One must keep those two considerations in mind, when making judgements about whether we're spending an undue amount of time and money on a given threat.

Simply looking at a stack-rank of 'killers' isn't enough.


Simply look at one line of data on anything isn't enough, but it isn't a bad place to start looking at what our actual risks are to see what the numbers on that risk are. Using that as a guide for the first things you should be concerned with helps quite a lot.

Looking at the auto stats http://en.wikipedia.org/wiki/List_of_motor_vehicle_deaths_in... I am not sure I would say massive.

But today, we are surrounded by "incident news" without looking at the trend lines. "Incident news" does not look at daily problems as its just noise and not headline friendly. It doesn't have a story narrative.

Risk vs reward is always going to matter. We watch TV and see "gun control" as the story. The long term trend lines for guns are more now than ever, more concealed carry, and less deaths.


From 25 per 100,000 to 10 per 100,000 is not a massive change? Or, using the more appropriate rate -- fatalities per hundred million vehicle miles traveled -- dropping from ~7 in the 1950s to ~1 today?

Again, the stack rank isn't a bad place to start if you have no other context. But we do. We're well past that point. We all know the NTSB, Auto Companies, Insurance Companies, et al have been hammering on car safety for decades now, have been advancing auto safety about as fast as society will allow and are always conducting newer/better research.

So it remains ill-advised to use that rate (or similar well-studied risks) as some sort of reference point when considering whether a smaller risk is getting outsize attention.

> "The long term trend lines for guns are more now than ever, more concealed carry, and less deaths."

In the 70s, when it became an issue (due in large part to media crusading and 'incident news'), the long-term trends for auto fatalities were also on a downward slope from the highs of the 40s and 50s. Yet, as it turns out, we could still actually do better by allowing research into the issue and taking common-sense precautions.

Similarly with the long term trends around smoking rates, when we finally deigned to allow smoking/cancer/cessation research.

Similarly was Airplane travel, even with a bump from terrorism, the safest way to get from A to B. Yet it remains a good thing that we studied the terrorism 'problem', despite its low risk, and changed air crew procedure and hardened cockpit doors.

And the TSA stuff is nonsense, but it was and remains nonsense precisely because it's not being studied and considered in a sober cost/benefit analysis.

What remains notable about the gun debate in the US, is that we're not arguing about "what to do" so much as arguing about whether to study the problem.

Look at how the trial balloon for the "whether we should study the problem" effort only has some sliver of a chance to succeed, because it's being first aimed at a scapegoat (violent video games). It's not even a direct study of the self-evident real problem (gang violence) or its equally-obvious underlying cause (the war on some people's use of some drugs).


I was looking at fatalities per miles traveled (given that is the measure of how much we use cars). It seems the trend line was dropping pretty nicely before the 70's, so I am not sure how much all the safety requirements have actually changed things. Seat belts and air bags might account for the delta from the 70's, but I am not sure it isn't the same reasons we see a reduction from the 30s. Looking at what happened after a change is good, but it needs to be in the context of what was happening before.

> What remains notable about the gun debate in the US, is that we're not arguing about "what to do" so much as arguing about whether to study the problem.

You must be watching some other debate, I consider new legislation being introduced "what to do".


> "I am not sure how much all the safety requirements have actually changed things"

Well, absent digging into details that neither of us have, that's going to come down to a philosophical position: do you think that trends will continue indefinitely in the absence of any additional efforts to sustain them?

> "You must be watching some other debate, I consider new legislation being introduced "what to do"."

I'm watching the debate where practical political reality says every legislative proposal, other than "study the problem", is trivially DOA due the lobbying power of the NRA and Congress' makeup and existing obstructionist strategy.


This thread is how I wish all discussions were on this site. Disagreement that is just polite enough, tons of facts on both sides, both sides with reasonable opinions founded in reality. Thanks roc and protomyth.


sorry for the late response

> Well, absent digging into details that neither of us have, that's going to come down to a philosophical position: do you think that trends will continue indefinitely in the absence of any additional efforts to sustain them?

