scientific productivity as these programs shift to address the scientific basis of technologies—are only now beginning to emerge.

In general, the Panel concludes that there is much room for improvement in intelligently targeting the government’s efforts to prevent unwanted technology transfer; priorities must be set and communicated in order to limit the adverse effects of controls on other vital national interests, including that of maintaining a position of world leadership in science.

More specifically, there are several areas in which improvement is needed:

• making controls more workable,

• improving the factual basis for decisions,

• improving mutual understanding between the government and the scientific community, and

• bringing better balance to U.S.-U.S.S.R. exchange programs.

At this stage, the government’s technology transfer controls have a very wide compass. There is a risk that by covering such broad expanses of technology and by implicating the scientific activities across the same range, the overall efficiency of the effort will suffer. Two principles can be applied to bring a more coherent focus to the problem. First, the government should concentrate on the most feasible forms of control and should eschew regulations that impose compliance burdens without significantly affecting leakage. Second, it should concentrate its resources more systematically on those technologies that are of greatest relevance to near-term Soviet military strength.

American scientists have broad, constitutionally based rights to disseminate information within the United States free from government control, unless the information is classified or they have agreed in advance to contractual provisions limiting disclosure. And, as a practical matter, information that is available domestically is also available abroad. For example, there is no practical way to prevent domestic publications from circulating internationally. Both ITAR and EAR recognize this fact to a limited extent by providing exemptions from the formal licensing process for certain types of generally available information, such as published data.¹ But information is available through many channels—lectures, seminars, conferences,