Earlier quoted context omitted.
Because when you have your target as zero then when you don't hit zero you push for ever more costly and suboptimal policy to achieve your goal. This is a problem both in that the incremental policy changes are highly unlikely to provide a larger benefit than their incremental costs and because it draws resources away from tackling problems that probably have a larger societal impact and smaller costs to achieve thos…
That makes no sense and I'd be interested in examples supporting it that are not some person on the fringe shouting opinions. What should the goal of the police department regarding killing innocent people? Should they aim for a dozen per quarter? Per-capita weighted? Incremental changes have been hard because it gets down to the police policing themselves, and that has shown to be a failure. Again, one of the "simpl…
All of these things have a pattern in common where the majority of the improvement can be had reasonably cheaply and capturing the last small improvement has huge costs that are usually much more costly than the benefit of eliminating the last bit and the population of potential events is huge. Road deaths are an illustrative example. Adding speed limits, traffic lights, crosswalks, moderate enforcement, etc. all dropped road deaths precipitously in the USA. If the USA changes its model from cars first to people first there are still some moderately costly infrastructure changes they could make to get their numbers down to Netherlands levels, which are still not zero. Beyond that you are left with hugely costly heavy enforcement among steadily more dystopian and invasive methods that will likely improve the number but not get it all the way to zero. It is really unlikley that these last improvements cost less than the benefit they incur between the direct cost of implementation and the negative effects on the total population of driving events.
On top of all this all those large costs you are incurring for very small gains come at the expense of being able to spend those limited resources on other things that likely have better cost to improvement ratios. You are foregoing a larger decrease in something else bad for society for small decreases in your target. In the road deaths example above that final small drop in road deaths might be coming at the cost of large drops in crime if you deployed those police there instead of on road deaths, as an example.