Assuming this is true, which has to be assumed to give any credence because you give no backing material or references here whatsoever.
I have multiple and varied connections to the pharmaceutical, biotech, and medical devices industries (inside & outside, academic, research, financial, executive, etc), so I have a feel for the numbers. However, you can find confirmation of this in academic literature if you look. DiMasi asserts in a 2003 paper that the fully capitalized cost of clinical trials is 70% of R&D (and this number has certainly grown according to all trends), see: "The price of innovation: new estimates of drug development costs".
then I think such government regulation should be funded.
You are not presenting any concrete proposal nor any support for your vague notions, but it's a bad one in general. Government generally does a poor job at allocating capital (see: central planning) and has a serious problem with efficiency (see: motivating its employees, agency costs, interest group politics, etc). Furthermore, if you mean that industry would still retain a role, such an arrangement could create a massive agency situation whereby industry could submit multiple compounds for clinical studies and not have to pay the costs thereby making the system far more expensive for a society as a whole. There is also real expertise needed to design the trials and manage them which, again, government tends not to do well with pursuits like this.
In example I think "No Child Left Behind" was a dismal failure because it tested for poor school environments but did nothing to fund those requirements, or to catch schools up to standards.
I know it is deeply unpopular with the teachers unions, but this does not prove anything. Virtually every test score has risen or stayed the same. Do you have any objective facts to prove that on whole it was a net loss? Furthermore, with respect to the unfunded mandate stuff, that is a weak argument. It is not mandating an act--it's demanding baseline performance (and generally weakly at that). Given the dismal performance of US public school education vs most of the developed world and it's comparatively high cost, it's difficult to argue persuasively that waste could not be removed from the system by refocusing it on a more effective curriculum and reducing the vast amounts of administrative overhead. If anything it has failed to the extent that it allowed the states to control the tests and generally weaken the standards.
what's the alternative -- beta testing?
Fewer, smaller, and shorter clinical trials (that recognize that RCTs have huge limitations) combined with increased post-marketing (approval) surveillance. Namely, the cost per patient is so high and the availability of volunteers is so limited (this is a HUGE problem now) that it is impossible to conduct these trials with sufficiently large numbers of patients (sample size) to detect relatively rare and/or delayed on-set problems. With current phase III trials (average 3-4K patients) adverse drug reactions (ADRs) can only be detected that occur more frequently than 1:1,000. To have a 95% chance of detecting ADRs that occur in 1 or 2 in 10,000, you would need to enroll 600,000 patients. This is a number that is completely impractical yet would still expose half those patients to risk and the other half would just be getting placebos (a problem in and itself). Furthermore (besides raw statistics), these patients do not tend to be represent real world conditions very well (far higher compliance, generally healthier, fewer co-morbidities, increased observation, etc).
In other words, there is no substitute for actual real world use. In the meantime the FDA, which has some understanding of this problem, has every incentive to delay and restrict products from market and almost none to allow market entry (it's CYA). This problem has gotten increasingly worse over the past several decades and shows no sign of changing.
Nobody should do any such thing. I actually don't think software is dependent at all on such monopoly grants. Software patents, which are essentially patents on Boolean Math, are inherently counter-productive, and arguably a limitation on free speech rights. I think currently proprietary software isn't even all that dependent on copyright. Proprietary software is dependent on Trade Secret, since they only release binaries and hide the source. Even when they release the source it tends to be under NDA style contract provisions. Open Source also has better business models than Copyright entails.
The software industry is almost entirely dependent on various forms of IP and related forms of IP support. Copyright is most critical, but so to is trademark, patents, trade secret law, and various other protections (even government support for contracts can be a form of this). There are almost no software sales (not to mention development) in countries that have weak support for IP, even in relatively industrialized ones like China and India that are starting to move in this direction. The vast majority of paying customers just care about being able to use the actual object code as it exists, not to understand how it works or be able to make their own improvements. It is trivial to copy software and break any protection schemes (see: the internet, china, etc). IP-free regimes support very little ability for software companies to recoup their R&D and is no guarantee of open source code either (not to mention that most open source business models would be equally dead).
Furthermore, you are entirely mistaken, software patents are typically not on math or even basic algorithms as you imagine them (I actually am a named inventor on at least one). You might argue that all software is ultimately built on math, but that is normally not contained in the claims of most patents, and is equivalent to claiming that a patent on, say, an engine design is equivalent to patenting atoms (or that copyrighting a book is equivalent to copyrighting individual words). In other words, it's a non-nonsensical argument. That is not to say there are no problems here, but that your argument is very bad.
Why would I pose the idea of Democratic government taking over an industry if I thought it was inherently bad? You're making that claim, not me. And you have the burden of proof here -- the need for big pharma has never been proven, in that it has never been proven that similar or better treatments wouldn't arise without giant drug companies and monopolistic patents.
Actually you have the burden of proof, since you are arguing against the status quo (not to mention against common sense, decades of basic economic research, empirical comparisons with various socialist regimes, etc). Furthermore, you need to present an actual proposal, not just half-baked ideas in order to even begin to offer an argument. In addition, I have presented many arguments against your vague ideas and demonstrated flaws in your thinking.
I do not accept your arguments, and the burden of proof is with those that say patents are needed, or that the same or better results cannot be achieved through less exclusionary means. That is what has not been proven.
Again, the burden is on you since you are arguing against the status quo. However, if you an arguing that shorter patent lengths would work better, then you need to address the many problems I have presented at the start of this thread (e.g., the actual effective patent life today (post-marketing approval), the R&D cost, risk, costs of capital, S&M, etc).