Hacker News Reader

Comment by CalRobert | Hacker News Reader

The car thing is very frustrating. It's the most dangerous thing most parents do with their kids (myself included) but for the most part we're blind to it.

Of course, pedestrian deaths have shot through through roof in the US and that makes it harder to justify walking to school.

View on HN

It's a purely emotional response about what a parent can control. Driving the kids to school, you have some control over the situation; e.g. you can swerve out of the way of a potential accident. When they walk alone to school, there is nothing you can do to help them or prevent accidents from happening and that lack of control opens up the imagination to all sorts of scary things.

Why can you swerve out of the way of a car, but you can't jump out of the way?

It's EXTREMELY uncommon for cars to Grand-Theft-Auto-style run people down walking on the sidewalk.

~80% of pedestrian deaths happen when it's dark.

There's about ~19 pedestrians that get killed by cars per day. Only ~4 during day time.

You have a higher chance being in a random mass shooting. Worrying about getting killed by a car in broad daylight is silly.

It being all about control is spot on.

While it’s possible and maybe even likely that I overrate my own driving ability, I’ve never been in an accident in 20+ years and drive in a paranoid, defensive, assume-everyone-is-crazy style that I feel must push the statistics in my favor quite a bit.

So despite statistics I tend to worry more about other things. I also have a hard time with my 3 year old daughter being driven by others unless I’ve seen their driving firsthand. There are people I know who I flat out won’t allow to drive her and it has created some awkward situations. I will also only let her go in an Uber/taxi if there’s no other option.

Your numbers and statements seem logical but can you provide any datasets to support? I’m slightly surprised a random mass shooting would be a higher likelihood than something automotive related.

I didn't check your pedestrian deaths during the daytime statistics, but ~4 per day would be ~1400 a year. I did check the number of mass shooting victims, and in 2019 (the most recent year for which data is really meaningful) the total was 180 (~.5 a day), of which 71 were deaths (~.2 a day)[1], or about 20x less likely than getting killed by a car during the daytime.

That's pedantic, but I just wanted to clarify because mass shootings are also in the news.

[1]https://www.statista.com/statistics/811504/mass-shooting-vic...

There's different measures for mass shootings.

By some measures (https://en.m.wikipedia.org/wiki/List_of_mass_shootings_in_th...) - this year there are ~1.8 per day.

So you're about twice as likely to be killed by a car in daylight as a pedestrian as you are to be killed in a mass shooting.

Your point stands. I overestimated.

The world is bigger than the USA. The chances of getting caught in a mass shooting are lower than you are estimating, and pedestrians are more impacted by automobiles.

We have those viruses and diseases too, you know.

It has to do with at what percentile of wary and being good drivers we think we are. Also abduction is not the only risk for children.

I believe that often than not people do the right thing, if it something done in mass there is a reason for it and it is just hard to pin point the reason.

There can be social reasons for it, also we don't know if there is some implicit higher chances of getting abducted for kids who get drove to school. Like them walking to school will increase the chance of them abducted.

That sense of control is a dangerous illusion, though. Cars move way too fast for people to be able to react when something unexpected happens. Drivers included.

> That sense of control is a dangerous illusion, though. Cars move way too fast for people to be able to react when something unexpected happens. Drivers included.

That's a dangerous fatalistic (literally) assumption. One must train and practice car control. Which unfortunately isn't required in the US, but training is available for those who seek it.

A driver can most certainly safely control mostly any slide or emergency manouver. But it does take training and regular practice.

I regularly drive on Hwy 17 (a windy highway in the hills just south of silicon valley) and have witnessed countless times cars which start to slide and the driver panics and doesn't know what to do so they end up crashing. But watching the slide, all of them would've been easy to control by a driver exposed to a decent amount of car control training. It's unfortunate we allow people to drive without requiring car control school and frequent referesher training.

I tried to explain the fact that covid risk comes in many levels to someone I'm living with by referencing the risk we take on by driving, and she just immediately jumped to drunk driving. I guess people find it uncomfortable to live with the fact that there's a small but nonzero chance of dying.

In the US, firearms are more likely to kill your child than car accidents are[1].

