public shootings study

[madd0ct0r]

ok. enough bitching and moaning.

let's do something collaborative.

I'll set up a google doc.

into this doc, we will put various data for the OECD countries and then run simple regressions to find correlations with other measures (ranging from crime rate to obesity rate).

Obviously correlation =/= causation, but we might find something intresting to argue about, certainly better then 'we don't need more gun control.' vs 'you don't need those guns.'

pm me your email address and I'll add you to the doc (i could make it public, but i expect it'd get vandalized pretty quickly)

What things would people like to test first?


OECD Current membership

Australia

Austria

Belgium

Canada

Chile

Czech Republic

Denmark

Estonia

Finland

France

Germany

Greece

Hungary

Iceland

Ireland

Israël

Italy

Japan

Korea

Luxembourg

Mexico

Netherlands

New Zealand

Norway

Poland

Portugal

Slovak Republic

Slovenia

Spain

Sweden

Switzerland

Turkey

United Kingdom

United States

[Skgoa]

IMHO the obvious question to ask are:
- does a (severe) economic downturn corelate with more/less shootings?
- does a better (i.e. more supportive) social safety net corelate with more/less shootings?
- does a better (i.e. socialized) health care system corelate with more/less shootings?
- does a higher emphasis on the rights of the child (i.e. ban of many forms of punishment, ban of home schooling...) corelate with more/less shootings?
- does a better education system (I believe there are annual studies by the OECD on this) corelate with more/less shootings?
- does a higher rate of (violent) crime corelate with more/less shootings?
- does stricter gun control corelate with more/less shootings?

[madd0ct0r]

I guess a lot will come from here:
http://stats.oecd.org/BrandedView.aspx? ... a-00590-en

ok. economic downturn can be measured as gdp growth rate, which can be got for most years for the past few decades from the world bank.

can't find a direct indicator for strength of social security net at http://data.worldbank.org/indicator - the OECD factbook has public and private social expenditure.

WHO had a measure of healthcare equity, but I can't find it on their new website. Government expenditure on health per capita is a poor measure for this as the USA government actually spends more then the British one does, and only covers a third of the population!
Maybe the ratio of public to private health expenditure? that would filter out bad socialized systems where most rely on paying for extra treatment (like america).

rights of the child.... not sure where to start looking for that.

better education system? - define better. the OECD factbook has a reading, science and math score for different countries. do we take into account average or standard deviation? (both available)

rate of violent crime - reported (to gallup poll) assualt rate ? - http://www.oecdbetterlifeindex.org/topics/safety/

strictness of gun laws - any idea where to start looking?

[The Grim Squeaker]

• No of events per year on average. (Reviewed per 15-25 years say) - (Useful to compare between countries. 25 Years is close to a generation, though you could use 10 if you feel lazy or micro for detail).
  Average Casualties per year. (Do you want to count killed or also wounded?)
  Average Casualties per Event (Number of casualties per event in other words. Interesting measure of lethality).
  HDI. (Human development index - A useful ballpark measure).
  GDP - Per capita.
  GDP - Normalized for relative income, Per capita.
  Ginni Index (Economic inequality - as a possible factor. Correlate with GDP).
  Literacy rate.
  Ethnic divisions/Heterogenity. (I'll eat my hat if countries that are ethnically diverse have Lower rates of events/crime than ethnically homogenous countries).
  
  Religion - % In the dominant religion relative to rest (Say 100% are religious. and 33% are Buddhists - that gives you 33% in the dominant religion). (This links to the whole ethnic/cultural heterogenity factor again).
  It might be possible to subdivide into Majority culture, if it's at least 70% of the population. (Culture on the level espoused by Huntington - 5-7 major cultures).
  
  Average Temperature. (Trust me on this )
  Abnormal hours of (or lack of) sunlight/daylight.
  Population Density.
  % Urbanization.

Those are all very easy to find and collate quantitive criteria. There's the CIA world factbook, but that's better for qualitive information.
Also,
http://www.imf.org/external/pubs/ft/weo ... index.aspx
http://www.indexmundi.com/g/g.aspx?c=xx&v=65

There's a book/site, that lists TONS of information about every nation on Earth, I can't remember it's name.. (I used it when working in International Relations)... Angus Maddison, it was on the internet, available in loads of formats (and super convenient for sucking loads of data about compared countries).

EDIT: Angus Maddison! That was the name - Quantitive Macroeconomics historian. http://www.ggdc.net/maddison/
Less useful than I remembered for something like this.

If anyone needs an explanation as to why the aforementioned factors are relevent, Feel free to ask, and I'll tell you why they are (or might be) .

[madd0ct0r]

hmmm. easy to find most of those, but incidents / events I've had trouble while looking for.

not sure how uesful average tempreture will be for somewhere like the USA

[The Grim Squeaker]

hmmm. easy to find most of those, but incidents / events I've had trouble while looking for.

not sure how uesful average tempreture will be for somewhere like the USA

Split it up. The USA is a damn continent. (You might as well lump all of Europe together, or all of Africa)

[madd0ct0r]

but then we'll have to split the other stats down for each state, which more then doubles the amount of work involved. (even if we do the larger areas, we'll need to build it up from state stats)

I did consider it too, but the shootings seem to be fairly evenly spread about: http://www.usatoday.com/news/nation/gra ... llings.htm
(2009 map)