Beware of folks who start worrying about government debt when the Prez is a Democrat, and stop when it's a Repub at the helm. Then, it's politics, not economics.
Now, how would we control for the probability of it raining, and the expected amount? I mean, it would obviously have been much harder to convince the faithful peasants of Murcia – which is one of the driest areas of continental Spain with an annual rainfall of around 300mm – of the prayers’ effectiveness than those of Santiago de Compostela (1,900mm). OTOH, there’s not much need for a prayer in Galicia in the first place.
Or are the Cassandras of this (and any other) world exempt from such rigor?
I think you are overthinking it. People simply register whether it rains or not and because Murcia is so dry, rainfall is much more appreciated than in Galicia.
In my understanding, Granger causality isn't much more than a lagged correlation.
I know instantaneous Granger causality tests that check contemporaneous causality between two time series. Do they coincide with simple zero lag correlation tests?
It's a bit more. It boils down to lagged correlation in one direction but not lagged correlation in the other. But that is unfortunately as good as it gets with causality tests...
The instantaneous Granger causality tests I don't know, so I can't comment on what they do.
Beware of folks who start worrying about government debt when the Prez is a Democrat, and stop when it's a Repub at the helm. Then, it's politics, not economics.
Hmmmm… call me a shameless sceptic now, but how’s this for a possible explanation of cause and effect:
1. No rain falls
2. No rain falls some more
3. Desperate people start praying
4. Mean reversion happens to some degree and rain falls a bit
Not dissimilar to some people’s belief in the RSI
¡Por el amor de Dios!
Now, how would we control for the probability of it raining, and the expected amount? I mean, it would obviously have been much harder to convince the faithful peasants of Murcia – which is one of the driest areas of continental Spain with an annual rainfall of around 300mm – of the prayers’ effectiveness than those of Santiago de Compostela (1,900mm). OTOH, there’s not much need for a prayer in Galicia in the first place.
Or are the Cassandras of this (and any other) world exempt from such rigor?
I think you are overthinking it. People simply register whether it rains or not and because Murcia is so dry, rainfall is much more appreciated than in Galicia.
I do, but only the Friday posts … 😜
I have a question for Klement the statistician.
In my understanding, Granger causality isn't much more than a lagged correlation.
I know instantaneous Granger causality tests that check contemporaneous causality between two time series. Do they coincide with simple zero lag correlation tests?
It's a bit more. It boils down to lagged correlation in one direction but not lagged correlation in the other. But that is unfortunately as good as it gets with causality tests...
The instantaneous Granger causality tests I don't know, so I can't comment on what they do.
Thank you very much