which extrapolates using least-squares approximation — ernestm

...

That tells no one anything about what you (allegedly) did. What whoever modeled this actually did in the forecasted years is exactly as @StreetlightX accused you of.

Even though you've previously posted a graph that shows crime trends nonlinearly with time.

Enough, man.

Really. The I provided the confidence factor for linear extrapolation from data since 2005, and as it is over 0.9, there is no need to spline model the rise and fall between 1990 and 2005. Which you should know.

The “defund the police” movement was a fringe idea two weeks ago. Now it is the rallying cry du jour. So it’s no wonder that, according to a YouGov poll, close to two-thirds of Americans oppose cutting funding to police—and of those who do support such measures, those who make $100k a year are over represented. So while some over-educated, trusty-fundy types are working to make their dreams become a reality, most Americans are wary of such proposals.

I could spend time educating you on R^2 and forecasting, but I won't. Suffice to say; a high R^2 in an strictly linear OLS model doesn't tell you anything about whether you've correctly modeled the data generating process even in most lab settings, nevermind when you've already got good information that the process is strongly nonlinear in time.

You'll notice that within the 3 years of the crime graph you've posted, the nonlinear components of the data generating process dominate the observed trend. You then assumed a strictly linear data generating process for 4 years in a similar setting for the purposes of forecasting; which is even more problematic.

You do not know what you are talking about.

Yes, but crime rate has not fallen in OTHER PLACES. — ernestm

Share some graphs of crime rates in other places please, I can only find US data.

Well thats interesting, thank you.

In California, San Jose rejects refunding in favor of ongoing reform bills in Congress, which is not surprising, because, the Speaker of the House is from San Jose.

Los Angeles has made a big deal of diverting $100~$150 million from the police to public welfare programs, instead of increasing the police budget by $735 million as previously planned. San Francisco has named similar intentions, but not stated specific results.

Minneapolis is reported to plan complete defunding of the police force and to be using the same model as the CAHOOTS program in Eugene, so I actually looked it up, and the CAHOOTS people specifically say they will not handle any criminal cases.



So if someone is in criminal danger, according to every fact we know now, Minneapolis plans to do nothing at all.
When Baltimore instructed the police only to respond to direct distress calls from victims, the murder and assault rates both tripled. I havent looked up Minneapolis' assault numbers, but I do know the murder numbers.

According to its homicide reports, it would be fair to say about 500 people will be murdered who the police would otherwise have stopped being murdered, of whom, about 350 will be black.

Ive not had a reason to do that before, all I know is about homicide rates internationally, mostly with relation to suicide and gun control.

A general graph of international rates would be fine, that’s what I was trying and failing to find.

Thats not necessary for my purposes, because suicide rates are my control group for measuring the effectiveness of gun control on homicide in the USA, there being a great deal of legislature to prevent mentai illness records being made available to background checks during firearm purchases for privacy reasons, I dont need an overseas control group.

So where is your claim that crime rates have not gone down elsewhere coming from?

If you look up homicide rates, you will find almost all the nations with higher homicide rates than the USA are in extreme political unrest and have no effective police force. Their trends are set by dictatorship takeovers, which happen too frequently for much statistical analysis of trends over time.

So no actual data on whether international crime rates correlate with atmospheric lead like US crime rates do?

They correlate with political takeovers. Which is pretty well what happens in the USA when Congress and President both switch from one party to the other. It doesnt seem very irrefutable to me, but you are welcome to look at the data yourself.

God he's making things up as he goes along.

If I produce a graph you just say my data doesnt matter, so why should I make any effort to demonstrate my knowledge by drawing graphs of it for hours on end?

You're a fraud, I don't give a flying hoot what you do, I'm just pointing it out.

Here is the international data from 2010, and if you have a point to make about lead poisoning or whatever, Wikipedia now has some data up to 2016. With best wishes.

Data across counties for a single year doesn’t help, we need data over time, for anywhere besides the US, or for the world globally.

The point of all of this is that there is already a well-known explanation for violent crime dropping (in principle anywhere, but definitely in the US) since 1990, one independent of anything to do with police or guns. That is a counterpoint to your claim that it went down in that time period because of increased police in that period.

