The first thing to do is: Tax "bads" instead of "goods". By that we mean that if you're going to tax something, get the incentives right and tax things like greenhouse gases (CO2, methane, etc), toxins (mercury, dioxin, etc) and things that are locally scarce (water in Las Vegas, for example). Don't tax things that you want more of, like labor, capital gains, investments, etc.The goal here is not to take more money from people. In fact, if we can take less, all the better. The goal is to shift taxes in a revenue-neutral way, so that if X amount of money is needed for some program, it can be raised by de-incentivizing people from destroying the planet without discouraging work and investment (these things aren't bad in themselves, they just have bad side-effects when we don't put a price on things like ecosystems, CO2 and toxins). Aligning people's self-interest with the planet's is the best way to have everybody make good choices, even people who couldn't care less. Right now, people's wallet and the planet are not aligned; you either pick the cheaper product, or the green product. The framework needs to change, because we doubt it's possible to convince everybody to think before they consume.
The second thing to do is to reform and streamline regulations that slow down the development of green technologies. We're specifically talking here about regulations that have been adopted "for the public good" in theory, but that were in fact lobbied by big players to keep competitors out. A good example of this was given by the founders of Google: They had to deal with endless bureaucracy and red tape when they installed their 1.6 megawatt of solar panels. A smaller company - or an individual! - could never have made it through all the delays and extra expenses. This biases the system in favor of the dinosaurs.
The third thing is to level the playing field by removing subsidies. All subsidies would be best. Dirty industries get a lot more of them than clean ones anyway, making the green underdog artificially less competitive with tax dollars. Even subsidies that seem quite good at first usually later turn out to be bad (corn ethanol) because of the law of unintended consequences, and governments are very bad at picking the right technologies because of political interference, pressure by special interests and multi-year lag in decision-making. On top of that, even the really good subsidies are usually temporary and uncertain, making investors wary to place big bets.
If we really must spend tax dollars, lets do it at the R&D; level and make the findings available to all via public domain licensing. That will speed things up immensely and lower the costs of going green for everybody.
We can have a strong economy that brings people out of poverty and reduces human suffering and a healthy planet.