Besides, assuming the device doesn't self-destruct Mission Impossible style, at worst it'll still be worth parts to someone. So there will be many more parts available--and possibly more victims to make up the difference (assuming muggers are working for something).

So, a couple of theft cases to frame: 1. People steal fake phones (shiny objects) that have NEVER worked, ie aesthetic value only. 2. Parts, such as replacement screens, are in great demand, and there are small urban retailers today that enjoy every opportunity to lower their costs, ie $5-50.

Following 1991 crime statistics, over 2/3 of those arrested for robbery tested positive for a controlled substance. Does an addict have Free Will? Delayed gratification? What's a $100/day habit like? What is an addict willing to do for a Klondike Bar? (Answer: Anything.)

That is clearly wrong, but it is not the assumption. The "assumption" (and not a very big one) is that a kill switch will substantially reduce the value of the device, which will reduce the incentive for theft. And that's almost certainly the case.

"Following 1991 crime statistics, over 2/3 of those arrested for robbery tested positive for a controlled substance."

What percentage of the population at large would "test positive for a controlled substance"? Particularly in the demographics in question (young and risk-taking)? That's not a terribly meaningful figure. Some portion of these crimes doubtless are drug related, some portion of those crimes would likely still occur. So? If we can get rid of some other crimes that's still a boon (and may, as I've said, nonetheless not be worth the trade-offs).

"Does an addict have Free Will?"

In a meaningful sense, less than the rest of us, but that doesn't mean they're going to react the same regardless of circumstances. If robbery is working less well to get them their fix then it certainly seems like they might be less likely to pick robbery next time. For those that do respond by engaging in more robberies to pay for a fixed cost habit, they are more likely to get caught as they are forced to attempt more difficult opportunities and leave more of a pattern.

Flip it around. Say phones were cheap, and someone was proposing something to artificially make them more valuable to criminals. Would you think that likely to make the number of robberies go down, up, or not move?

I think if we were actually serious about victims, instead of using that warrant for a Kill Switch, we'd use it to query the networks and track the breadcrumbs; it's even richer than metadata sourced from billing systems. (It works for finding missing hikers.) And I'm pretty sure existing law covers this already.

"I think if we were actually serious about victims, instead of using that warrant for a Kill Switch, we'd use it to query the networks and track the breadcrumbs"

I don't know that I'd object to that. I'd still demand an owner-provided code. Note that I've never been saying "this system is amazing and exactly what we want", I've been saying "there is a legitimate up-side" - which I stand by, and which the existence of better alternatives don't undermine.

And people carry much of value besides electronics today. You can use NYC seizure records and ATMs as proxies.

Of course something is better than nothing in the one dimension where it's something rather than nothing. I've never been arguing that this is the best solution or even a desirable solution. I've been stating (repeatedly) that there's a dimension (that doesn't involve being recipient of the contracts or similar) along which it's a positive thing. This was in answer to the strong statement, "How do people see this as a good idea? There are already voluntary methods for this, so there is literally no good justification for forcing a state-controlled version on everyone."

"why not pay robbers not to rob?"

Depends what you mean. A benefit solely for people who would otherwise rob seems obviously impractical - how do you verify that 1) they would rob absent the money, and 2) they refrain from robbing after you've given the money? Something broader like a Basic Income is actually something I favor, though I'm not certain that it will have a tremendous value in reducing robbery (though that's certainly conceivable).

"And people carry much of value besides electronics today. You can use NYC seizure records and ATMs as proxies."

You'll have to expand on this. Note that my argument was not "no one ever carries anything of significant value", but that the typical (black market) value of what's carried by even wealthy individuals has fallen if you exclude electronics.

If we unpack "nothing" further, we'll find any number of 'somethings' that are worse than what you're attempting to measure and affect. That's what we call Unintended Consequences, and not simply tautology or Begging the Question.

It is far from clear that this measure will have any significant affect on crimes such as robbery. And if we begin to unpack the "something" we quickly get to questions like, "Is it really a good idea to design a method for centrally destroying critical communication infrastructure?"

Following your logic, shouldn't we expect more robberies to make up for lost revenue?

If you're still in search of "cause", I'd suggest that the rate of robbery is inversely proportional to the typical waist size over time, i.e. we're wealthier now than we were.