Hacker News Reader

> the author goes much further than that and concludes that because Australia does not provide enough evidence to reject the null hypothesis that the null hypothesis (gun control doesn't work) is therefore true

I don't think this is the claim. I think the claim is that the evidence is insufficient to merit supporting a particular policy intervention. Banning things is not without cost: there will be some economic consequences for people making the thing, and there will need to be an enforcement regime which will either increase law enforcement costs or reduce LEO time spent elsewhere, and the practical impact of that enforcement regime will, if past gun interventions are predictive, involve incarcerating a bunch more people, many or most of whom will be urban men of color, which will result in both monetary and societal costs. And if there's a buyback program, that will also be enormously expensive, and require either incurring new debt or not spending that money on some other means of life-saving.

I think it's reasonable to want to be pretty confident a particular policy intervention will work before agreeing to incur that cost, and I think the author is saying the Australia example does not constitute sufficient evidence to merit that confidence. That's not the same as saying it didn't or won't work.

> I don't think this is the claim.

That is exactly the claim, almost word-for-word.

"Neither nation [Australia or Britain] experienced drops in mass shootings or other gun related-crime that could be attributed to their buybacks and bans. Mass shootings were too rare in Australia for their absence after the buyback program to be clear evidence of progress."

But if you actually look at the data, this is simply absurd. Australia averaged one mass shooting every 18 months before the ban without about a dozen separate events. Since the ban 21 years have gone by without a mass shooting. To say that "mass shootings were too rare in Australia for their absence after the buyback program to be clear evidence of progress" is manifestly absurd. At the very least it requires an explanation: just how many mass shooting does she require before their total absence for over two decades can be attributed to an intervention?

You can verify the authors credentials. The idea it was paid for by the NRA has no basis in facts.

You're free to disagree with it.

All of the people who published anti-climate-change studies on behalf of the oil industry also had credentials. Even people with credentials can succumb to temptation.

And I didn't say that this piece was commission by the NRA. I said that it would not surprise me to learn that it was. I'm advancing that as one hypothesis to explain the very strange too-good-to-be-true correspondence between the author's rhetoric and the classic tropes and talking points of the NRA. That might all be a coincidence, and it might be that all of the oversights and faulty reasoning are really Leah Libresco's work alone. But that would imply that Leah Libresco is incredibly stupid and a really bad statistician, and that seems unlikely to me.

I know her personally, and she's a better statistician than you.

Also, I've independently done a lot of the same research, going to primary sources. It's correct.

Really? Do you also know lisper personally?

I don't have a horse in the race, but if we are doing argument by authority, I would trust the CS PhD and JPL researcher more than the blogger with a PoliSci degree undergrad who dabbled in data visualisation and SPSS and thinks she is an authority.

To say nothing of the fact that one of the reasons I think she might be a shill for the NRA is that I don't think that she's stupid.

FiveThirtyEight has a good piece that reached the same conclusion. It's likely that it's by the same research team, but is published under a different author:

https://fivethirtyeight.com/features/mass-shootings-are-a-ba...

That piece is completely different. It's mostly just a factual report. It comes to a much vaguer and more general conclusion:

"Gun violence isn’t one problem, it’s many. And it probably won’t have a single solution, either."

Libresco's piece has a long list of specific policies that she rejects (banning silencers, assault weapons), endorses (personal interventions), or ignores (background checks).

Is your issue with the conclusions or with the data? Your initial comment seemed to challenge the veracity of the data and the motivations of the author, but the same data is presented on FiveThirtyEight and same research team compiled both.

My issue is with the reasoning.

For example: if you drill down into the data you will find this (https://fivethirtyeight.com/features/gun-deaths-mass-shootin...):

"Did Australia and Great Britain’s reforms prevent mass shootings? It’s hard to say, simply because mass shootings are relatively rare. In the post-buyback period, Great Britain has had one massacre with guns while Australia has had none. It’s hard to calculate how many would have been expected without a ban. Australia looks more successful in this regard, because it had more frequent mass shootings before the ban (averaging about two mass shootings every three years from 1979 to 1996) Mass shootings in Great Britain, prior to the ban, were rarer. Prior to 1996, there hadn’t been a widely covered mass shooting in Britain since 1987."

So... between 1979-1996, a 17 year period, they averaged 2 mass shootings every three years, or one every 18 months, with about a dozen data points. Since the ban, 21 years -- 250 months -- have gone by without a mass shooting. I cannot even begin to imagine how someone could conclude from that data that it's "hard to say" whether the ban worked. At the very least this extraordinary conclusion requires an explanation, not merely a glib assertion.

I think the claim that the UK only had one massacre with guns is questionable - we had Hungerford and Dunblane, both of which led to bans, and neither type of mass killing has been repeated since

> Is your issue with the conclusions or with the data? Your initial comment seemed to challenge the veracity of the data

I don't see anything in lispers posts challenging data, only the authors selective portrayal of facts and reasoning from the data.

Further, your upthread comment didn't reference the data either, it referred (falsely) to the 538 piece coming to the same conclusion; you backed off to claim it was about shared data and not a shared conclusion when it was pointed out that your claim about the conclusion was false.

"Just because a classification is invented doesn't mean it is meaningless or useless."

In the case of "assault weapons", though, the classification does indeed happen to be meaningless and useless. With the possible exceptions of silencers/suppressors and maybe bayonet/equipment mounts, the vast majority of what constitutes an "assault weapon" has effectively zero real effect on that weapon's lethality (either to an individual or to a large group).

> It just seems too sloppy

citation needed.

https://news.ycombinator.com/item?id=15403907

Thank you.

You bet.