Here's a back-of-the-napkin calculation. Figure a month at least is necessary for an addict to get the help they need at one of these centers, and 20% of people need a return visit. That means the average number of days is:
0.8*30 + 0.2*60 = 36 days
Second, the data from the link below seems to suggest that as much as 7% of the population admits to illicit drug use in the past 30 days (we'll call these regular users).
We'll also say that at least a quarter of regular users have an actual "problem" that needs to be addressed by the war on drugs.
Given the current population of the U.S. at ~300M people, here is the math of people with problems:
300M * 0.07 * 0.25 = 5.25 million people
At $100 per day for 36 days, it would cost the following to "solve" the problem:
$100 * 36 * 5.25M = $18.9B
We can just round up to $20B. That's the good news, because it seems like if we passed a bill to send the country's people with drug problems to a rehabilitation center, it may be affordable to do so.
The bad news is that it would likely cost many times this amount to A) find the 5.25M people who need it, B) bring these people to court so that the law can be enforced, and C) make sure that there are follow-ups to ensure that the drug problem of each individual has actually been addressed.
I don't have a figure handy for this, but it is legitimately part of the solution. And while I'm not a member of the CBO, I think the figures mostly say that the treatment is conceivable, but the execution of it into law is the intangible part that would be extremely hard to do. We also have history to tell us that drug users aren't so easily found, and they surely won't volunteer to get treatment without the law being enforced.
Market Data 
Markets 
AI
