for the system. Finally, our assumption of one additional subscriber in areas yet to be wired for every mature subscriber before extant plant is a maximum estimate given the franchises acquired. The total number of homes in franchised areas not yet wired by the firms acquired in these 17 transactions is estimated to be approximately 700,000 from 1970 Census of Population counts. Even if all of these homes are wired eventually, they will not equal total homes subscribing at maturity in currently wired areas. Moreover, our calculation requires that these homes be wired instantly--otherwise, the required rate of growth to justify the system purchase price must be higher.

Unless cable owners are willing to accept a lower rate of return than 15 percent on capital before income taxes, we must conclude that their actions reveal that they are expecting perpetual growth of net revenues per subscriber in excess of 8.5 percent with no increases in cost. This increase is likely to be realized in the form of new services which add to both costs and revenues, but we cannot on the basis of current evidence estimate the precise magnitude of each during future years.

This 8.5 percent growth rate for net revenues is approximately consistent with a rate of growth of gross revenues of 3 to 4 percent per annum with no cost escalation. Given the built-in wage escalators in our operating-cost formulation, we shall take 4 percent as the maximum growth rate of total revenue per subscriber, but we present estimates for a more modest growth rate of 2 percent as well.

E. RESULTS OF CALCULATIONS

In this section, we calculate the internal rates of return on cable systems in different operating environments under the revised assumptions described above. We utilize Mitchell's assumptions and data for all but the following parameters:

1.

2.

Penetration at Maturity - a 0.648 in the top 50 markets,

All operating costs and revenues are discounted from the middle of each year. Capital costs are discounted from the beginning of the year in which they are incurred.

3. All distribution plant investment is spread over two years-

50 percent in the first year; 50 percent in the second year.

4.

Underground cable percentage 5 percent in all markets. Twenty-channel capacity for all systems; even those outside the top 100 markets.

5.

6.

7.

8.

Exclusivity Effects - Two additional independents imported via three microwave hops each as standby capacity to substitute for blacked-out independents.

Origination All systems have standard origination as defined by Mitchell.

Annual Revenue per Subscriber - $67.92 which grows at a
rate of 0 percent, 2 percent,
and 4 percent in three

different calculations.

9.

The number of subscribers increases at an annual rate of 2 percent in years 16 through 60.

Clearly, our assumptions on mature penetration and subscribe revenues are more generous than Mitchell's and they lead to higher calculated rates of return. However, as we have argued in Section C, these assumptions are more consistent with cable system owners' projections and currently reality.

We utilize the cable system owners' projections for penetration in 1982 for our first calculations, employing the datum for markets 51-100 for all markets outside the top 100. These calculations embody more optimistic penetration assumptions for the top 100 markets, but in some cases we are more conservative than Mitchell for the smaller markets. As an alternative projection, we utilize the pessimistic assumption that, despite growth potential, penetration in the top 100 markets will remain 24 at 55.3 percent.

For the path toward maturity, we utilize the growth path used by Mitchell despite its lack of statistical foundation. We feel that this path is a reasonable compromise between Comanor-Mitchell and Park. In addition, it is quite similar to the one utilized by Park in his Dayton25 Miami Valley predictions.

Revenues per subscriber take three forms--all beginning with $67.92 in the first year. In the first variant, it is assumed that there is no net growth. In the second, revenues per subscriber grow at an annual compound rate of 2 percent while in the third they grow at 4 percent. These rates of growth are consistent with our observations about prospective subscriber fee increases and future sources of ancillary revenues. More importantly, they are conservative reflections of revenue growth 26 discounted by buyers and sellers of systems in the past two years.

Equally important and also leading to higher rates of return are our assumptions concerning cable investment and its timing. While we use Mitchell's data on costs per mile of each type of cable plant-

24 Don Andersson, op. cit.

25L. L. Johnson, et al., op. cit., Paper #2.

26 See Section C above for an estimate of growth rate anticipations.

aerial and underground --v
-we phase the investment in that plant over two
years in accordance with our observation that few plants are ever
completed in a single year. In addition, we allow for a maximum of
5 percent of cable underground for a typical system given the obser-
vation that few systems currently under construction or in operation in
major markets have more than this percentage of their plant buried in
underground conduits.

Minor changes involve exclusivity and origination. We simply allow the importation of two additional standby independent signals to offset the effects of exclusivity protection afforded local stations. While this may not allow the cable system to prevent blacking out one channel at all broadcast hours, it will greatly offset any effect of exclusivity protection upon subscriber penetration. The cost of importing each additional signal is included in the additional three microwave hops per station required.

For origination, we utilize the Mitchell "standard" origination-requiring capital costs of $38,000 and annual operating costs of $43,000. Mitchell, on the other hand, utilizes a minimum origination expenditure for systems of fewer than 10, 000 mature subscribers, but we exclude these smaller systems from our results.

A few minor differences exist between our calculations of operating and capital costs and those of Comanor-Mitchell upon which Mitchell relies. All of these derive from the difficulty in translating the Appendix description of cost parameters in Comanor-Mitchell into actual cost data, For employee benefits, office rentals, and house drops, our cost data differ to a minor degree with the Comanor - Mitchell calculations. These differences are minor and have no perceptible effect upon calculated rates of return.

Otherwise, we utilize the Comanor-Mitchell cost data for

the calculations reported below. Instead of replicating the plant over

an infinite horizon, we simply allow for four generations of 15-year plant life. This reduces the rate of return to a very minor degree given the present value of net revenues realized 61 or more years from the present.

In order to demonstrate the effect of alternative proposed copyright fee schedules, we calculate the rates of return under the four different assumed copyright fee schedules employed by Mitchell:

represents no copyright fee payments;

#1

These four alternative fee schedules are detailed in Table E-1.

*S. 644 relates fees to subscription revenues. For simplicity, we have related the fee formula to gross revenues. This will tend to understate slightly the calculated profitability of the systems when copyright fees are included.