In the post-Cold War era, the justification for the space industry to continue to receive the large government budgets on which it has depended to date is being questioned. In particular, the value of government investment in the development of new launch vehicles when there is over-supply in relation to the world-wide launch rate of just a few tens of launches per year is questioned. Yet in the absence of a new generation of fully re-usable launch vehicles with much lower launch costs than today, the space industry has little possibility to grow into a commercially profitable and self-sustaining industry. Justifying such investment will require the development of new launch markets which have the potential to grow to traffic rates many times greater than the present limited demand for satellite launches.

One possible source of demand for future launch services, and so of future revenues and profits to finance the development of re-usable launch vehicles, is tourism - the making of short visits to low Earth orbit by fare-paying customers. In the modern world tourism represents one of the largest, most international industries, and it is growing fast, particularly in the more economically advanced countries. Furthermore, space travel has a high level of popularity among the general public. The desire of many people, particularly the young, to explore the unknown world of space and face new challenges seems likely to create growing demand for popular space travel. Thus it seems possible that the commercial demand for space tourism could be very high, and that the widespread desire of humankind to have the experience of going to space for themselves could become a driving force for the space industry in the future.

However, the attractiveness of "space tourism", or "orbital tourism" (at least in the early stages), as a potential target market for commercial companies will depend critically on the actual scale of future demand for space tourism services at different prices. To date very little data has been collected on this subject in any country. During the summer of 1993 a market research questionnaire on the potential demand for orbital tourism was completed by more than 3000 people from all age groups in Japan. The questionnaire was distributed face-to-face mainly in large companies, universities, schools, obtaining a response rate of well over 90%. Details of this survey, and initial results have been published in (Ref. 1), but the idea of space tourism was very popular. Nearly 80% of those under 50 said that they would like to travel to space, and 45% of those over 60, with no significant difference between the 1465 men's and 1565 women's answers. Some 70% of those who wished to travel to space said that they would pay three months' salary or more, and the great majority said that they would like to stay in orbit for a few days. In the following we discuss some further details of the results.

In considering whether a new service such as passenger flights to Earth orbit could become a profitable business, companies must estimate both the costs that they would incur in providing the new service, and the revenues that they could earn. The costs depend on how low engineers can reduce the manufacturing and operating costs of re-usable launch vehicles, and the revenues depend on the size of the potential market for the service, that is on how popular the service could become. Consequently in order to assess the feasibility of this business we need to do both engineering and market research. Currently the Japanese Rocket Society is studying the engineering and other aspects of space tourism in its Space Tourism Study Program (Ref. 2). The authors of this paper have been carrying out preliminary market research.

In principle, if firm data were available concerning the future demand for space tourism at different price levels, it would be possible to plan a business to provide such a service, provided that the vehicle costs could be reduced sufficiently. In practice, however, any conclusions that can be drawn from market research data on this subject can be only tentative, for a number of reasons. First, the reliability of all market research concerning future products and services is inherently limited. There have been many cases of new products' sales being very different from the results of market research, both greater or less. This is due fundamentally to the impossibility of predicting the future, and more particularly to peoples' inability to foresee their own future behaviour accurately.

Second, this uncertainty is particularly strong in the case of a new and futuristic service such as space tourism about which peoples' expectations are particularly likely to be mistaken. Thus the reliability of market research results must be more uncertain than usual. This is not to say that the results must be inaccurate, but rather that it is not possible to know how accurate they are.

Third, it is uncertain how representative the market research data that have been collected are of the Japanese population as a whole. The authors tried to ensure a wide distribution to people outside the space industry. As discussed in (1), by distributing the questionnaire on a face-to-face basis, the response rate was nearly 100%, which avoids the common bias of receiving answers only from those who are interested in the subject.

