The January 1986 accident involving the space shuttle Challenger has prompted serious reconsideration of space transportation issues. Deliberations have ranged from rigorous scrutiny of NASA management policy to an in-depth review of shuttle engineering design. In the course of debate, several significant decisions have had to be made that involved matters fundamentally of economic choice. By year's end, while efforts to restructure NASA and resolve technical problems appeared promising, the decisions of economic import were less than encouraging. Moreover, they are likely to perpetuate a long-standing shortcoming of space policy: the failure to bring notions of efficient resource allocation to bear in managing space transportation.
Among the decisions made by late 1986 were approval to build a replacement shuttle for Challenger, and the go-ahead to begin research on a new, transatmospheric spaceplane. Taken together, these projects represent a vast commitment of resources for the nation's civil space program; the estimated $3 billion expense of the replacement shuttle alone adds up to nearly half of NASA's average annual budget in recent years, and development of a prototype of the new spaceplane is expected to cost anywhere from $2 to $17 billion. The higher spaceplane estimate is perhaps the more likely, for development of the shuttle required fewer major technical innovations in rocketry than are anticipated for the spaceplane, yet shuttle R&D expenditures, including interest, totaled $22 billion.
President Reagan also set forth a policy of rationing access to the shuttle. As tentatively outlined, shuttle use is to be primarily limited to military payloads, scientific and planetary research missions, and space station construction; commercial payloads—predominantly communications satellites—are to use conventional rockets, also known as expendable launch vehicles (ELVs). Two reasons underlie the rationing decision. First, NASA foresees a backlog of payloads extending throughout the mid 1990s, and decisionmakers generally agreed that some means of allocating access to the shuttle system is necessary. The schedule for the remaining fleet of three shuttles through 1990 cuts the number of flights by 80 percent from the pre-accident schedule. Even starting in 1991, when the replacement shuttle is expected to be operating, only two-thirds as many flights are likely. A secondary goal of the president is to nurture a commercial ELV industry by sending business its way.
Repeating History
This tripartite response to the Challenger accident—implementing rationing, embarking on supply-enhancing R&D, and spending to rebuild—is a familiar one which has brought adverse consequences in parallel settings. An instructive perspective is provided by experience with the energy crises of 1973 and 1979.
Analogous to post-Challenger deliberations in many respects, the response to sudden energy scarcity centered on non-market policy approaches. The strategy involved centrally administered gasoline rationing, the start-up of research in new synfuels technology, and, after the 1979 crisis, a rapid buildup of the Strategic Petroleum Reserve. Unfortunately, gasoline lines appeared in the Northeast, while the farm belt confronted a surplus of fuel; the Synthetic Fuels Corporation consumed a multibillion-dollar budget, but produced only a few successful projects; and in the haste to stockpile oil in the years immediately after 1979, large quantities of reserves were purchased despite the fact that oil prices were at all-time highs.
This experience with non-market solutions to energy problems gives substantial cause for concern regarding the desirability of a like prescription for space. To be sure, as in the approach to energy problems, all three decisions following the Challenger accident sought to promote the overarching objectives of ensuring national security and enhancing the economic wellbeing of domestic industry. Yet decisionmaking for space, as for energy, proceeded without due regard for incentives that can thwart the intentions of government intervention, and without full consideration of alternatives that might advance governmental goals more effectively.
Before turning to the 1986 decisions in greater detail, it is important to note that setting the stage for them were three years of heavily subsidized shuttle prices (prices charged to users). A series of pricing policies since late 1982 had resulted in subsidies averaging 50 to 75 percent of the estimated true resource cost of a shuttle flight. Moreover, the subsidized price had become so low compared to prices for conventional rockets that ELV assembly lines had virtually closed. In a policy announced before the accident but since rescinded, a certain number of flights for commercial use, beginning in 1988, were to be auctioned and to be subject to a minimum price per flight, but even this price floor would have represented a subsidy of 30 to 50 percent which would have continued to harm the ELV business. Just as important, the general sense of uncertainty brought by changes in government shuttle policy deterred longer-term plans for rocket production. At the time of the Challenger accident, then, these developments had circumscribed the potential role of ELVs as ready sources of supply and as providers of technological diversity to insure against the risk of space transportation accidents.
Projections of the overall demand for space transportation were also affected by shuttle subsidies. By January 1986 such projections reflected not only an enthusiasm artificially bolstered by the subsidies, but also a bias toward using the shuttle. Unfortunately, these estimates would play a prominent role in the post-Challenger debate.