You claimed that is hasn’t gone down elsewhere. (Which it should, if it is all about atmosphere lead). I can’t find any data on trends over time elsewhere. You presumably have some, if you’re making a claim about it. I’d just like to see it.

What I observe, having been working on these numbers since 2015, is that there appears to be a reverse in the downward trend since Trump took over, at which time there was also a GOP majority in house and senate. The last time the GOP lost control of all three branches of government, the homicide rate started declining instead of rising. So now maybe the homicide rate started to climb again in 2016, but the GOP lost the house in 2018, so maybe they were not in total control for long enough to effect a complete change in direction of homicide, and there's only data until 2018 now, so there is not enough data yet to know whether the homicide rate changed direction or not. Because I have also been watching the trend in change of the R2 factor, and from what I know, there is not enough change yet to make spline modeling more accurate.

Why are we talking statistics?

So you've got a model and expected variables that aren't in your model? In other words, you have a shit model. Cool.

Apologies for interrupting a spirited debate, and this may have come up earlier:

"Crime rates in the U.S. have fallen by about half since the early 1990s. A new working paper from the National Bureau of Economic Research finds that legalized abortion following the Supreme Court’s landmark Roe v. Wade decision in 1973 accounts for 45% of the decline in crime rates over the past three decades.

The paper’s authors, Stanford University economist John Donohue and University of Chicago economist Steve Levitt, take new data and run nearly the same model they used in their influential — and controversial — 2001 analysis published in the Quarterly Journal of Economics, where they first suggested an association between abortion and crime."

Interesting! Roe v Wade and unleaded gas started the same year.

Thats because, there is NO data on homicides since the GOP lost control of all three branches of government, and it never had control of all three branches for as little as two years before. So its unknown. Im not going to argue further for someone who calls my attempt to share knowledge shit any further. Piss off.

You went to Oxford? Fucking bullshit. Like, on a day trip?

I'm rather coming round to the consensus that you're a bullshit artist. And no, treating non-existent crime as the absence of 'would-be-criminals' is not a 'fact', it's a bloody insult to faculty of inference.

↪ernestm OK buddy you do you. — StreetlightX

:rofl:

My actual point, before I started responding to your delusions, and which has been totally ignored, is that ~500 people on Minneapolis are likely to be murdered, who would not otherwise have been murdered due to closing down the police. That was based on prior evidence when the murder rate tripled in Baltimore, and that tripling of murder rate, was after the police were only told not to stop crime unless explicitly asked by a victim. Minneapolis is planning to do no more than send a counselor and medical evacuation team, and that, so far as currently stated, only to events that are not known to be criminal.

That means at least 350 black people will die because of what 'black lives matter' wants in the next two years.

What i said was, in my first post on the topic, when I try to raise concern about black people being killed, I am ridiculed as an over-privileged white animal with a useless education from Oxford, instead of anyone being concerned about black people dying at all.

And all of you have proven my point. Thank you very much.

As 'Black Lives Matter' has also threatened to flatten the country if even one more event is deemed unforgivable by them, it is rather uncontroversial now that it may have started with good intentions, but now is a national terrorist organization, and just like terrorist organizations do, it is exploiting naivety and ignorance as a recuitment tool, deliberately ignoring facts that are contrary to its mission, while insisting on the rules it wants, and disregarding the actual death and suffering it is causing.

And as its proponents have done nothing but ridicule the fate of those who will be murdered without police intervention, that its motives and yours have nothing to do with saving black lives, and you merely seek political power and furtherment of your own agenda, without any real concern for negative consequences at all.

Distracted by low hanging rotted fruit.

'Black Lives Matter' ... now is a national terrorist organization — ernestm

Ah, I see, well at least the cherry picking, flawed reasoning, unjustified speculations and invented stories of persecution had a purpose. Maybe we can get back on topic now.

I presented my logic. With that, no one has been able to say it is wrong except because they say so. The Black Lives Matter movement has threatened to flatten the country if there is one more event it does not like. Fact. People will be murdered without police intervention that Black Lives Matter ignores. Fact. Also, a white person was killed by a neckhold in Texas, about which Black Lives Matter has nothing to say at all. Fact. It is a racist, terrorist organization whose actions will kill the faction it claims to defend in Minneapolis alone, by at least two orders of magnitude more than lives it will save nationally.

Ok, thanks.