There are nevertheless some biases in the results. For example, 64% of Japanese households are salaried, and their income is some 5% above the average, while 36% are self-employed and their income is some 10% below the average, but only 2% of the participants in our survey were from the self-employed population. However the proportion of self-employed participants who said that they would like to go to space was not very different from the self-employed, 77% versus 81%. Consequently, while recognising this limitation in the survey, it was not felt necessary to correct the data as a result.

Fourth, it is uncertain how representative our data are of the populations of the advanced industrial countries as a whole. Although similar market research has not been carried out in these countries, the market for space tourism seems sure to be international. Consequently, in order to understand the true potential market we need to consider at least North America, Western Europe and Australasia as well as Japan, the total population of which countries is some 6 times that of Japan.

There may be significant differences in demand between these countries due to the cultural differences between them, which are significant. But there are also important common strands in modern popular culture. Science fiction and particularly space fiction, international organisations such as the Young Astronauts Club, and the world-wide popularity of big-budget space-related films such as "2001" or "Star Wars" suggest that people in many countries enjoy these ideas. Thus, although it is possible that Japanese people are more interested in space travel than other countries, it seems likely that such an interest is also widespread in America and Europe.

However, there are also significant differences between Japanese consumer behaviour and that of Americans and Europeans. One difference is the much higher savings ratio in Japan (which is related to the low interest rates, the high rate of investment, and the low level of unemployment in Japan). Because of their high savings, typically amounting to one year's income held in financial assets, Japanese are able to spend relatively large amounts on things that have particular value for them, such as education, weddings, foreign travel. Another difference is the relatively equal income distribution in Japan compared to America or Europe. From this we might guess that demand for space tourism services may be higher in the latter countries while the price is so high that only the rich can afford it, but may be relatively greater in Japan if the price is low enough to become affordable by a large proportion of the population.

Fifth, since the number of people who said that they would pay one year's income or more to go to space was considerably fewer than those who said that they would pay less, the figures must be considered statistically less reliable. That is, a small number of unrepresentative answers could bias the results significantly. Likewise, the 3 - 4 % of non-student participants in each age-group over the age of 20 who said that they would pay 3 years' or 5 years' salary for a trip to orbit may be unrepresentative, so that the proportion of people with similar opinions may be significantly less than this. Even so, and even if these people are wrong in the sense that in the event of such a service becoming available many of them would not actually pay such a high price, we must still recognise that for a significant number of people the idea of going to space is clearly a powerful dream. However, although there may well be significant demand for space tourism even at very high prices of more than 10 million, this will probably be mainly from uncommonly rich people, rather than from people with average incomes paying a multiple of their annual salary. Consequently, in analysing the results we have included these multi-year answers with those who said that they would pay one year's salary.

Despite these limitations of our market research data, it is nevertheless interesting to use the results as a basis for estimating the future demand for space tourism services. Indeed we have no choice but to do this if we are to develop a successful commercial service. We must also remember a fundamental difference between scientific research and business. Businesses do not require proof before they invest in a project, but reasonable probability. Investment entails risk because the future is uncertain. Businesses that wait for proof risk being beaten by bolder competitors.

Initial results of our survey are described in (1). For the present analysis, in order to derive a demand curve for space tourism, we excluded teenagers, as lacking experience about financial matters, and those over 70 years old, as being unlikely to travel to space. By making the assumption that the participants in the survey are representative of the Japanese population as a whole, we can estimate the proportion in each age group who say they would pay each price, in terms of months of salary. National income statistics allow us to estimate the monetary value of the monthly salaries for different age groups. By using national population data broken down by age we can then estimate the total number of people in Japan who would go to space at different prices.

In order to estimate the number of people who would travel to space each year, we need to develop a traffic growth model, and we plan to do this in the next phase of our analysis. For the present, in order to obtain a single figure representative of a future space tourism business, we simply divide the total demand by 25, that is we assume that 4% of those who wish to would go to space each year. Figure 1 shows the results of this simpler case with an exponential curve fitted.

