[throwaway35777]

Don't successful drugs also have to pay for the failed trials?

[jrsdav]

This is essentially true. Pharma is incredibly expensive (for lots of different reasons), with R&D taking up a huge portion of those costs.

So yes, it's safe to assume that part of the accounting around those published costs in the billions are all of the failed candidates that never even made it to trials (the failure rate varies depending on the area of biology and the type of drug, but it's generally around 9 out of every 10 candidates [1]. By the time you get to trials, that ratio gets even more abysmal).

Disclaimer -- I work for Recursion, a company built around this very problem.

- [1]: https://www.sciencedirect.com/science/article/pii/S221138352...

- [2]: https://www.recursion.com

[mc32]

R&D takes lots, but so does compliance --for good reason. But compliance costs a lot of money, directly and indirectly. Lots of people, lots of inefficient processes, etc.

[tomrod]

Why do the phase testing not prevent overindexing failed projects?

[darth_avocado]

Yes. They also have to pay for the $4.2B the companies spend on lobbying efforts in 2023.

https://www.nbcnews.com/investigations/lobbyists-spent-recor...

[ProjectArcturis]

Seems like a lot, but in the context of $200B spent on R&D, 2% spent on warding off confiscation is probably money well spent.

[refurb]

Uhhh.. $4.2B is all lobbying, not just pharma?

[kurthr]

And also for failed executive pay, and failed lobbying, which are much more expensive. In some cases (like Aduhelm) the marketing started before it was shown to have efficacy... so you have to pay for that very expensive failure too.

https://www.washingtonpost.com/business/2024/03/08/eli-lilly...