In a prior post, I did a back-of-the-envelope calculation of the savings we could get by getting rid of drug patents and instead publicly funding drug development. There would be other differences from the status quo, not relating to the money, though.
Enables more kinds of drugs
There are classes of drugs not getting currently developed because the expected return from their on-patent sales would not provide good ROI over the expected development costs (including crazy-expensive clinical trials). New antibiotics provide a good example - lots of generic competition, and we as a society want new antibiotics to be used as sparingly as possible to slow down the evolution of resistant bacteria, so that means the market will be small.
Rare diseases have traditionally been a problem, but the government has been putting some funding into drug development for these, so more have been forthcoming.
Removes incentives for "me-too" drug development
The go-to example here is of course TRICOR II. In general, funding drugs via patents encourages a sort of Minimum Patentable Increment. Though see here for discussions of the upsides of me-too drugs.
Positive spillovers from cost savings
This is related to the fiscal savings, but the resulting lower costs to drug users would lead to better access and better health. A recent working paper made a splash showing how increased drug cost-sharing leads to usage cutbacks without regard to necessity. Better population health will help the economy, but no way am I going to try to figure out how much.
Most negatives I can think of are predictable consequences of involving politics in the funding.
Government funding can be unreliable; among other reasons, sometimes the Federal government is controlled by people who don't want its programs to work.
Drug classes may be unpopular with some constituencies: those who believe hormonal birth control to be abortifacient might be able to block any research into that area, or those who believe fat to be a moral failing (most everybody, unfortunately) blocking research into conditions caused by obesity. Is allocating research based on percieved virtuousness of the affected better than allocating research based on expected ROI of treatment? Hard to say.
If contracted out, we could end up right back where we started at a public-private partnership, though with somewhat more competition. Barriers to entry would be high (as in defense contracting), no doubt.
Was there supposed to be a conclusion? I think the positives outweigh the negatives enough to support trying it, but of course I would.