At first sight the demand shown in Figure 1 seems very high. For example, at prices of \5 million and \2.5 million the demand figures of 100,000 and 500,000 passengers per year are some 10 - 20 times higher than the estimate published by Citron in 1985 of 5,000 - 10,000 passengers per year at a price of $50,000, and 20,000 - 50,000 passengers per year at $25,000 (3). However, in order to compare these figures accurately we need to make some adjustments. First we should adjust the earlier figure for inflation from 1985 to 1993, which reduces the difference to some 5 - 12 times for the lower price. If we further adjust the figures to use a "purchasing power parity" exchange-rate between the yen and dollar, which is arguably more appropriate for comparing consumer prices between Japan and the USA, the difference falls to perhaps 3-7 times.

However, Citron also proposed that demand at a price of $25,000 might grow as high as several million passengers per year (3). Thus, though demand in Figure 1 is greater at higher prices, the demand at lower prices is less, since it shows demand in excess of 1 million passengers per year only at prices below \1.4 million ($14,000). But the results are also incompatible in a number of ways.

The earlier estimates were for world demand, whereas the market research data was from Japan alone. The earlier data relate to just a trip in a launch vehicle, whereas a large proportion of the market research data represents people who wish to stay a few days in orbit. Finally the earlier estimates included projections of growth over the next 30 years, whereas the market research data has no explicit time-scale.

Based on the data in Figure 1, Figure 2 shows the annual revenue that would be earned at each price level, with quartic (solid) and quadratic (dotted) curves fitted. Annual revenues reach a peak of some \1.35 trillion ($13.5 B), though there is some uncertainty about the price at which the maximum revenue would be earned. In the smoothed curves it lies between \1.2 million and \2.4 million. The correct figure will in practice depend on the different services that are offered, as discussed in the following section, but which are not distinguished in Figures 1 and 2.

Since these figures seem very high, it is interesting to consider possible reasons for discounting them. It might be argued that for the various reasons discussed in the previous section these figures should be discounted by a factor of 5 or 10. To discount them even further, to perhaps 1/20 of the market research data, seems excessive: In a country where many people pay three months' salary for a foreign holiday, it does not seem unrealistic to expect that many would pay the same or more for a trip to space, which is clearly a particularly popular dream.

Alternatively it might be argued that once it became a commercially available service, space flight would lose its glamour, and so people would find the reality less attractive than the dream today. While this argument probably has some value, on this qualitative level it is possible also to argue the opposite: that in the coming years, a visit to orbit to see the Earth from space and to experience living in weightlessness could become the defining experience of the new post-cold-war era, in which advanced technology will be used for peaceful more than for military purposes, as foreign air travel might be said to have become in the late 20th century. In addition, the marketing industry, which grows ever more influential with the spread of the mass media, would surely be pleased to be offered the challenge of keeping the idea of space travel exciting. In view of its existing popularity, and of the almost limitless range of interesting future entertainments that can be developed in orbital facilities (4, 5) and beyond, it does not seem a serious danger that space tourism might fail through being considered boring.

Another possible response is to say that the survey is unrealistic because low-cost, airline-type launch operations are not feasible. Although this has been the opinion of many in the space industry over the past quarter-century, the tide of opinion is now changing. Hudson's work on SSTO VTOVL vehicles (6), which had been largely ignored through the 1980s, has been endorsed (7), and even senior figures now openly support the concept (8). To design a vehicle to achieve the cost targets necessary to start a commercial space tourism service is one of the objectives of the current Space Tourism Study Program of the Japanese Rocket Society ( JRS).

After considering these arguments for considering the results to be unrealistically high, we should also note a number of arguments for correcting the figures upwards. These are first, the figures used for average salaries are those for Japanese manufacturing industry in 1990, which are significantly lower than national average incomes in 1993 at the time of the survey. Second, we have not included bonuses, which typically represent some 25 - 40% of annual income in Japan. Third, since we are considering the demand for a service to be available some years in the future, we should allow for economic growth in estimating salaries. If incomes grow at 3.5% per year, this would represent growth of 50% over 12 years, and 100% over 20 years. Fourth, it is a well-recognised feature of consumer expenditure in advanced countries, that once a certain income level is reached, spending on leisure activities increases proportionately faster than on other expenses, taking a larger share of income. Future spending on space travel would fall into this category. If we were to make allowance for all these factors, it is therefore arguable that we should increase the price at each demand level by as much as 100%.

