RK Crosby is not OK

After the surprise election of the Coalition government in Australia in May, the Director of the Centre for Economic Policy research made this rather harmless, but insightful scatter diagram of select demographics and swings.

Some potential drivers of swing to Coalition at electorate level - low income, low education, Christian religion. Age, rentals little correlation. Usual correlation caveats apply! pic.twitter.com/ZRwZPB7xj0

— Ben Phillips (@BenPhillips_ANU) May 20, 2019

I thought it was quite useful, and matched was was reporting elsewhere on how TPP swings to the Coalition occurred in electorates which were not going to benefit from their policies and swings against the Coalition were occuring in those seats targetted by Labor for their proportion of rent-seekers. Professor Phillips' remarks came as a useful contribution to the submission of the Isocracy Network to the ALP's Campaign Review in August.

Not everyone was happy however, RK Crosby, who leads the voterchoice website, came out swinging.

This is stupid and irresponsible correlation that will be misreported and misunderstood. Correlation is never causation. Delete this nonsense. If you want to demonstrate causal links, spend the money and do a panel study. Seriously @ourANU, get some standards.

— RK Crosby (@ktxby) May 24, 2019

Sure, a panel study would be good, but it is a bit much to describe Professor Phillips' correaltions as "stupid and irresponsible", let alone to demand "delete this nonsense". But being of a philosophical bent, one statement stuck out as amusing tangent in the history of philosophy.

> Correlation is never causation.<

Ahh, the return of David Hume and billiard balls. To paraphrase: "No matter how often one moves after being struck by the other, we cannot say for certain that the 100% correlation of movement is caused by impact".

But it'll do for physics.

— Lev Lafayette (@lev_lafayette) August 8, 2019

Hume is correct of course, kinda-sorta. Whilst Hume is being a determinist, he is skeptical of our abilities to determine causality from reason alone. Instead, we must develop a system of probabilities derived from experience. Thus if one billiard ball strikes another and it moves 100% of the time, that correlation is not causality as such, but rather a certainty. Technicaly, it is possible that some other (unobserved or even unobservable) force is at play. Correlation is never causality, but it'll do for physics.

RK Crosby however, was not up for any amused banter.

I’m not sure why you are replying to a tweet from May, unless there has been some movement in university land to apologise to the academics ‘counselled’ for making similar observations as I, but well done for demonstrating Hume’s point that causation is not well understood.

— RK Crosby (@ktxby) August 9, 2019

Huh, one may have thought that "voter's choice" would know that Australia's largest political party was having a review of the election results.

Maybe if I say that I'm agreeing with them?

FWIW Hume argued that humans can't understand causation at all; all is correlation.

"only experience which teaches us the nature and bounds of cause and effect, and enables us to infer the existence of one object from that of another" (EHU ch 12)

— Lev Lafayette (@lev_lafayette) August 13, 2019

Oh please, yes, would you continue trying to assert some kind of nerd dominance? Mansplain my own area of expertise to me some more? I presume that’s why you replied to a random tweet from May, to swing your big nerdy dick around and fuel your ego...

— RK Crosby (@ktxby) August 13, 2019

OK, that's a somewhat unexpected reaction when you're agreeing with someone.

Are you OK?

— Lev Lafayette (@lev_lafayette) August 13, 2019

Apparently not.

Oh well, their profile does describe themselves as a 'migraineur'.

As for nerds, we may add:

... And a good nerd accepts the corrections because it's not a competition between people.