1. https://www.nejm.org/doi/full/10.1056/NEJMc2201761

Regarding the "your", I bet vehicle collision related deaths are more randomly distributed than homicides and unintentional shootings of children (and probably firearm suicides too, for that matter).

If you take out 18 and 19 years old, that is no longer true.

We have way too many firearm deaths in this country but the way these studies choose their age ranges always feel like folks are in search of the conclusion they desire.

That article also includes suicides.

Not really a discussion I particularly want to get into, but many people think “an increase in suicides” should be counted as a negative consequence of firearms.

But the US has a lower suicide rate than countries with fewer guns?

Does shooting up some place count on the suicide figures, almost always it's going to be a decision to die (you might not die, but surely death of an 'active shooter' is the expected outcome).

No but it wouldn’t matter if it was

The US has a pretty high suicide rate, especially for developed countries

https://worldpopulationreview.com/country-rankings/suicide-r...

US isn't the highest but there are certainly countries with lower rates.

Right, other countries with higher rates have fewer guns. My point entirely.

Other countries with fewer guns have lower rates as well. Its a nonsensical argument.

Not sure why you're confused.

Claim: "More guns means more suicides"

Fact A: "Countries with far fewer guns have more suicides"

Fact B: "Other countries with fewer guns have less suicides"

Fact A disproves the Claim.

Fact B does not, at most it shows guns aren't a factor either way. Is that what you're claiming? Guns have no impact on suicides?

All we can know is it's not the only factor.

I think we can be more nuanced and discuss casual relationships for suicide and those for homicide and recognize that there are many differences. If we aggregate the two we will likely have bad policy and be inefficient at fixing either.

Depends if those lives were genuinely a net negative due to suffering.

How many people suffer longer than they should due to lack of access to suicide? I suspect the answer is "fewer than commit suicide due to temporary and fixable chemical imbalances and an easily available gun" but the question should be asked and evidence gathered.

As it happens, a large scale experiment is underway..

https://www.spectator.co.uk/article/why-is-canada-euthanisin...

There’s no evidence of correlation or a causal link between guns and suicides.

In countries with guns available, men use guns for suicide.

In countries without, men use hanging.

Either way, it’s intentionally misleading to lump suicide in with murder and accidents.

This somewhat begs the question that there are the same rates of suicides in the two areas you are mentioning. So, no, it is not intentionally misleading, but it does need more data to show it is meaningful.

> This somewhat begs the question that there are the same rates of suicides in the two areas you are mentioning.

It’s not hard to find examples; Japan has a much higher rate of suicide. No guns.

Almost universally, men use guns or hanging, depending on what’s locally available.

> So, no, it is not intentionally misleading

They’re juking the stats to be able to make proclamations like “guns are the leading cause of death for children”.

When most people hear that, panic bells start ringing. They assume this means school shootings are killing an incredible number of children each year. This is a crisis!

People don’t realize that to manufacture that number, activists had to extend “children” to 19 years old, and include “suicide”.

It’s intentionally misleading.

If only they’d dedicate the same energy to exploring why our educational and cultural institutions are failing teenagers so spectacularly that they’re suiciding at unprecedented rates.

I'll grant it can be misleading. I don't think it necessarily is. I say this as someone that has had gun suicides in the family. I also have Saturday Night Special running through my head, now...

I don't understand the fixation that counting kids as up to 19 and including suicide is somehow a concern. Seems generally alarming that people under 21 is that high for this statistic, period.

Edit: Per https://www.nationmaster.com/country-info/compare/Japan/Unit...; no, Japan is not much higher than the US. It isn't even higher. (I'm confused, do you have a better source?)

Edit2: Ok, https://en.wikipedia.org/wiki/List_of_countries_by_suicide_r... has it higher, but doesn't look "much" higher. Though, I grant that is a subjective term.

Edit3: https://www.who.int/news-room/fact-sheets/detail/suicide is an interesting source on this. I find it amusing that they are able to have articles that mention limiting access to pesticides as an effective way to reduce suicides, but don't look at limiting access to guns. It is listed in the top three methods. (Sadly, I'm stepping away from computer for a while now. Will be delighted to engage more on this later, but will probably be a day or so before I remember to check back here.)

