The one thread that runs through all these discussions is that "true costs" are not well represented, or captured at all, by market prices.
Let us review the various examples of this.
The cost of carbon emissions, and the consequence threat of global warming, is not reflected in the price of carbon fuels. It is just shifted off to a great hereafter.
The danger of nuclear proliferation is a potential cost for building nuclear reactors that is not factored into the price, but it might prove to be the most significant cost of all.
Adverse consequences for the balance of payments, excessive dependence when oil prices are low, a huge transfer of wealth to petrostates when prices are high, the vulnerability of the economy to sudden disruption via an "oil shock"--all these are relevant in assessing the long term costs of oil dependency that are not registered in the nominal price.
The costs of maintaining expensive armed forces to guard over, patrol, or otherwise dominate oil-bearing regions are also very real, but are not reflected in the market price. They are a cost of extraction paid for by public authority (and, so long as they achieve their ostensible object, also a hidden subsidy to consumers of oil everywhere).
No modern economy can function without markets, and no approach to energy can be successful that does not give "market forces" a wide berth. But there are basic environmental, strategic, economic, and political costs that hover round the energy market, not registered in the price, which any rational public policy should attempt to register.
In principle, the conclusion is unavoidable that government action is needed to provide rational pricing for those "externalities"--rational in the sense of reflecting our estimation of costs not captured by the market. Has government done this well when it has tried? Sometimes yes, sometimes no. Is the assessent of costs surrounded by considerable uncertainty? Of course. But somehow we need to figure out a way to get our best estimate of real price registered in the market price.
An evocative portrait of the long term costs of gasoline consumption is given by the Canadian scholar Thomas Homer-Dixon: "Every time we push down on the accelerator pedal, we emit a blast of carbon dioxide that contributes to global warming. Our children and grandchildren will pay for this warming--in the form of higher food prices from drought, heat waves and floods, greater health expenses from diseases that thrive in warmth, more property damage from storms and rising seas. Those huge future costs aren't reflected in today's gasoline price. In effect, our children and grandchildren are subsidizing our current mania for driving."
The idea Homer Dixon expresses here has long been registered in environmental economics and is the foundation of the field. You don't have to accept as inevitable the consequences Homer-Dixon foresees in order to accept the main point. In principle, there are obvious and not-so-obvious environmental costs associated with energy use, whether we attach greater importance to the short term or the long term. His general analysis, moreover, applies also to costs associated with military spending (counting part of that as a cost of extraction from the Persian Gulf) and to the other examples we have mentioned. In all these areas, government action is needed to provide rational pricing for those "externalities."
In the environmental field, as Homer-Dixon notes, economists have two bright ideas for going about this: "We can apply green fees or taxes to reflect a product's environmental harm, or we can create a market for nature's environmental services that we now treat as free."
More on the advantages and disadvantages of these approaches later.