The latest column from The Straight Dope exemplifies more or less how I feel about the environmentalist movement: We should all be looking for ways to improve the environment, but those little, easy things you do really don't make much of a difference. The stuff that matters is going to be the hard stuff that no one wants to do.
In this case, the column is about whether the switch from incandescent light bulbs to CFLs is good for the environment. While the dangers of CFLs were certainly overplayed, the fact is that in the end it makes very little difference.
There are plenty of reasons for this, and they are all more or less rooted in common sense. The main thing many people forget about is opportunity cost: by making a consciously "green" decision, you're only doing something that is better than the next worst thing. Another issue is the substitution effect. When gas mileage goes up, we don't burn less gas, we drive more; this cancels out the effect to the environment. Something similar may happen here (as pointed out in the above column): with worries reduced due to better efficiency, we'll just use light more often and be less careful about letting the porchlight on.
It's psychological, too; for example, I am a recycling skeptic. I think the costs are higher than the benefits (or at the very least neutral), but more importantly people who recycle think they are "doing their part," so don't engage in those things that could legitimately help the environment (and, of course, are more inconvenient than chucking a pop bottle in a green bin).
On top of all this, of course, is the ugly layer of government action: I don't think that it's right that the government bans incandescent light bulbs. Sure, they are inefficient, but to extrapolate the case that energy consumption is enough of a free-rider problem and therefore saving mere fractions of energy in the overall scheme of things is worth forcing consumers to a different kind of lighting is just wrong.
Since nearly all behavior-centric "solutions" to saving the environment will fall victim to this, my theory has always been that only a large-scale massive change in how we consume energy will help us. (Energy isn't the only environmental concern, but it's by far the worst offender.) I'm not saying all green initiatives will fail. Many newer building codes make economic sense, for example, since regardless of whether you want to save the earth, you still want to save cash on energy costs. And yet if they made economic sense, people would do it anyway--so why should the government step in?* Likewise, certain painfully wasteful activities have micro effects: X amount of chemicals won't hurt us, realistically speaking, but when they are concentrated in a specific location it's certain to damage the water supply. So this isn't a call to banish all environmental regulations: it's a call to 1) make sure a realistic cost/benefit analysis has been done, and 2) concentrate on the stuff that matters.
*I'll grant that there may be an issue of scale here. If a new environmentally-friendly device is developed but the costs are way too high to be economical, it's possible that via subsidy/tax credit/brute force, the government can get people to buy enough that the economies of scale kick in and it does, legitimately, make economic sense. I am torn as to whether this is a legitimate function of the government even though the end result is that of combating an externality; my libertarian heart says no, but my realistic side knows that a gentle ushering of "adopting" such things via tax credits to help cure pollution wouldn't be the worst thing in the world.