Hacker News Reader

Comment by invalidOrTaken | Hacker News Reader

The U.S. is a multicultural society. Aggregate measures don't tell much.

So is Canada -- more so in fact -- and yet...

That's my point though: and yet what?

It seemed to me that you were attributing the lower life expectancy due to US' diversity compared to many wealthy nations. My point was: Canada is similarly diverse and yet its life expectancy is higher.

No, I was saying the diversity of American society means that aggregate measures won't tell us much. If Canada is similarly diverse (that seems contested, but accepting it for argument's sake) then aggregate measures there also won't tell us much.

> So is Canada -- more so in fact

No, that's not a fact.

https://en.wikipedia.org/wiki/Demographics_of_the_United_Sta...

https://en.wikipedia.org/wiki/Demographics_of_Canada?#Visibl...

I'm really not sure what this means. Yeah, the U.S. is a multicultural society, but so what? Are we supposed to focus only on the certain cultures and ignore others when we talk about health and life expectancy?

I think I've heard this argument before in the context of gun violence in the US - as in, the US wouldn't have a gun violence problem if we excluded cities like Chicago and Baltimore from gun violence research. Is this the same basic argument?

It makes a lot of sense actually, a lot of this stuff comes down to what the smaller groups are doing. It's for the same reason no one cares about global gun violence statistics.

Talking about behavior patterns of US people like they're all part of a single nation stopped making sense long ago if it ever did.

It matters in discussion forums like this involving ordinary people. There are two reasons we might care about statistics and trends like this.

One, you're a public health official or maybe even just a voter trying to weigh various policy options or assess the success of what has already been done. In this case, sure, you have to care about national averages and other measures of central tendency and ways to characterize distributions.

Two, you're an individual American trying to figure out how much you should worry about an early death. In this case, national averages or even per-culture averages seemingly mean next to nothing for you. I can't speak for all Americans, but I expect to outlive the average by quite a bit.

For gun violence, it's often travelers and visitors worried they're going to get shot if they come to the US, and we're trying to tell them that whether they really need to worry heavily depends on where they're visiting. Even if they're actually visiting Chicago or Baltimore, gunshot deaths in tourist districts are damn near unheard of, and besides which, the vast majority of people ever killed by guns at all are either killed by themselves or someone who knows them. There aren't a whole lot of random stranger attacks happening anywhere.

It"s a racist dog whistle, not an actual argument

Per Google: "In the United States, life expectancy varies significantly by race and ethnicity, with Asian Americans generally having the highest life expectancy and American Indian/Alaska Natives (AIAN) the lowest. In 2021, life expectancy was 83.5 years for Asian Americans, 77.7 years for Hispanics, 76.4 years for Whites, and 70.8 years for Black Americans, according to KFF. AIAN populations experienced the lowest life expectancy at 65.2 years. These disparities are largely attributed to factors like socioeconomic status, access to healthcare, and the disproportionate impact of certain diseases like COVID-19 on specific racial groups. "

Facts have a clear racist bias.

Say Canadian Whites and American Whites both have mean lifespan of 81, and Canadian Blacks and American Blacks both have a mean lifespan of 71. Using that data and the fact that the US has 3X the proportion of blacks, you could calculate how much of Canada's higher life expectancy is due to demographics. It could explain the entire difference!

Alternatively, if Canadian Whites and Blacks live proportionally longer than American Whites and Blacks respectively, you would know that demographics could not explain the difference.

Hence the need for the disaggregated data.

I mean, okay, but what if black people live 10 years less than white people in such statistics?

That's still an abysmal thing, as the US would have something like 30% of their population getting clearly worse health and not caring or doing anything about it?

If the difference is down to demographics, surely that's really bad?

Maybe so, but touting Canada's "superiority" over the U.S. would lead to the "obvious solution" of getting rid of black people. If the difference is down to demographics, you want to know that, rather than look at people in country-sized blocs.

>Are we supposed to focus only on the certain cultures and ignore others when we talk about health and life expectancy?

If you want to get anything done, then hell yes! If life expectancy is low in Concord because of too many Doritos, but it's low in Phoenix because of a bad medical system, and it's mysteriously high in Fresno, you very much want to disaggregate. And you definitely don't want to ask "What is going so well in 'the U.S.'(Fresno)?" and then look at Concord.

Reply on news.ycombinator.com