As discussed above, it seems reasonable to assume that world demand for space tourism services might be 6 times greater than demand in Japan alone, although to be sure about this will require further surveys. Consequently, if we took the demand shown in Figures 1 and 2 as provisional estimates of world demand for space tourism, we would be effectively discounting our questionnaire results by some 90%. Thus it seems a reasonably conservative interpretation of our data to assume that world demand for orbital tourism services could reach a level of more than \1.2 trillion ($12 billion) per year at a service price of between \2.4 million ($24,000) and \1.2 million ($12,000). It therefore seems reasonable to conclude that the commercial revenue earned by space tourism services would be several times larger than the entire launch industry today, and has the potential to grow to many times its size. Consequently, in considering the design of a reusable launch vehicle from a commercial point of view, developing a vehicle for the tourist market seems to be an attractive target.

The JRS Space Tourism Study Program assumes that the first phase of space tourism will comprise short trips to orbit lasting a few hours (9). However, it seems likely that once space tourism begins, like other commercial activities the variety of space tourism services will grow progressively and, in particular, orbital accommodation will become available. At first no more than a simple "hostel" comprising a few accommodation units, these orbital vehicles will later become large and sophisticated "hotels", offering a range of entertainments that exploit the unique features of the orbital environment, as discussed in (5). It is therefore interesting to consider the possible pattern of development of such services.

Although Figures 1 and 2 do not make any distinction between demand for different services, in our questionnaire participants were asked to state a preference between a day-trip to orbit, a trip lasting 2-3 days, a week-long stay in orbit, and a stay of 2 weeks or longer, without reference to price. The replies showed a strong preference for stays of a few days or longer, as described in (1). From a commercial point of view the important questions are: "How much more would guests pay for a longer stay in orbit?"; "How much higher will the cost of providing such services be?"; and "How much greater would demand be for these services?"

Participants in the survey were not asked how much they would pay for different lengths of stay in orbit, but were asked to state their preference for a single visit length, independent of price. Consequently our data does not allow us to determine confidently the relative demand for trips of different lengths. Nevertheless, in the absence of more precise data, it seems reasonable to use the increasing popularity of different trip lengths at the same price as a proxy for the growth of demand that would occur as better services were offered.

Following this approach, Figure 3 shows the relative annual demand for the different stay lengths offered, on the same assumption as above, namely that annual demand is 4% of total demand. The revenue figure is not identical to that in Figures 1 and 2 above because we use a single average national income figure, which is not broken down according to age, but it is reasonably comparable.

As orbital accommodation services grew, the share of space tourism revenue accruing to launch vehicle operators would fall, as a larger proportion is paid for accommodation. Thus a significant proportion of the revenues shown in Figure 2 would accrue to operators of orbital accommodation rather than to launch companies. It is notable that the demand for stays of 2-3 days and 1 week show the same pattern of dependence on price, with a revenue peak at a price below \2 million. This is different from the demand for day-trips, which shows a different pattern, with demand being relatively inelastic and growing only relatively slowly as the price falls. However, this is of course affected by the increasing attractiveness of the alternative offerings at the same price. It is an important gap in the present data that, because the demand for longer stays in orbit is, not surprisingly, greater than the demand for day-trips at the same price, we can deduce little from our data about what the demand for day-trips would be at a time when this was the only service available - other than it would be greater than the figures in our results, which represent only those who would prefer a day-trip to a longer stay at the same price.

