1. Demand from end users appears to be driving satellite technology toward the provision of services directly to end users (e.g. direct to home (DTH) television broadcasts) rather than the historical emphasis on network connectivity through large earth stations.

  2. Whether rights to use GEO slots should be a consideration in the restructuring debate depends, inter alia, on the level of scale economies in operating a global satellite network. If the minimum efficient scale were 25 satellites, then a case could be made for letting INTELSAT retain its slots, and for regulating it as a natural monopoly. However, if the minimum efficient scale were four or even eight, then a better regime might be divestiture of excess satellites and slots to promote competition among multiple providers. The fact that private satellite systems are proposing to operate global systems with far fewer satellites strongly suggests that scale economies may not be very significant.

    A different type of "scale economy" might arise from the need to coordinate positioning in GEO orbit, coordination that may be smoother within an existing system than among different heterogeneous systems. This issue perhaps merits further investigation.

  3. Drawing an analogy from the U.S. telephone industry, when the integrated telephone monopolist AT&T was broken up into a long distance provider and seven regional local-service Bell companies, competition became the vehicle of choice for keeping prices low in long distance service. To guard against Bells abusing their market power and charging excessive interconnection rates, the FCC regulates these rates. An important difference in the world arena is that there is no global regulator (nor would we advocate one), and individual countries may be tempted to charge excessive access fees (if they act in their national rather than global interest). Such excessive fees could harm foreign countries in two ways: by reducing the profit that foreign telecom suppliers might have earned, and by increasing the cost of global interconnection, thereby injuring also foreign consumers.

    But INTELSAT's role in mitigating such incentives is uncertain. It would seem preferable to rely on explicit agreements between countries that ensured satisfactory access terms to all, perhaps within a multilateral context such as the General Agreement on Trade in Services.

  4. To be sure, rents could arise from two other sources: INTELSAT's greater efficiency, or economies of scope. The latter, if important, argues against the viability of setting up a separate subsidiary.

  5. On the other hand, if all cross subsidization between a proposed restructured "core INTELSAT" and its affiliate(s) is prohibited (and assuming such a prohibition is enforceable), it is not obvious what is gained by retaining the affiliate status rather than creating completely separate private entities.