"No retreat laws" and race
A while ago I posted regarding the Tampa Bay Times' claims that Florida's no retreat law was racially discriminatory because conviction rates were higher when a minority member was shot. I concluded, from reviewing their raw data, and looking at the defender rather than the person killed, that blacks and whites had almost exactly the same exoneration rate, and hispanics had one much higher.
Economist John Lott has used a much more sophisticated means of analyzing the issue. The Times also gave other data on each shooting, such as was it a home invasion, etc., whether the person shot was armed, whether they initiated the confrontation, etc.. He ran multiple regression on that to essentially ask "in identical situations, what are the odds of conviction? Note that he's taking the Times' own test -- ethnicity of a person who was shot. He concludes:
"Everything else equal, in cases with only one person killed, killing a black person, rather than a white person, increases the defendant's odds of being convicted, though the result is not statistically significant. If you also include multiple murder cases, killing a black person increases the chances of conviction even more.
These regressions also show that white defendants are more likely to be convicted than black defendants, and both effects are significantly greater than for Hispanics.
Whether the person killed initiated the confrontation and having an eyewitness were the most important factors in determining whether there was a conviction."
He also notes:
"Blacks may make up just 16.6 percent of Florida's population, but they account for over 31 percent of the state's defendants invoking a Stand Your Ground defense. Black defendants who invoke this statute to justify their actions are acquitted 8 percent more frequently than whites who use that same defense."