Our models are too primitive at this point to give us any more than an educated guess that humans are actually causing global warming.
Meanwhile, you're asking for a sh!tload of expensive change and real sacrifice on the basis of that guess.
For example... let's say we replace all the cars on US roads with SmartCars and Tata Nanos. Based on the crashworthiness of those vehicles, we can reliably estimate that such a move will cost us X number of additional highway deaths every year, even as it saves us Y amount of gasoline and Z amount of air pollution.
How much Y and Z does there have to be to justify a given quantity of X?
If you're not prepared to answer this question, you're not prepared to make a policy change.