Thus, as space tourism services develop, as in other forms of tourism the overall price that passengers pay will comprise two separate components, that for flight to orbit and that for accommodation in orbit, and the price of each will fall progressively as demand increases. Stays in orbit will become available at a given price only once flight to orbit has fallen sufficiently below that price. Longer stays will become available later, when the price of shorter stays will be less than that of the longer stays.

If we make a simple assumption that the unit cost of launch services will fall according to a learning-curve of 90%, the cost of a flight to orbit will fall by about 30% as demand grows approximately 10 times from the day-trip phase to 2-3 days in orbit, and by about 37% as demand more than doubles again by the 1-week stay phase. Thus if the price of the service remained constant, as assumed in the questionnaire, the cost of accommodation would be successively 43% and 60% of the cost of flight to orbit. This is comparable to the rough estimate published in (10) that the cost of a few days' stay in orbital accommodation will be approximately 50% of the cost of a flight to orbit. In the future we will analyse this in more detail, breaking down the data collected in our survey according to age and other factors. It will be interesting to compare the results with "bottom-up" economic analysis of orbital accommodation using inputs from the hotel and real estate industries (11).

It is noteworthy that the growth in demand for orbital accommodation, measured by number of guest-days will be faster than the growth in launch demand, due to the growing length of stays. Consequently the service of providing accommodation in orbit has the potential to achieve greater reductions in cost (measured per guest-day) due to learning-curve benefits than have launch services. At a time when 1 million people visit orbit each year, orbital accommodation for 10,000 people or more will be required.

From the point of view of those considering investing in the development of a passenger launch vehicle, a question of central importance is "How high an investment cost can we expect to recover from profits from sales of vehicles?" In order to answer this we should make bottom-up estimates of vehicle operation and maintenance costs, propellant costs, staff costs and indirect costs. By comparing these with estimates of traffic rates at different prices, it would be possible to estimate justifiable vehicle price against passenger fare. By making further assumptions about vehicle production and sales, it would be possible to derive a figure for the maximum development cost that would be commercially recoverable. This is one of the intended outputs of the JRS passenger launch vehicle design study.

For the present paper we take a simpler approach. We assume a certain profit margin on the price that passengers pay. By doing this we can estimate the profit per year obtainable according to the vehicle utilisation. From the overall traffic we can estimate the number of vehicles needed, from which the maximum supportable investment can be calculated by making conventional financial assumptions. However, we must also take into account the fact that the flight price will fall progressively below the overall service price, which we do by assuming a 90% learning curve, as above. For Figure 4 we use the initial assumptions of the JRS passenger launch vehicle study, that each vehicle carries 50 passengers, and flies to orbit 300 times per year (12). In addition we assume a 2% profit margin on the price of the passenger flight (which is less than the overall service price).

On this analysis, the most profitable flight price is about \1,200,000, corresponding to a service price of some \1,700,000. Development of a vehicle designed to fly passengers at such a price could be financed commercially at a development cost of up to some \150 billion ($1.5B), 25% more than a vehicle with passenger flight costs of twice this level. To provide the service would require 50 such vehicles, compared to only half this number of a vehicle with twice the passenger cost. The results of such a simple, preliminary analysis can only be considered tentative. However, the underlying concept of Figure 4 is fundamental to the future of the launch business.

If further research on both the demand for space tourism and on the cost of supplying the services that people want supports our results, it is clear that carrying passengers to orbit and back has the potential to become a mainstay of the space launch business. No other payload which has so far been proposed in the space industry literature offers such a potential. The only other candidate for which launch demand could be of a similar or greater magnitude is the project to deliver solar-generated microwave power from space to Earth. This will become feasible only after a number of technologies have been developed beyond their present stage, in addition to the development of low-cost launch vehicles. By contrast, space tourism can start as soon as an appropriate vehicle is developed.

Another interesting and potentially valuable subject not mentioned in the above discussion, which can also be analysed using the survey data, is market segmentation. As in terrestrial tourism, there seems to be a relatively inelastic demand for a high-priced service, and a separate, more price-elastic demand for low-price, mass-market "pack tours". From our results the possibility of a number of other more detailed categories of customer are apparent - graduation travel, young family holidays, "full moon" travel. These suggest directions for developing a range of distinctive space tourism offerings, which we will investigate further.

