| Geoscience and Environment |
Suppression of demand for ASTER products is defensible. The same cannot be said for the products of Landsat, a 30-year old program, nor for RADARSAT, nor for some European and Asian data products. Apparently economists in government service and in universities are not involved in providing advice to guide decisions concerning pricing of satellite data products. The probable consequence is that pricing for such data is less technically sophisticated than pricing for toothpaste.
What might a more sophisticated approach entail? First, what is the objective of the pricing policy? What is the price supposed to do? Should the pricing achieve revenue maximization? Recovery of operating costs? Some combination of these two or other objectives? Or should price send a signal via the market to tell data providers what kind of data is wanted and what level of processing is desired? Does a price of zero achieve anything worthwhile? Second, how is the price determined? Ostensibly, some cost figure is divided by some number of datasets to get the "cost per dataset" that forms the basis for present prices. What costs go into the numerator and how are the units in the denominator counted? After allowing for data distributed free, historical reductions in prices of Landat and other datasets could be used to calculate the price elasticity of demand. Does demand increase as the price is reduced? If demand increases, the unit cost must fall. If total revenue increased with a fall in price, this would justify further reduction in price. Third, if production of satellite data must be subsidized, what mechanisms distort the market the least? An example of an alternative mechanism would be reduced-price vouchers for certain consumers. What justification is there for free data, if any?
The issues can be illustrated by analogy with a capital intensive project like a bridge. After construction, the numbers used to justify the project become irrelevant. Net-revenue maximization becomes the major objective for pricing. What counts is cash flow. The reason is this: the cost of investment for constructiing the bridge is very large in relation to the cost of operating the bridge. Moreover, once the bridge has been built the construction cost is a sunk cost: the owner cannot get his money back except by taking in the cash flow or selling the right to the cash flow. For a bridge, cash flow equals the amount of toll fees less the cost of operations. And the cost of operations is what the owner would save by closing the bridge. The operator of the bridge needs to discover what price structure for traffic will maximize the net-revenue (revenue less operating costs). Clearly, a price of zero will maximize traffic, but not net-revenue. At some high price no traffic at all will use the bridge. Between these extremes there is a revenue maximizing price, structured for various classes of users, say cars and trucks of various sizes. Because operating costs are not closely related to the volume of traffic, the net-revenue maximizing price will be not much higher than the revenue maximizing price. If demand is elastic with respect to price, bridge operators will want to keep prices low; if demand is not elastic with respect to price, they will want to raise prices. The traffic determines the price. Otherwise, how would anyone ever know when there are enough bridges or what kind of bridges are needed or where they should be located? To stimulate demand, bridge owners might sell reduced-price vouchers to specific classes of users, such as pensioners who might otherwise avoid using the bridge. If required to provide free use of the bridge to police, military, and other government vehicles, the owners would nevertheless maintain proper bookkeeping to sustain claims for public funding for additional operating costs. This hypothetical bridge owner must contend with many factors that satellite and remote-sensing device owners must also consider. One thing the bridge owner knows for sure is that suppressing demand by high prices is not the way to operate a bridge.
The thrust of this argument is not necessarily that satellite data should be sold at the revenue-maximizing price, though that might prove to be optimal depending on the objectives. The point is merely that the kind of thinking required to work out a sound policy for pricing satellite data is different from the kind of thinking needed to run a government agency, to design a remote-sensing package, to rocket a instrument into space, or to use the data once it is downloaded. The approach needed is different even from that of the commercial data-vendor who is compelled to focus on his own profit or switch to another business.
The issue is this: after the taxpayer has paid for the rockets and the satellite, and while he is funding the data processing and distribution, how can she be certain the data actually flows to users in optimal quantities and that the data provided is the data that users actually want? Note that this hypothetical taxpayer does not ask why the users want the data any more than she asks why travellers want to cross the bridge. Our hypothetical taxpayer wants reassurance that satellite remote-sensing is in some way driven by wants and needs rather than by technology.
Satellite data is accumulating at an ever increasing rate in all countries supporting remote-sensing programs. If the hypothetical taxpayer were also a retailer she would ask, how much of this product is being moved? How does it pay its storage costs? While accepting that many questions of satellite data supply and demand cannot be resolved by considerations of markets and pricing mechanisms, we ought also to consider if remote-sensing is now overly driven by technology, and if so, will such an approach be viable over the long-term. At the very least, we should be certain that pricing mechanisms do not suppress demand, as this author believes. Using price deliberately or unintentionally to suppress demand for satellite data products would place the future of the whole satellite remote-sensing enterprise in jeopardy.