Tapsearch Com EditorRevisiting incarceration
June 23rd 2007 21:13
Warning: This is a nerdy stats post.
Awhile back I blogged about international incarceration rates. I got some data -- diversity, murder rate and democracy index ratings -- and found the very odd fact that diversity and murder are not correlated with each other. However, both are correlated with incarceration rate.
This suggests that diversity makes people throw each other in prison more, but it doesn't actually cause more crime in and of itself. (Of course, in any immigration situation, this leaves the issue of where the diversity is coming from.)
Recently I downloaded Gnumeric, a free spreadsheet program that makes multiple correlation a breeze. This further confirms the results:
--Diversity, murder rate and degree of democracy together explain 44 percent of the variance in incarceration rates. Individually, they explain 21, 24 and 14 percent respectively.
--Diversity and murder rate together explain 42 percent, so degree of democracy doesn't mean much once these two factors have been controlled.
What I wonder now is -- is this just an oddity in the data set, or is it true that diversity increases incarceration without increasing crime?
New spreadsheet here.
Awhile back I blogged about international incarceration rates. I got some data -- diversity, murder rate and democracy index ratings -- and found the very odd fact that diversity and murder are not correlated with each other. However, both are correlated with incarceration rate.
This suggests that diversity makes people throw each other in prison more, but it doesn't actually cause more crime in and of itself. (Of course, in any immigration situation, this leaves the issue of where the diversity is coming from.)
Recently I downloaded Gnumeric, a free spreadsheet program that makes multiple correlation a breeze. This further confirms the results:
--Diversity, murder rate and degree of democracy together explain 44 percent of the variance in incarceration rates. Individually, they explain 21, 24 and 14 percent respectively.
--Diversity and murder rate together explain 42 percent, so degree of democracy doesn't mean much once these two factors have been controlled.
What I wonder now is -- is this just an oddity in the data set, or is it true that diversity increases incarceration without increasing crime?
New spreadsheet here.
| 34 |
| Vote |
Add To: 
