The Fractured Power
How to Overcome Tribalism
Last year, America's foreign trade deficit reached alarming new levels. Responding, President Carter announced in September a multifaceted program to encourage U.S. exports, and 15 members of the House of Representatives formed an Export Task Force to pursue the same objective. But also last year the President and Congress imposed various new limitations on exports, and 70 members of the House introduced a "Technology Transfer Ban Act," calling for broad prohibitions on exports to communist countries. In 1979, the Congress will have to reconcile these conflicting tendencies as it legislates an extension to the Export Administration Act, the principal authority for controls on civilian exports, which expires September 30.
Unfortunately, the political context in which Congress will face this task may not be conducive to dispassionate and objective analysis. Alarmists in Congress and the executive branch have seized upon the national uneasiness over Soviet actions in Africa, and the national revulsion over the treatment of Soviet dissidents, to subject the entire concept of East-West trade to the most serious attack it has faced in the last 15 years at least. The President's announcement last December 15 of the normalization of relations with the People's Republic of China has excited passions ranging from euphoria over the trade possibilities, to eagerness to play the China card against the Soviet Union, to charges of selling out our friends on Taiwan. As this article goes to press, events in Iran and Afghanistan threaten to damage the climate further.
In this emotional atmosphere, it will be extremely difficult to maintain a focus on the tough but crucial questions of export control policy as it applies to the communist countries. To what degree is it possible for the United States to influence the growth of Soviet military capabilities by means of export controls? Since any trade has some potential for contributing to Soviet military capabilities (Khrushchev is said to have remarked that the United States should embargo buttons because they are used to hold up Soviet soldiers' pants), where do we draw the line between acceptable and unacceptable risk? How can we fashion an export control system that protects the national interest, without over-burdening the bureaucracy that would implement the controls with either impossible judgments or unmanageable paperwork? How can we impose export controls without unnecessarily frustrating U.S. exports and losing business to other countries, including some of our closest allies?
These questions fall under the traditional heading of "national security controls"-dealing almost wholly with the export of products and advanced technology which might increase the military capabilities of communist countries. But, in addition, the Export Administration Act provides expressly for "foreign policy controls," designed to further the foreign policy of the United States, which may apply to any form of exports to any destination. Today the application of such controls is steadily growing, both through the denial of export licenses for individual items and through broader embargoes. Can these be handled wisely and with reasonable consistency?
And, finally, there are the cases where security and foreign policy interact and become hard to separate. Should we seek to exert economic pressure on the Soviet Union through the selective control of exports, specifically in areas relating to oil production? And how can we keep trade and export policy toward the U.S.S.R. and China in tune with the overall policy of balance defined by President Carter and Secretary of State Vance?
In addressing these questions, this article will stick mainly to the context of the Export Administration Act, proposing key guidelines for its revision. Such an approach necessarily omits detailed discussion of many other laws affecting exports-such as restraints on the Export-Import Bank, conditions attached to authorizations for economic aid and military aid and sales, and the special case of nuclear exports, which fall under the Atomic Energy Act of 1954 and the Nuclear Non-Proliferation Act of 1978. Each of these laws does indeed involve some form of export control based on foreign policy concerns, but the issues are inevitably more diffuse.1
The core of the problem of formal export controls continues to lie in East-West security problems and in the major foreign policy problems raised within the Export Administration Act.
In these days of difficulties in our relations with the Soviet Union, there are those who argue that the United States has no policy for controlling the transfer of strategic technology to the East, and that the policy we should have is one of "don't sell them anything." Both propositions are wrong.
The Export Administration Act makes it the policy of the United States "to restrict the export of goods and technology which would make a significant contribution to the military potential of any other nation or group of nations which would prove detrimental to the national security of the United States." To give effect to that policy, the act authorizes the President to control the export of "any articles, materials, or supplies, including technical data or any other information."
The reason U.S. policy is as stated in the Export Administration Act, and not one of "don't sell them anything," is that we tried the latter policy before, and it failed. For 20 years, under the Export Control Act of 1949, it was this country's policy to deny the Soviets any trade that would contribute to either their military or their economic potential. Since countries presumably trade because it improves their economic potential, this indeed came close to being a prescription for not selling the Soviets anything. This policy was abandoned when Congress passed the Export Administration Act and President Nixon signed it into law in 1969.
There were essentially three reasons for changing from a policy of economic denial to a more narrow strategic embargo. Usually, only one is mentioned today by critics of U.S.-Soviet trade: that we hoped by liberalizing trade to involve the Soviet Union in a "web of interdependence" with the West, and to promote détente. There is room for debate over the success of that strategy.
But there were two further reasons independent of the first. One was that our policy of maintaining a broad trade embargo on the Soviet Union clearly had not significantly retarded Soviet economic growth or inhibited Soviet foreign policy. The other, related to the first, was that our allies had refused to go along with a broad embargo, and were simply selling the Soviets what we refused to sell, thereby taking business away from U.S. companies. Given this experience, it seemed prudent to relax the embargo policy, and to seek the benefits of trade with the Soviet Union to the extent that such trade would not damage the national security. It is interesting that Congress took this initiative even at a time when U.S.-Soviet relations were strained, in the aftermath of the 1968 Soviet invasion of Czechoslovakia.
While our policy is reasonably clear, the effectiveness with which it is being carried out is justly the subject of considerable debate. In determining what reforms are necessary, we must consider the validity of some of the often-repeated criticisms of our export licensing system.
