Hacker News new | ask | show | jobs
by kafkaesq 3221 days ago
We have so many datasets available now that it's so easy to graph things against each other and notice things that may be coincidentally related, or even causally related, and then tell some kind of story about where the relationship came from.

Yeah, it's a fun website -- but you could say that about just about any claimed correlation.

Meanwhile, for those who have been around long enough to have a sense for the 'barometric' changes he's talking about -- the shift in the basic, underlying "we're-all-in-this-together" ethos has been not just noticeable, but profound. Probably a lot more work needs to be done to find a solid statistical basis behind this observation (if this is at all possible).

But by and large (aside from the singe n-gram example, which I agree smells like cherry-picking), it's not like his arguments are simply frivolous.

3 comments

> the shift in the basic, underlying "we're-all-in-this-together" ethos has been not just noticeable, but profound

Depends on your value of "we."

There is certainly lower solidarity/interest in public life among the straight white landowning Christian men from good families who used to participate in things like freemasonry and Rotary.

On the other hand, both the left and right have been adding previously marginalized demographics to their constituencies and platforms.

I don't think there's any less solidarity going around. My intuition is that more people than ever have their voices heard and interests represented in public life. But rather than one line of solidarity across the center of the political spectrum, we now have two disjoint lines reaching from each party's center to what used to be its marginalized fringes. (Not just in ideology, but in identity).

We also just aren't in this together so much anymore. The rise of coastal cities means the country is increasingly split between (at least) two totally distinct patterns of geography, economics, culture, built environment, and lifestyle. The correlation between population density and vote share on the latest electoral map is staggering.

> Meanwhile, for those who have been around long enough to have a sense for the 'barometric' changes he's talking about -- the shift in the basic, underlying "we're-all-in-this-together" ethos has been not just noticeable, but profound. Probably a lot more work needs to be done to find a solid statistical basis behind this observation (if this is at all possible).

I need to see some very solid evidence before I change my assumption, which is that people have always felt this as they get older, because we romanticize our memories of youth and we are uncomfortable with the societal changes that younger generations inevitably introduce.

Fair point - it's very tricky to measure.
I should also have been more careful to more clearly distinguish spurious correlations from correlations where both things are related to a third factor (which I think is probably often the case in John Robbins's graphs and somewhat plausibly here).