Sir

David Leverington claims to be measuring cost-effectiveness of astronomical observations, and your provocative headline, “Star-gazing funds should come down to Earth,” shows that his conclusion can lead to important consequences (Nature 387, 12; 1997). But, as his Correspondence shows, he is really merely measuring the number of times various papers have been cited.

He has no way of accounting for the uniqueness of an observation or considering the changes in our understanding that can result from breakthrough discoveries. His reliance on citation is like poll-driven government: it has some value but isn't necessarily the best.

In particular, he says that, “although the Hubble Space Telescope has been extremely successful, Figure 1 shows that it has not yet justified its high costs…”. I submit that this conclusion is ridiculous and that, after seeing the wealth of unique and revealing observations that have been and are still pouring from Hubble, an observational scientist is driven to finding out what is wrong with Leverington's arguments rather than accepting his conclusion. After all, if we assign a factor of 3 for uniqueness, then the space observations would meet even his absurd criteria.