I would expect continued downward trend, but it cannot go down indefinitely (limit on flukes). I expect a upswing in the first couple of years of automated vehicles just because new realities tend to make for bad times (see the 1930's) then another sharp downward trend.

> I'm watching the debate where practical political reality says every legislative proposal, other than "study the problem", is trivially DOA due the lobbying power of the NRA and Congress' makeup and existing obstructionist strategy.

Just as I hope the EFF and ACLU will be obstructionist in defending some of the other amendments. The NRA is not powerful by itself. It is powerful because of the number of members. I dearly wish we had an NRA for the 5th amendment, but it seems we are getting beat there very badly. I would say the Senate has been more obstructionist due to their failure at passing a budget. They should remember which house is supposed to be preeminent in that regard.


In 2010 most deaths by firearm were suicides [19,392]. There were 11,000+ homicides by firearm. And 412 deaths from legal intervention by any means.

In other words, the person you are most likely to kill with a gun is yourself.


I imagine, also, that if one intends to kill oneself, having a gun or not is not a serious impediment to suicide.


Research suggests that more guns = more suicides.

"The preponderance of current evidence indicates that gun availability is a risk factor for youth suicide in the United States. The evidence that gun availability increases the suicide rates of adults is credible, but is currently less compelling."

http://www.hsph.harvard.edu/hicrc/firearms-research/gun-owne...


I might try to get my hands on that study, as we had some trainers in the past using previous studies by the same group that didn't correspond to state stats at the time, but I'd like to point out two things about their studies:

1) Their gun in self defense studies http://www.hsph.harvard.edu/hicrc/firearms-research/gun-thre... have stats that do not match the US Dept of Justice

2) I would love to get ahold of the full study for #13 "The public does not understand the importance of method availability." because it would explain their base beliefs on a lot of the rest of the page.


I can't read the study, but lots of cultures in the US attach stigma to suicide(weakness) and to gun accidents(stupidity). Families in these cultures sometimes feel that they share the blame for not properly securing the firearm or not effectively mitigating the suicidal ideations of a depressed family member.

Consider a family with a youth who accidentally shot themselves. Rather than state that it was an accident, the family tells authorities that it was a "suicide".

Consider a family with an adult that chooses to commit suicide. Rather than state that it was a suicide, the family tells the authorities that it was an "accident".


Suicides misreported as accidents; accidents misreported as suicides. If both happen equally often, they'll cancel each other out.

Is there a good reason to believe that one of these effects is stronger than the other?


>>>Is there a good reason to believe that one of these effects is stronger than the other?

That's a great statistical question! If the data skew was proportionally equal in both directions, the two skews would cancel each other out, right?

Unfortunately, we don't know if the data skews towards accidents, or skews towards suicides, or skews towards them both equally. Here is the snarky one-liner answer to your question:

>>>Is there a good reason to believe that one of these effects is not stronger than the other?


> Looking at the stats and what kills us, we spend a lot of time looking at the stuff that is actually going down versus the stuff that is increasing.

We need to be careful: maybe things that are going down are going down because of how much time we have spent looking at them.


Possible in some things. They do make a special not on HIV in the paper that should not be ignored.

On guns, we have more than we had before, so the liberalization (old school meaning) of gun laws has not increased any gun related violence. On the other hand, we spend a lot on educating the public on proper eating. I get the feeling our approach isn't correct given the changes, but I would have to look up some study data.


I don't deny for a moment that Americans (and indeed, people everywhere) worry about the wrong things. We are biased in a number of ways to pay more attention to risks based more on the availability heuristic than any realistic assessment of what is more likely to happen.


It’s also important to remember that the total rate of death is constant at 1 death/person. Some death types going down will automatically result in others going up, so an increase or decrease in one type might not necessarily mean that the risk has changed.

I guess the goal is to reduce the deaths we can do something about.


Or weight the stats by age or health death. Old people dying of cancer is better than young people dying in car crashes.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: