In most EU countries, the employee receives much less than what they cost the employer. In France, if an employee gets 30K euros, the employer has to provision ~45K Euros.
You can’t just recruit “200 homeless” and have it pass research standards. The homeless population is the most difficult of all to track, maintain accurate records, and even recruit for that matter. You think the homeless just line up for novel drug trials and report back for updates on a strict schedule?
You need good candidates for the trial. You need them to follow up. You need admins to properly track them and ensure it’s at least mostly accurate. Even the best trial candidates won’t follow the protocol correctly.
If you genuinely think the medical and pharma industry don't exploit homeless and other marginalized populations out the wazoo as lab rats. They also only stopped doing it to prisioners when it was made illegal, because that's what they did before. Of course using them has a lot of practical problems, as you outline, other than the much bigger ethical issue.
You managed to miss the whole point of the comment though.
Which is, how is it possible that a multibillion dollar industry, exploiting both the test subjects AS WELL as the researchers in the form of practically free PhD candidates can still claim it's one of the most dificult / expensive endeavours? Where does the money go? When you look, it goes to the bureocrats and to the "bio-investors".
You have to try really hard to make a video game no-one wants[0]. You might not recoup all your investment, but you won't sell zero copies. A drug can have all that money poured in, and nothing come of it.
[0] https://arstechnica.com/gaming/2024/09/two-weeks-after-launc...
If you have to pay 200 homeless to take your 0.0001% better than placebo antidepressants in the context of a huge corporation, and maybe redo the trial a few times, I can see how that gets expensive, but I don't see why it's a de facto rule.
But even the big trials it's weird how expensive people say they are. Most other products require a lot of high paid labor to produce, think of a video game studio for example, also without any guarantee it won't flop, and it certainly takes longer to develop than to do a clinical trial.