The Opportunity Cost of Replacing Challenger
Against the backdrop of substantial dependence on the shuttle fleet, the Challenger accident has led to debate focusing on a "stockpile" argument reminiscent of that surrounding the Strategic Petroleum Reserve. Emphasis has been placed on the immediate need to begin construction of a fourth shuttle to back up the shuttle fleet in the event of future accidents. Not evaluated, however, have been alternatives that might function faster to rebuild space transportation. Such alternatives could have included contracting for the reopening of ELV assembly lines, funding the construction of additional ELVs instead of a fourth shuttle, and modifying payloads originally designed for the shuttle to fit on rockets.
Nor was the extent of the subsidy-induced bias of the demand assessments put to any test. Once information about tentative shuttle flight rates and payload capacities (required even for the rationing scheme) was available, all demanders could have been asked to bid for openings in the schedule. By thus revealing willingness to pay, a more accurate picture of total demand and the priority of different payloads could have been obtained. Such a bidding scheme could be manipulated by government agencies or private sector interests with deep pockets, but this problem also arises, albeit less overtly, with the rationing approach.
Any of these alternative actions might have more expeditiously and decisively demonstrated the need for a fourth orbiter. Instead, the central planning approach led to more laborious, and unconfirmable, estimates. After identifying major sources of demand, including the commercial communications satellite industry, NASA science and space station payloads, and Department of Defense missions, planners attempted to build estimates on the basis of information specific to each source. Aside from their subsidy-related overstatement, the estimates tended to reflect the limits, and in many cases the unavailability, of information.
The case of communications satellites illustrates these shortcomings. Projections used in deciding to build a replacement orbiter indicated that communications satellites would account for 40 to 50 percent of all civilian space payloads through the 1990s. Yet a confluence of industry developments at the time of the January accident strongly suggested that these numbers were too large.
The satellites being built and launched today—a process that typically takes several years from start to finish—had been planned in the heyday of satellite communications during the late 1970s and early 1980s. Then as now, communications companies were required to apply to the Federal Communications Commission (FCC) for allocations of broadcast spectrum and outer space orbital locations, and as proof of due diligence in spectrum and orbit use the grantees were also required to demonstrate commitments to launch their spacecraft. Down payments on launch reservations for future years were frequently submitted as evidence.
These down payments also formed a tangible basis for estimating space transportation demand. During that period however, demand was also affected by two other developments: fiber optics technology began to compete successfully in supplying many of the services provided by satellites, and innovations in satellite vi technology resulted in spacecraft with larger capacities. By early 1986 statistics for the communications industry were revealing that 60 percent of existing satellite capacity is unused; in fact, some corporate officials cited the hiatus in space transportation as a grace period during which overcapacity could be allowed to diminish. The "sunk costs" represented by payments for launch reservations thus became dubious indicators of actual business plans.
Demand estimates have been just as uncertain for scientific and Department of Defense missions, for reasons related as much to the use of ELVs as an alternative to the shuttle as to assumptions about overall levels of demand for transportation. Projections may well have been overstated by the extent to which shuttle subsidies were buried in intragovernmental transfers. The wariness with which scientists and defense planners regard the shuttle program has been a countervailing factor, however. Spending on r n spaceflight is often viewed as comPeting with budgets for space research, and 10 both budgetary and apparent technological reasons ELVs have received support from defense managers throughout the history of the shuttle.
The Space Science Board of the National Research Council has reported that of seven major unmanned scientific missions scheduled for launch aboard the shuttle in the near term, all but one (the space telescope) could be launched on ELVs. Similar arguments have been advanced regarding military payloads. Many experts claim that even space-based manufacturing operations could be automated, and Could therefore be launched by conventional rockets. That debates have been far from unanimous reveals unsettled judgements about the cost of designing payloads for ELV use, and argument about the longterm need for human involvement in space.
If shuttle fees had been more aligned with actual costs and in turn reflected in intragovernmental transfer payments, objective evaluation of these points of view could have been based on past scientific research and defense practices. Without knowing past practices, however, information on about demand for space transportation and the best mix of shuttles and ELVs is difficult to evaluate.
Another set of problems is presented by projections for shuttle requirements in connection with building and operating the space station. Budgetary circumstances since the Challenger accident have introduced new questions about the division of NASA resources between the space station project and space transportation. Consequently, as a fourth shuttle has been debated, discussion has also focused on a scaled-back design for the space station and on plans for a Shuttle II. This new launcher would incorporate emerging technologies to tailor it specifically to duties anticipated for the space station. Together with the uncertainty of other sources of demand for the existing shuttle fleet, these considerations add to the difficulties in centralized planning.
The Problems with Rationing
Complications in measuring demand and tile debate over the role of ELVs lead as well to two indictments of the decision to ration shuttle access through non-market allocation. First, and most important, the rationing scheme is likely to give rise to same set of problems in reconciling future space transportation demand and supply decisions as did shuttle subsidies. Second, neither the cost-effectiveness nor the fairness of the scheme is obvious. When payloads are shuffled, ranked first in priority are to be national security missions; ranked second are scientific missions, which are in turn ranked according to their scientific merits and astronomical "windows" (the required position of the earth relative to other planets at launch date). Communication satellites for the use of civil government are third in line. Commercial communications satellites, as noted earlier, are to be banned from the shuttle to bolster the ELV industry, although there are indications that the shuttle will launch certain foreign communications satellites. Access to the shuttle for other commercial activities is to be judged on the basis of their "shuttle-unique" launch requirements.