> Japan has a much higher rate of suicide.

Maybe it was _much_ higher a very long time ago, but it has been on a downwards path for a long time. And the US's on an upwards path... for so long that it has actually past Japan. So not only is Japan's suicide rate not _much_ higher, but it has been, for a few years, slightly lower.

Source here: https://apps.who.int/gho/data/node.main.MHSUICIDEASDR?lang=e...

2019, per 100k data (average of both/male/female):

US: 14.5 / 22.4 / 6.8

Japan: 12.2 / 17.5 / 6.9

> This somewhat begs the question that there are the same rates of suicides in the two areas you are mentioning

I think if you look at gun deaths and suicide you're going to have significant autocorrelation. 60% of gun deaths are suicide, so the data is baked in. But I have another correlation that may be more fruitful and interesting. It isn't 1 to 1, but there is significant overlap with (per capita normalized) overdoses[0] and gun death rates[1]. We'd have to dig deeper, but I think it wouldn't be surprising to find that rates of gun suicide and overdoses has many similar root causal factors. The north east seems to be the biggest mismatch

[0] https://drugabusestatistics.org/drug-overdose-deaths/

[1] https://www.pewresearch.org/fact-tank/2022/02/03/what-the-da...

I'm not sure what you are claiming. Seems safe to assume that suicides do correlate in numbers. Question is if they substitute in magnitude.

That is, if it isn't 1:1 in correlation, then it is reasonable that you can prevent suicides by limiting access to guns. Just as you can prevent suicides by limiting access to drugs.

The a reason that is not valid?

> That is, if it isn't 1:1 in correlation, then it is reasonable that you can prevent suicides by limiting access to guns.

The problem here is you're assuming that there's a singular dominant causal factor. You don't need a one-to-one relationship for either a weak or strong causal relationship.

On the other hand, what I'm claiming is that there are causal relationships between gun suicides and drug overdoses. That solving one will help solve another if we look at the causes from a standpoint beyond the first order viewpoint. We need to ask why people are depressed and what in society is causing that. I'm sure you'll recognize that this issue is likely rather complex. I think you'll agree that getting to the root of this issue will save more lives overall (some people will for sure just choose an alternative method of suicide -- such as drugs -- though you are also likely to save a few that are acting in the moment and getting a gun -- which is why waiting periods can be helpful. But either way, they are still depressed, which is a problem we need to solve).

From the studies I'm seeing, limit access to the means of suicide, and you reduce suicides. Do you eliminate them? No, but a reduction is still a reduction.

Note most of the studies I saw were not about guns, but pesticides. I see no reason to think they would not apply similarly. Would be nice to see studies in them.

Now, should we /also/ try and tackle more issues down the chain? Yes. Not sure why this has to be a "pick one" scenario.

> There’s no evidence of correlation or a causal link between guns and suicides.

There's definitely evidence of correlation. But the thing is that there are clearly other factors. Access to guns is definitely a causal variable (you can't shoot someone unless you have a gun) but the question is if it is weak or strong. I think it is pretty clear that this is a weak causal variable. It's clearly naive to think that simply having a gun makes one more likely to murder someone or kill yourself. Similarly women do not are not subject to gun deaths (suicide nor homicide: killer or victim) at nearly the same rate as men when controlling for household ownership.

This is why I've encouraged people to look at a more complex picture. Clearly suicide, homicide, and mass shootings have different principle causal variables. There's definitely overlap as well. People are comparing to European countries and Australia simply based on number of guns. But this ignores a lot of other factors like access to health care, household debt, trust in government, social mobility, and many other things. The US consumes far more opioids per capita than European countries. We also have higher suicide rates. I'm willing to bet that there's some coupling relationships here.

But I agree it is misleading to lump homicide and suicide together. There's plenty of statistical paradoxes that depend on aggregating data that shouldn't be aggregated. If we over-aggregate data we'll have no hope of solving any of these issues. I'd make the argument that it is of the upmost importance that we be nuanced in these conversations and communicate in good faith. We must have self doubt in our own models. It is a complex problem, we've been handcuffed in obtaining data (forcing aggregation), and people with agendas are trying to oversimplify the problem and have us make decisions based purely on emotions. We live in the information age yet it is sad we have so little here and think so broadly about this problem. It is getting people killed.

