Correct me if I'm wrong, but if demand for a product goes down, shouldn't prices go down along with demand? Not only that, but if they're selling and producing less product, wouldn't their overhead to extract and transport same decrease, as well?He tosses it over to King Banaian, who offers a first approximation at resolution.
What he's just identified is the stranded asset problem (.pdf), in which capacity built with the expectation of continued expansion in a regulated environment has little or no going concern value in a deregulated environment with policies that encourage conservation.
Those utilities were encouraged to expand energy production. An unregulated monopolist makes money by lowering the amount of goods and services it provides and thus pushing up the price of its product (for instance, newer sports facilities often have fewer seats than those they replace.) That is what maximizes profits -- lower quantity, higher prices. Regulators hold prices down and encourage the monopolist to provide more. But because of environmental concerns, we've pushed energy distributors in the opposite direction.
In return for submitting to regulation, energy producers are guaranteed that the physical plant they build will receive a normal rate of return. Again, when the goal was to get them to produce more energy, we would do this by permitting some small passthrough of the costs of construction to rates. But not too much, because we want access to electricity and heat for all. But now because of conservation measures the access problem can be solved with less physical plant than was there before. "So shut the inefficient, dirty producing facilities down!" you might say. But the regulator promised them a fair return on the investment, and that may mean the old plant has to stay open longer.
There's a further problem, which is that the regulated monopolists often operate under conditions of natural monopoly. Thus the title of the post. Economies of large scale are sufficient but not necessary for subadditivity, and subadditivity is sufficient for natural monopoly (which might have to be regulated unless some rigorous regularity conditions hold.) This site is more accessible to readers less well versed in theoretical industrial organization although it has some major errors. There's a proper theoretical treatment here (.pdf) that requires a good background in math.
Psycmeistr's problem is worsened under subadditivity. Suppose people buy less natural gas, such that the output under conservation Q1 is less than the previous output Q2. It then follows that Q2 = Q1 + Qd. Under subadditivity, the cost of producing Q2 in a single firm, C(Q2) is less than the cost of producing Q2 in two firms,including the partition C(Q1) + C(Qd). If we stipulate further that the subadditivity is a consequence of increasing returns to scale, C(Q1)/Q1 > C(Q2)/Q2, and the gas rates have to increase to reflect the higher cost per unit that accompanies the smaller output. (In practice, the "overhead" to transport and distribute gets allocated to fewer cubic feet of gas.)