The effects of the rationing scheme will be difficult to quantify, but are likely to be substantial. Expected delays of several years in the launching of major planetary and other scientific missions will be costly in terms of both the additional funding needed to mark time and the harder-to-measure costs of setbacks in scientific advance. In particular, precluding use of the shuttle by the communications satellite industry may not lead to large net social costs in the near term, but over time could redound to significantly harm the industry, the ELV business, and consumers of communications services. The conspicuous lack of ELV business from communications satellites in the wake of the rationing decision attests to the declining demand for satellite services in recent years. The lack of transactions also calls into question whether the second goal of rationing—the assignment of communications satellites to the ELV industry to boost that industry's commercial success—is attainable.
In the long run, the doldrums in the satellite industry may dissipate. New satellites will be needed to replace obsolete spacecraft, and the industry may increasingly specialize in services that fiber optics cannot offer as cheaply. In this event, plans formulated by business for satellite launches in the late 1990s and beyond would be better served by a choice between ELVs and the shuttle (at unsubsidized rates). Competition between the shuttle and ELVs would benefit both the communications satellite and rocket industries, and would provide incentives to government to manage the shuttle program efficiently.
Finally, the interpretation, equity, and cost-effectiveness of the uniqueness criterion for commercial use of the shuttle may bring debate. As a result, development of private sector space endeavors may be delayed, expensive litigation may arise, and the difficulties of sorting out the different scopes of conventional rocket and shuttle transportation may be increased rather than alleviated.
The Spaceplane
The difficulties inherent in forgoing the use of demand and supply as decisionmaking criteria carry over to the decision to fund research on the spaceplane. As a detailed MIT study of the spaceplane makes clear, there is a menu of innovative possibilities for next-generation launch vehicles, including different propulsion systems, the extent of reusability or expendability, payload size, and degree of human involvement as a substitute for automation. In the aerospace industry trade literature there are numerous references to ELV modifications and to new ELV designs that would greatly expand the flexibility and scope of conventional rocket technology. By obscuring the choice among vehicles, however, both past shuttle subsidies and the proposed rationing mechanism threaten to obscure as well the most desirable directions for research in new launch technologies.
A Role for Economic Choice
The decisions made since the Challenger accident sidestep the most overriding need in rebuilding space transportation: a broader role for economic choice. It is not too late, as a starting point, to set shuttle prices that reflect real resource costs. By allowing realistic fees to govern access to the shuttle, the backlog of payloads would be sorted out more objectively according to merit and priority, resources would be redirected to the ELV industry on a more logical footing, and a more accurate indicator of research needs in space transportation would be provided. In tandem with this strategy is the necessity for commensurate intragovernmental valuation of shuttle and ELV services, to better enable space scientists and defense managers to trade off shuttle and ELV payload designs. Subsidies to space activities for such public benefits as planetary exploration or research in materials manufacturing could still be made, provided they are directed to the activities themselves rather than to a particular means of space transport.
To be sure, the implementation of sound pricing policy for government enterprise has a checkered past, and in the case of shuttle pricing may require radical attitudinal and institutional change. But the notion of using market-like mechanisms in allocating other public resources has begun to take hold; the Environmental Protection Agency's "bubble" policy for managing air pollution is a notable example of the willingness of industry and policy makers to explore the potential of economic reform.
The role of codified pricing formulas in public utility management is another example that could be brought to bear on space policy. Such formulas could play an integral role in the oversight of NASA or in the governing of a federal agency set up to operate the shuttle fleet (an organizational change considered by Congress). In either case, a formula for shuttle pricing could be based on an objective assessment of resource cost, and the advantages of economic over technological regulation could be used in shuttle management.
Federal budgetary realities also argue for a system of economically meaningful shuttle fees. One estimate of the societal cost of the shuttle program during its three full years of operation from 1983 to 1985 is $1.5 billion. This has been calculated as the difference between the resource cost of shuttle flights and the alternative cost of using ELVs (when possible) for a given payload. It is true but irrelevant that revenue from launch fees reduced this deficit, because the real loss is the excess cost of space transportation per se, not just the unrecovered portion of subsidies to the shuttle program. The amount is equal to half the replacement cost of the shuttle. It is also quite large in comparison with spending on projects for NASA's scientific research. In this light, the argument is indeed strong for bringing a greater degree of economic rationale to be in governing space transportation.
A version of this article appeared in print in the January 1987 issue of Resources magazine.
