Q24. WHAT ARE FILL FACTORS?
A24. The FCC defines a "fill factor" as the proportion of a network facility that will be filled with network usage.12 In the context of a TELRIC cost model, a higher fill factor translates into lower average estimated costs. However, that is not necessarily true in the real world, because higher fill factors themselves impose costs on a system and these costs should also be accounted for.
The FCC's TELRIC methodology requires that unit costs be derived from total costs by using "reasonably accurate 'fill factors' (estimates of the proportion of a facility that will be 'filled' with network usage)" and, more specifically, notes that unit costs can be derived "by dividing the total cost associated with the element by a reasonable projection of the actual total usage of the element."13 If this is not done, and, instead, if unrealistically high fill factors are used, one would also need to recognize and account for the other costs that would be increased as (real world) fills increased.
Q25. PLEASE PROVIDE AN EXAMPLE OF WHAT A FILL FACTOR MEANS IN THE REAL WORLD.
A25. In the real world a firm typically carries spare capacity, which means that its plant is not running flat out at all periods of time. There are a number of reasons that spare capacity is efficient and necessary in a telecommunications network. First, it is entirely unrealistic to suppose that a real-world firm will know with 100 percent certainty what or where its actual future demand will be. In economics, there is a principle attributed to Nobel Laureate George Stigler that a real-world efficient firm will incur costs that are higher than what it might otherwise incur if it were to build a rigid, inflexible plant that is incapable of responding efficiently to changes, variability, and uncertainty in the market.14 A real-world firm will incur higher costs to build an adaptable plant that can accommodate changes in the economic or market situation. This principle applies not only to spare capacity needed for flexibility in accommodating uncertain growth in demand, but spare capacity is necessary to accommodate variability in demand, with or without net growth. Airlines, for example, historically have run on average at 50-70 percent fill factors (or what is called in that industry "load factors"). Airlines maintain spare capacity not necessarily because they expect demand to grow next year, but because demand is higher on Friday evenings than Wednesday mornings, and because some Fridays are simply and unpredictably busier than others. Similarly, networks must maintain spare capacity because it is impossible to predict which households will demand new or additional lines.
For example, consider two Sacramento suburbs, Elk Grove and Carmichael. It could turn out that 30 percent of households in each neighborhood demand a second line. Or it could turn out that 60 percent (or 100 percent) of households in Elk Grove demand second lines, and none do in Carmichael. Of course, all scenarios in between are possible as well. Planning for 30 percent in each suburb will not suffice to handle the possibilities (and statistical likelihood) of unbalanced patterns of demand, because plant installed in one area cannot be used to serve demand in another. To handle at least some of those possibilities in an efficient way, more than 30 percent spare capacity must be installed in each neighborhood if 30 percent additional lines are expected on average.
A TELRIC model in which the engineering specifications are based on an assumption that the network is built to precisely meet a pre-determined demand, without considering how to accommodate future moves and changes, violates Stigler's adaptability principle. The costs that need to be incurred to flexibly accommodate uncertainty are costs that are part of a forward-looking, efficient network. An efficient provider knows that its demand will materialize over time (not all at once) and that end-users may move around, thereby requiring new capacity in one geographic area while capacity is unused in another. The efficient provider will account for these factors in its capacity planning decisions.
Efficient forward-looking firms utilize spare capacity as a way to hold down other costs, manage risk, and maintain service quality. Accordingly, spare capacity is a legitimate, economic, and efficient cost of doing business.
Q26. WHAT ARE SOME OF THE OTHER REASONS FOR HAVING SPARE CAPACITY?
A26. The following represent some examples that are based on issues of (1) an economic tradeoff between inputs; (2) technological considerations; (3) service quality; and (4) prudent response to risk.
· Economic Tradeoff Between Inputs. It would be unduly costly to install distribution plant overnight at new locations whenever such new plant is called for, with the thought of coming back again in the near future (and opening up a street perhaps) to install additional plant - all in the anticipation of running at a higher average utilization rate during a particular time period. The placement of new plant involves such tasks as siting, obtaining permits, trenching (or placing of utility poles), and other construction activities. Some of these costs will have to be incurred again and again every time new plant is called for in a particular location.
· Technological Constraints or "Breakage." Technological constraints affect the opportunity set that firm managers have to select from and therefore affect costs. For example, I understand that telephone distribution plant is "lumpy": It is available only in a finite number of sizes. It is, therefore, infeasible to obtain physical maximum capacity utilization from lumpy capital except in the rare instance where demand exactly matches the physical supply. For example, if the efficient number of distribution lines in an area were 80, but the smallest size of cable that provides 80 lines were 100 pair cable, engineers might reasonably install the 100 pair cable, resulting in 20 lines of spare capacity. This spare capacity is not the result of inefficiency nor can one expect it to be eliminated going forward. Rather, breakage is a cause of spare capacity that will result in any efficient network subject to technological constraints.
· Service Quality Considerations. One finds, in most industries, that customers have to wait for the product or service they want. Anyone who has ordered from a catalog understands the meaning of "back-ordered." However, back ordering is not considered an acceptable option in local exchange telecommunications service provisioning. To keep quality up, telecommunications carriers must have on hand sufficient capacity, in advance of demand, to meet expected demand. The costs of advance capacity are a necessary part of the "ready-to-serve" obligation and so are costs of providing service at current quality levels.
12 Local Competition Order, ¶ 682.
14 George J. Stigler, The Theory of Price. (New York: Macmillan Publishing, 1966), pp. 130-131.