Say the government decided to let Wal-Mart park delivery trucks on your property for free (i.e. decline to enforce your property right in favor of a business). Economically, it'd be indistinguishable from a subsidy. Shareholders don't care whether the government policy increases Wal-Mart's revenue or decreases its expenses. The same is true for power generation.
But with the level of compensation people think energy producers should pay to offset their "negative externalities" -- if governments forced them to pay it, the price of energy would rise by a huge amount, disrupting the economy, and making the previous calculation of the externalities meaningless.
Even the renewables would then need more subsidies, to pay for increased manufacturing costs and the fuel for the trucks that repair their transmission lines.
Point is its obviously misleading to say that fossil fuels are more heavily subsidied than renewables. People who hear that will assume that fossil fuel companies are being given taxpaper money, as the renewable companies are, when that is not the case. I don't think lumping together "companies that receive free money" and "companies we think should be fined" is an honest way to frame the situation.
When the true costs of coal are not being paid by those that use coal based electricity the economy is already being disrupted. More coal than would be economically efficient gets consumed because coal based electricity is artificially cheap. Implementing pigovian taxes repairs the existing disruption.
"People who hear that will assume that fossil fuel companies are being given taxpaper [sic] money, as the renewable companies are, when that is not the case."
That is in fact the case. Here is just one component of those direct subsidies, as reported by the U.S. Treasury Department:
But then every negative externality not taxed is subsidized. It dilutes the meaning of subsidy because with this new definition government accounting doesn't make sense - the government is 'paying' trillions to subsidize all kinds of negative externalities (but it never actually does).
Also, if the government actually subsidizes a negative externality, you're in a thicket, since some of it is a real subsidy and some of it is the estimated negative externality subsidy.
