Timeline for Is this chart showing the likelihood of a terrorist attack statistically useful?
Current License: CC BY-SA 3.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 5, 2017 at 10:27 | comment | added | matt_black | The chart is not about terrorism in general but terrorism committed by refugees. Its purpose is to demolish the public perception that refugees are a big terror risk not that we shouldn't worry about terror in general. In economic terms, specifically targeting refugees makes essentially no impact. | |
| Feb 3, 2017 at 17:51 | comment | added | spraff | We should not worry disproportionately about terrorism, and we shouldn't invest disproportionate resources into it. What we're doing at the moment is ridiculous. You may have a brilliant grasp of statistics, probably better than mine, but you're displaying a lack of understanding about how security works. | |
| Feb 3, 2017 at 16:21 | comment | added | Aksakal | Following your logic we should not worry about terrorism, and shouldn't put much resources on it. This is ridiculous. Look at any country with high terrorism activity, their economy and quality of life gets impacted in major ways. The reason why terrorism has low fatalities is because we fight with it in active manner. The tactics and strategies of terrorism and counter-terrorists change all the time. It's quire reasonable to assume that terrorists might target refugee programs to infiltrate our country. | |
| Feb 3, 2017 at 14:29 | comment | added | spraff | Your assumption that $f$ is a decreasing function is wrong. And to the extent that your analysis is correct, $\frac{f(x)}{x}$ small doesn't necessarily imply anything, but nevertheless it is quite apparent to me that, in the world in which we actually live, the US spends epic sums of money on the premise of fighting terrorism, to negligible or even negative benefit. Do you honestly think we couldn't save more lives by diverting some of those dollars to healthcare? | |
| Feb 3, 2017 at 13:18 | comment | added | Matthew Gunn | That's not quite right; you have an average vs. marginal issue. Let $\alpha$ be the statistical value of life. Let $f(x)$ be the number of expected deaths when $x$ is spent (where $f$ is a decreasing convex function). Cost-benefit analysis would lead you to solve $\max_x -\alpha f(x) - x$. Solution is $-\alpha f'(x) = 1$. The average benefit of your spending is $\frac{\alpha \left(f(0) - f(x) \right) }{x}$, the marginal benefit is $-\alpha f'(x)$. And your notion of the average $\frac{f(x)}{x}$ is something else too. $\frac{f(x)}{x}$ small doesn't necessarily imply that $-\alpha f'(x)<1$ | |
| Feb 3, 2017 at 12:17 | comment | added | spraff | @whuber I never said US spending had no effect, I'm saying that the terrorism casualties are so small and US spending is so large that we can safely conclude that US terrorism spending is, at best, less efficient than spending on alternative life-saving strategies. | |
| Feb 3, 2017 at 11:49 | comment | added | DrMcCleod | Your answer points out why 'deaths' is not a good metric when measuring the impact of terrorism. The cost of terrorism - in loss of social trust, reduced liberty due to increased security and, occasionally, war, far outweighs the (generally) small number of people actually killed. | |
| Feb 3, 2017 at 6:45 | comment | added | Matthew Gunn | @whuber I agree that you want to avoid statements on cost-benefit analysis of existing measures because you don't know what the counterfactual is. You perhaps could question the benefit of additional measures by questioning what the additional measures would be reducing (we appear to be fairly effective in disrupting plots...). | |
| Feb 3, 2017 at 0:22 | comment | added | whuber♦ | The fundamental (and obviously incorrect) premise behind this response is that the US spending had no effect. It's like concluding that incorporating engineering safety tolerances into bridge designs is worthless because bridges don't collapse. (I'm not necessarily disagreeing with your conclusions, but only with how you appear to misinterpret these data in order to reach those conclusions.) | |
| Feb 2, 2017 at 23:22 | history | answered | spraff | CC BY-SA 3.0 |