Comment by AnthonyMouse

You can argue "harm" about anything. Especially if you include risk of harm. Under that model it would be valid to prohibit driving at any speed, owning a steak knife, etc., because there is some risk that activity poses to others.

The fundamental problem is proportionality. And the reason it's a problem is that perfect enforcement is impossible. The actual harm posed by driving down a residential street at 60MPH is not proportional to the existing penalties (i.e. reckless driving, jail time, suspension of license). You could do it a hundred times and likely nothing bad would happen. The problem is that a) if you do it 50,000 times then likely someone will be injured, and b) you can do it a hundred times and likely not be caught by law enforcement.

So here comes the trouble. If you set the penalty equal to the risk-adjusted harm (i.e. penalty for 60MPH in a residential zone is a $25 fine) then it serves as no deterrent because you'll only be caught less than 1% of the time which makes the risk-adjusted penalty 100 times less than the risk-adjusted harm. However, if you adjust the penalty to account for the low risk of getting caught, you now have a penalty which when imposed is disproportionate to the offense by a factor of 100 or more. Which is problematic when the effectiveness of enforcement increases without adjusting the penalties. And extremely problematic when the reason for enforcement ineffectiveness is that enforcement is expensive or unpopular, meaning that someone targeted for investigation (e.g. for political reasons) could be found violating many such laws in a way that wouldn't be financially or politically viable to investigate every member of the general public, and then be subject to 100X the proportional penalties for the offense(s) because the penalties were set under a set of assumptions about the effectiveness of investigation and enforcement that are invalidated when the investment of investigative resources is politically motivated.