One thing to note is that there's a pretty decent overlap between gun deaths/capita and drug overdoses/capita

https://drugabusestatistics.org/drug-overdose-deaths/

https://www.pewresearch.org/fact-tank/2022/02/03/what-the-da...

Regardless, from the POV of day to day risky things, none of this crap matters.

All of these "average risks" people are slinging back and fourth like idiots are dominated by outliers who are engaged in specific behaviors that have a strong causal link to that means of death. Said outliers drag up the average making it completely unrepresentative of the typical person's risk.

Don't drive drunk, don't pedestrian while drunk, don't get involved in drug industry business disputes and don't let mental illness go untreated and you will almost certainly not get killed by a car or bullet.

Would that offer some consolation to the parent?

Every single child (and people under 18 are just that: children) dying because their parents couldn't be arsed to follow even the most basic gun safety laws is one too much. Toddlers shooting themselves with the gun that their reckless father left under the pillow? Even in countries with very liberal attitudes to guns such as Switzerland don't have that problem.

Not to mention, the cutoff at age 18 is obvious because most people have finished education at that age - and the most shootings happen at high school.

The cutoff at age 18 is so that gang warfare that kills gang members who are under 18 can be counted as shootings of "children".

The US has a child accidentally shoot a gun almost every day, with 765 people ending up dead in five years [1].

That alone is horrific and completely unheard of outside the US and active war zones.

[1] https://www.npr.org/2021/08/31/1032725392/guns-death-childre...

So in a country with around 75 million under-18s, that works out to an approximate risk of 1/500,000 per kid per year. And per the article:

> The most likely age group to be both shooters and victims are teenagers ages 14 to 17

Maybe we should make firearms safety and hunters education classes once again part of the school curriculum to prevent these accidents? Just like with driving classes, training and experience in a controlled environment is a great way to reduce the frequency of accidents. Although that assumes the incidents in question are accidents, rather than "accidents".

That's teen gangs blasting each other in certain hotspots. Not kids accidentally being killed by firearms in average homes.

To elaborate, it's around 65% homicide, 30% suicide, and 5% accidental or undetermined.

Compare homicide rates

FIGURE 8: Male Gun Death Rates by Race/Ethnicity, 2020

A Year in Review: 2020 Gun Deaths in the U.S

https://publichealth.jhu.edu/sites/default/files/2022-05/202...

Live map of Chicago: https://heyjackass.com/

Scroll down and look at who the real victims are (it's sad)

2022 Race of Victim/Assailant https://heyjackass.com/?p=382

That last link is so upsetting. Not just because of the racial composition, but also because of the assailant/victim ratio being 25/220, which means an average of 11 victims per assailant.

9, not 11. But that is still significant.

The loss of math abilities is probably a side effect of the bout of COVID I'm currently experiencing. Probably.

That is just where assailant's race is known and recorded.

The homicide clearance rate in Chicago is at or under 50% and approximately 50% of those identified as assailants had no charges filed against them. This implies the assailant figures may be 4x under represented. Who knows?

https://chicago.suntimes.com/crime/2022/3/31/22996487/cpd-po...

yes. Texas and Florida have the highest gun deaths in the country.

Double false. First's it's not true. #1 and #2 are Texas and Californa

Second, unless you at least divide by population the comparisons are meaningless. A state with 1 person in it is going to have less death in general than a state with a million people

Dividing by population the top 10 are

    Mississippi
    Louisiana
    Wyoming
    Missouri
    Alabama
    Alaska
    New Mexico
    Arkansas
    South Carolina
    Tennessee

Texas is 26th, Floria is 29th

Source:

https://www.cdc.gov/nchs/pressroom/sosmap/firearm_mortality/...

The bottom 7 are

  California
  Connecticut
  New York
  Rhode Island
  New Jersey
  Massachusetts
  Hawaii

I wonder if the strong gun laws in California and NY, NJ have anything with them being at the bottom.

