This is backward, the point of a patent is that it guarantees the invention will become public knowledge. The exclusivity period is compensation for that.

There is no law forcing a person or company to file for a patent. They are free to keep their trade secrets a secret forever, if they want.

This may be backwards, but it's how regular people see it, from what I've observed. They dream of patenting something and then licensing/selling it to companies. It's easier than making use of the invention by themselves.

As for guaranteeing the invention will be public knowledge, these days it's just bait-and-switch.

Using an invention to create a product is a way (maybe the most popular way) of making the invention public knowledge.

The dream of licensing is not a perversion of the patent system, it's part of the point: to get inventions out of people's heads and secret workshops, and into forms that benefit the public at large.

Why are pharmaceuticals an exception? What makes them different?

Not the OP, but the vast majority of the up front cost is in development. Reproduction is relatively cheap, so who's going to pour money into R&D if your formula is just going to be immediately copied and sold at a fraction of the price?

But isn't that the purpose of all patents? To promote innovation by protecting the output of up-front R&D spending from just being immediately copied by competitors?

Yes, on paper. The problem is in current practice. In most fields, patent system extends too much protection for the patent owners, allowing them to block an idea from being used by others for far longer than it's needed to recoup development costs and earn a solid profit on top of it.

Also, you forgot about the other part of the purpose of patents - to allow the public to benefit from the innovation, by making it public domain after patent protection expires. Increasingly, patents don't hold that end of the bargain, by being deliberately written in such a vague form as to be useless as a blueprint.

Back to the topic of pharma - this is the only case I know where I feel that cost/benefit of patents favors keeping them. Everywhere else, I feel things would be better if we abolished them, or at least severely reduced their duration.

Absurdly high costs of drug research (and then absurdly high costs of getting successful ones approved). We're talking about billions of dollars of investment for something that, once developed, can be copied for pennies by cheap chemical plants. I accept that there needs to be a way for "big pharma" to recoup the development costs, or else they won't even bother with drug research. It seems to me that cost/benefit of patents make them justified in this case. But not in most other industries, not with the patent lengths of today (and lack of any details that would let people reproduce the invention after patent expiration).

That same argument can be used for other inventions. Why is pharma an exception just because you added "absurdly high" to it? It is possible to fund pharma research through the government just like how much of current pharma research and plenty of other types of research are funded. And the good thing about being fully government funded for pharma research is that you get the sell the resulting medicine for very cheap.

> Why is pharma an exception just because you added "absurdly high" to it?

Yes, precisely that. If you see the problem of patents as a cost/benefit analysis and not black-and-white issue, it's not surprising that the system may be socially useful in some sectors, but not for others.

As for alternative systems of funding drug research, this is a separate topic. I'd all for abolishing pharmaceutical patents if a different effective method of funding would be available.

EDIT and lest you think I'm a fan of big pharma business, I'm not. I abhor the trickery they pull with marketing to doctors, or flipping an inert fragment of a molecule and patenting it again. But I'm inclined to cut them some slack as the society needs new drugs.

Maybe the solution is to just make patents more expensive to obtain. That way companies won't bother patenting trivial "inventions", while products that require lots of resources to develop would be worth the extra cost.

This would disadvantage individual inventors though, who would be less likely to afford the application than a company.

My current and somewhat layman opinion is that we need to take a serious look at the duration of patent protection. For instance in software, a patent longer than 2-3 years is IMO absurd, and just serves to enrich the patent owner while retarding the progress of technology. Similar calculation could be made for other industries, based on typical time-to-market and rate of progress. The application process should IMO be also more strict, to a) not let through so many patents on obvious things, and b) ensure that patent description actually allows to reproduce the invention. Lastly, we need to find a way to stop companies from stockpiling patents as if they were nuclear weapons.

And speaking of that, I wish regulators could also take a look at the duration of copyright protection, and curb the practice of extending it in perpetuity.