Although the interpretation of the data that we have collected is uncertain in a number of ways, at least some of the conclusions that we have drawn seem reasonably robust, and many are suggestive of the possibilities for developing a commercial space tourism business: Large numbers of people appear to be prepared to pay relatively high prices for visits to space. Once orbital accommodation is available, the demand for space tourism seems likely to grow to several times the demand for short flights in a launch vehicle. If the price of passenger flights can be reduced sufficiently to around \1.2 million ($12,000), the number of vehicles required could exceed 50, and the commercially justifiable investment in passenger launch vehicle development could be very substantial.

The simple examples described above are only a selection of the interesting and useful analyses that can be done with suitable market research data. However, they are sufficient to illustrate the potential of this approach, and to suggest areas for closer attention in follow-up surveys. We hope that more detailed and accurate analyses will be done using better market research data collected on this subject in several countries in the near future. In that case the prospects for developing a successful space tourism business will be greatly improved.

In considering such a prospect, the early history of aviation seems to offer many lessons. Basic decisions such as the number of passengers that a vehicle could carry were of fundamental importance to the economics of passenger carrying, and they depended on market estimates. Small differences in the design of different vehicles were sometimes responsible for major competitive advantages. The unending quest for better performance and lower costs to which the aircraft companies were driven by competition led to today's global aviation industry.

As in passenger aviation, energetic competition between a number of commercial companies building and operating reusable launch vehicles seems more likely to produce rapid reductions in cost than the continuation of launch operations by government monopoly organisations. And in order to provide sufficient scale for several companies to compete, a large market is needed. We suggest that space tourism has the potential to provide such a market.

The authors are interested in making their data available to interested parties, either for an appropriate fee, or in exchange for data or cooperation in research on space tourism of comparable value.

1. P Collins, Y Iwasaki, H Kanayama and M Ohnuki, 1994, " Potential demand for passenger travel to orbit", Construction Engineering and Operations in Space IV, ASCE, Vol.1 pp 578-86
2. M Nagatomo (ed), 1993, " Space Tourism Special Issue", Journal of Space Technology and Science (Vol.9, No.1)
3. R Citron, 1985, " Space tourism could drive space development", Proceedings of L5 Space Development Conference, AAS Vol.68, pp 193-205.
4. S Matsumoto et al, 1989, "Feasibility of space tourism: cost study for space tour", Proceedings of 40th IAF Congress paper no IAF-700
5. P Collins, T Fukuoka and T Nishimura, 1994, "Zero-gravity sports centers", Construction Engineering and Operations in Space IV, ASCE, Vol.1 pp 504-13
6. G Hudson, 1991, "History of the Phoenix VTOL SSTO and recent develoments in single-stage launch systems", Proceedings of 4th International Space Conference of Pacific-basin Societies; Advances in the Astronautical Sciences Vol.77 pp 329-51
7. J Penn, C Leonard and C Lindley, 1989, " Review of Pacific American Launch System SSX:Phoenix VTOL Concept", Aerospace Corporation
8. I Bekey, 1994, " Why SSTO rocket launch vehicles are now feasible and practical", NASA Headquarters
9. M Nagatomo, 1993, "On JRS space tourism study program", Journal of Space Technology and Science Vol.9 No.1 pp 3-7
10. P Collins, 1993, "Towards commercial space travel", Journal of Space Technology and Science Vol.9 No.1 pp 8-12
11. C Lauer et al, 1994, " Mixed-use business park developments in space: a real-estate paradigm", Construction Engineering and Operations in Space IV, ASCE, Vol.2 pp 1471-81.
12. P Collins, T Akiyama, I Shiraishi and T Nagase, 1994, "Services expected for the first phase of space tourism", Proceedings of 19th ISTS, ISTS 94-g-25p