I'd bet that (like damn near every other violent crime) it correlates stronger with income than anything else.

Teasing out the correlation and causation here is difficult. People generally don't like getting shot, so it's plausible that stricter gun control is at least partially responsible for the higher wages, since places with better gun control laws are more attractive areas to set up businesses in.

Wouldn't the correlation run in the opposite direction if that were the case, since the employees would basically demand the higher wages as a sort of hazard pay?

> it correlates stronger with income than anything else

If it were only income, we wouldn't see reductions in deaths from mild gun-control law (like banning high-capacity, automatic "assault-style" weapons) within the same state.

We also wouldn't have seen the large increase in gun deaths after such laws were invalidated, and we have seen that[1].

So income is certainly some component of it, but access to untracked, powerful firearms is a big component of it.

1. https://www.judiciary.senate.gov/press/dem/releases/studies-...

Just a quick point of correction, not to derail the conversation, but "automatic" weapons have been heavily restricted since 1934. Buying one requires a tax stamp, background check (even if private sale), notifying the ATF of where the weapon is at all times, and giving the ATF the right to search the premises of the firearms at any time without need for a warrant or reasonable suspicion.

As far as I can find data for, legally-licensed automatic weapons have only been used in 2 homicides since 1934 (both by police officers, weirdly) and a single suicide.

To address your data directly, it has some issues.

The first is that there was no "large increase in gun deaths". The lowest gun deaths during the 1994-2004 period was 5.9/100k, and the highest it's been since 2004 is 6.2/100k. About a 5% increase. 2001 was actually the highest since 1994, at 7.1/100k. That's per CDC data. The incidence of mass shootings is what they're talking about, not gun violence overall.

The second is that we're working with small numbers here. Homicides with a rifle are exceedingly uncommon; per FBI data, it's about 300/year. Your link is specifically discussing mass shootings, which are rarer yet. For the 3 periods, they list 19 events, 12 events, and 34 events, each spread across a decade. Columbine happened in the middle of that period.

The third is that the correlation isn't as tight as you might presume. Have a gander at https://www.statista.com/statistics/811487/number-of-mass-sh... Unless someone pointed it out, it would be hard to guess where the assault weapons ban even is in time. The numbers aren't really notably different until 2001. Mass shootings aren't even noticeably high until 2007 when we hit the recession, which is a massive confounding factor.

The fourth is that if assault weapons are to blame, you would expect numbers to roughly match before and after. They don't. 2005-2006 follow roughly the same trend as before. It's not until 2007 when they go up (again, recession), and 2012 before it gets really crazy. 2012 is coincidentally when US median wealth hit it's lowest after the recession, and hasn't really recovered since then. California has more mass shootings per year now than they did before 2016 when they passed their high capacity magazine ban. I don't think it's causal, just of note.

There are just so many other factors that play into this. Increased media attention on mass shootings, niche online communities forming, increased political polarization, increased wealth inequality, zero tolerance school policies, etc.

I also find it a bad thing to focus on. Every death is a tragedy, but rifles are low on the chart of causes. Homicides by rifles are roughly equal to deaths by ladder. About 2 times as many people are beat to death with hands/feet. About 5 times as many people are stabbed to death. About 10 times as many people die from foodborne illness. About 20 times as many people die of anemia.

It just doesn't make sense. Even if you want to focus on guns, why rifles? Handguns are used to kill roughly 50 times as many people.

This is almost universally due to gang activity and not the kids picking up weapons from friends and family and shooting themselves or others.

> This is almost universally due to gang activity and not the kids picking up weapons from friends and family and shooting themselves or others.

This doesn't seem to be accurate. Research [1] looking at data from a few years back (before firearms became the leading cause of death for children in the US) found:

Firearm deaths in children were 53% homicide, 38% suicide, 6% unintentional.

Of homicides in older children (13-17) the leading circumstances were argument (40%), precipitated by another crime (31%), and gang related (21%.)

[1] https://publications.aap.org/pediatrics/article/140/1/e20163...

Reply on news.ycombinator.com