But is it a solution?

If Andre gets out of prison and goes back to stealing, then that process may well have wasted more money than many social programs for the exact same outcome. Imprisoning people is really expensive.

If Andre never gets out of prison at all, then we're committing to indefinitely spending a ton of money on him for the rest of his life, which also is not an amazing outcome. And that's only a theoretical possibility anyway, selling stolen goods (even habitually) is probably not going to get you a life-sentence in prison.

I think people have this perspective on incarceration that it's the final catch-all solution, and the reason we avoid it is purely compassion. And compassion is part of it, sure, but also if prison sentences don't reduce recidivism, then they're not a working solution. You should look at prison recidivism rates and apply the same level of skepticism and hold prison spending to the same standards that you apply to social programs and public housing. You should have the same expectations for both public housing and for what is essentially a much more expensive form of "coerced housing".

Asking how far we should go to reform someone before we give up and incarcerate is kind of begging the question -- it assumes that incarceration is always solution that just blanketly works and the only reason we avoid it is out of compassion and morality. In reality, incarceration is just another tool in the toolbox, and sometimes it might solve a problem, and sometimes it won't -- sometimes it's a complete waste of time and money.