Tuesday, August 25, 2015

Q&A - 25/8


Musicians, writers, and other creative folk are still scratching their heads over the cover story in Sunday’s New York Times Magazine: “The New Making It” — packaged online as “The Creative Apocalypse That Wasn’t” — looked at how the Internet economy, instead of destroying creative careers, had redrawn them in “complicated and unexpected ways.”[..]

Some of the statistical stuff is wonky and hard to explain succinctly. But one of the key objections comes from Johnson’s claim that more people are making their livings as musicians. But since the statistical categories were changed midstream to accommodate school teachers, the early numbers and the later numbers become an apples-to-oranges comparison: the numbers of working musicians, it appears, is not going up but down.

All Models Are Wrong, Some Suck

Bruce B. de Mesquita

If Hillary Clinton was nominated by the Democrat Party for 2008, she would have been elected as President.


The Abramovitz model agrees: Republicans were going to lose 2008 no matter what. That means Democrats could have nominated a tree trunk for 2008, and that would have been the next President of the United States. No offense to Hillary, she would have been great prez, but that's what happened.  


Democracy is a luxury, can only work in rich societies, it is not a necessity.

Most likely untrue

In The Logic of Political Survival BBM makes the opposite case where he uses the winning coalition W as a measure of better governance. According to this research increases in W effect GDP growth positively. Here is the causation part - the book calculates increase in W based as difference between W of two years ago and calculates its relation between the current year's growth.

I was able to replicate his results (see the notebook from the previous link). [geek] the only part I am not too crazy about is R^2 being too low, but the model overall is significant [/geek]. Change in W's coefficient is at 1.8, this means going from lowest to highest W adds over 1% to GDP growth. Blimey. The effect does not necessarily go the other way; that avenue is explored in the book as well.


Why is the situation described here by Fukuyama not cool? He said "[Nigeria is] divided into more than 250 ethnic and religious communities that do not want to work with one another. Their ties are instead vertical, to clientelistic networks controlled by the elites, who dole out just enough patronage and subsidies to mobilize support at the next election".

Vertical divisions decrease W

Let's imagine country Z has 9 ethnic minorities with near equal members. Each has its own party, blacks, whites, Asians, Hispanics, so forth.. Every black votes for the Black Party, Hispanics for the Hispanic Party, etc and after the election, 5 out of 9 parties form a coalition. Sadly in this case the winning coalition can only revolve around whoever is deciding the leaders of the parties - because people, by sticking to their own party, are not making a choice here, really. Then, we have low W, low W means low growth.

A brief seque here: I'd like to expound on the previous two-party idea we mentioned before. Let's say the Chinese wanted to go multi-party. Their leaders are tired of this dictator bullshit, they don't want to party like it's 1899. Here's how to do it.. In 6 easy steps! Call now! Kidding. But here are are the steps: shut down the Communist Party (it had a really effin  stupid name anyway given that China today is anything but Communist), create two new parties, distribute the old geezers among both. Have quotas for each minority in each party exactly mirroring their percentages in the greater population. Then maybe have a symbolic president who has some amount of power, whatever.. Then have an election with these two parties. Boom. Then watch the money flowin'

♩ ♪ ♫ Of Gucci and Prada I gotta lotta  
♫ Got a Mac 10 and some Pina Colada 
♩ ♪ Drinkin, countin the green, chillin with my man Benjamin ♩ ♪
♩ ♪ Drivin by lookin chill, this brother's all fly ♩ ♪

I even wrote the song for it. C'mon

Q&A - 21/5

Question How do you empirically prove interest rates do not cause increase or decrease in GDP growth? There is a test for that Data ,...