The market doesn't always know what is right. To be more specific: the market is short-sighted and greedy, and can self-destruct an economy if left unregulated. We need to subsidize activities that have large costs in the short term, but return more over the long run to society. Or else they won't happen.
State funded 'generic' scholarships simply rise the cost of tuition. Good state schools are a much better use of funds as they drive down for profit education costs to be cost competitive. Schools can easily offer scholarships at minimal costs as a form of price discrimination, but public scholarships tend to simply result in administrative bloat.
Should we discontinue subsidies for solar energy companies?
Yes, instead we should tax CO2 production instead of subsidizing two industry's. see: https://en.wikipedia.org/wiki/Externality
How about the billions in R&D tax subsidies that we give to pharmaceutical/industrial/healthcare companies each year to develop new drugs, treatments, and machinery?
These should also be stopped, it's fine to fund public R&D but private R&D has terrible ROI. Drug companies spend more on advertising than R&D and really don't need help. For a solid example, the US could have a 10 year strategic food reserve for 1/10th the cost of current agro subsidies that directly result in massive health issues.
PS I agree that markets are often inefficient, but subsidies then to be horrible ways to correct them. Taxes tend to be much more efficient.
I won't address your specific points as everyone has different opinions on policy, but no one has the data to rigorously defend their opinions.
I simply want to refute the notion that we can allow free markets to allocate all resources within the economy.
Taxes are much more scalable than subsidies. AKA if A is better for society than B but B is cheaper to produce then taxing B can balance things fairly easily with consumers paying society for the external costs. But, subsidizing A is an unbound cost, adds administrative overhead, and promotes consumption of A and or B.
The downsides of Taxes is there often regressive, but for something like alcohol, pollution, or speeding societal costs are often independent from a persons net worth.
Sure, let me count them... OMG "It's Over 9000!!!"
PS: Ask a dumb question... http://dragonball.wikia.com/wiki/It's_Over_9000!
If you want more R&D, you should subsidize R&D.
If you want more people living in burn zones, you should subsidize living in burn zones.
I think I join most Americans in saying that I'd rather more people be educated, I'd rather have more R&D, and I wouldn't rather more people live in burn zones.
I'm sure other people have thought this through much better than what I just said. But it seems to me to be a reasonable argument.
If your goal is people living in flood zones, you should subsidize living in flood zones (e.g. by insuring it, which we also do).
We need an answer to the latter question in this case, not the former. And one is readily at hand: the political interest of those who live in flood zones. Maybe there's some other justification, but I'll confess I'm skeptical that this outcome is globally utility maximizing relative to alternative policies.
The efficient thing to do is end the subsidy, and have all the less valuable activities whose benefits don't exceed the costs relocate or cease.