Some observers argue that U.S. controls on exports to the Soviet Union, particularly of advanced technology items such as computers and machine tools, are too loose, so that we are in effect helping the Soviets build up their technological-and consequently their military-capacity to the detriment of our national security. Others, particularly the advanced technology industries, feel the controls are too strict. They argue that we attempt to control too much obsolete technology which is generally available elsewhere and that, when we attempt to control what other technology producers are unwilling to control, we merely lose business to other countries without affecting what the Soviets can acquire.
In the authors' view, the question of whether this country's export controls for national security purposes are too tight or too loose is the wrong question. They can be both. This was the conclusion of a 1976 study by a Defense Department task force, commonly referred to as the Bucy Report after the chairman of the task force, J. Fred Bucy, President of Texas Instruments. The Bucy Report concluded that we control too many end products, but do not control the export of design and manufacturing technology as effectively as we should.
The essential recommendations of the Bucy Report were three. First, export controls should focus on "critical technologies" that "transfer vital design and manufacturing know-how most effectively." According to the report, these critical technologies include: (1) "Arrays of design and manufacturing information that include detailed 'how-to' instructions on design and manufacturing processes"; (2) " 'Keystone' manufacturing, inspection, or automatic test equipment"; and (3) "products accompanied by sophisticated operation, application, or maintenance information." Second, since no technology can be protected forever, it should be recognized that the purpose of the controls is to protect key "strategic U.S. lead times" in critical technology areas. Third, export controls should focus on "active transfer mechanisms"-such as turnkey factories, licenses that involve an extensive teaching effort, joint ventures, technical exchange agreements, and training programs.2
The Bucy Report has led to intensive efforts in the Defense Department to define critical technologies and active transfer mechanisms in operational terms which would be useful in guiding export control policy. The results of these efforts are scheduled to be available early in 1979.
We will not know whether the Bucy Report provides a real basis for a rational system of export controls until we learn whether its key terms can be adequately defined. Still, the report was a useful one in that it emphasized the need to focus our resources on controlling what we have determined to be critical technology. Implementation of the Bucy Report may entail new forms of control on technology transfer. But it should also entail a significant reduction in controls on products. Proposals simply to add more controls to the existing ones would actually damage the national security, in our judgment, further encumbering the licensing bureaucracy with paperwork and distracting it from its essential task.
It is sometimes argued that, if a Soviet capability is supported by technology or products of U.S. origin, U.S. licensing decisions must be to blame-that the licensing system itself is too "leaky." The presumption that U.S. export licensing actions can by themselves control the flow of technology and its products to the Soviet Union is a vanity and a delusion we would do well to dispense with. The fact is that the Soviets usually have access to technology and products comparable to those licensed for export by this country-whether from their own sources, from their East European allies, or from other Free World sources. Where this alternative access can be had quickly, there is no way the United States can stop acquisition of the products by the Soviet Union. In that case, there is no national security reason not to export the same thing. When we do, it is not valid to argue that the export causes the capabilities.
Where, however, the United States possesses a key strategic lead time-i.e., where the Soviets can obtain the technology or products from alternative sources only after a considerable delay-then the government's practice is to deny the license. Witnesses at hearings conducted in 1976 by the Subcommittee on International Economic Policy and Trade, although representing different points of view, were agreed that it is impossible to prevent Soviet acquisition of technology over time, but that it may be possible-and worthwhile-to delay such acquisition. By the time the Soviets do acquire it, the United States may have moved on to a new generation of technology.
Another criticism of U.S. policy and practice with respect to controls is that exports to the U.S.S.R. are often permitted when the items in question could conceivably be put to military use. The whole system of national security controls is indeed-almost by definition-one for controlling the export of products with both civilian and military uses (or, in the jargon, "dual-use items"): if there were no potential military use, the question of control would not arise. But, as a matter of policy and practice, the United States does not license the export to the Soviet Union of products with important military uses without satisfying itself that the risk of diversion from the authorized civilian end use to a military use is small-on the basis of a combined judgment of the probability that diversion would be detected and of its consequences if it should take place.3
A classic example of a dual-use product is the computer. Because the same machine can be used for many purposes, computers confront our government with the most troublesome export-licensing decisions and have given rise to elaborate "safeguards" against diversion, which form part of the sales agreement. For example, safeguards may include a procedure whereby the exporting company programs the computer in such a way that the company can monitor its output and detect any diversion; if diversion occurs, spare parts can be withheld so that the computer quickly becomes inoperable. As another safeguard, permanent residence by an American technician, or periodic on-site inspection, may be required.
It may also be noted that the Soviets have good reason not to divert U.S. exports to military use in violation of the terms of the sale. Few diversions would entail benefits great enough to balance the loss of regular access to U.S. technology that would surely follow detection of any major diversion. The greater the benefits of diversion to the Soviets, the more stringent would be U.S.-imposed safeguards, and hence the higher the risk of detection. Granting export licenses in such cases is not to rely on Soviet good faith, but rather to consider what the Soviets would regard as in their own self-interest.
Ultimately, however, the important point to recognize is that there is no strategically impregnable export-licensing decision; there are only acceptable risks. The question about a computer (or any other dual-use product) is not whether it could be used for military purposes. The answer to that question is always, by definition, yes. The question is, by exporting that computer, do we run an acceptable risk? No amount of information we could possibly acquire would enable us to be absolutely certain that diversion would not take place.
One suspects that those who would place an absolute ban on the export of dual-use items to the Soviets do so in an attempt to escape from this uncomfortably uncertain game. But there is no escape. Other countries stand ready to sell what we do not, and the risk is the same whether the export that produces it comes from the United States, Western Europe or Japan. By refusing to export, we do not eliminate uncertainty; we only opt out of the game.
Because of the various concerns about the national security implications for the United States of technology transfers to the East, proposals occasionally surface in Congress for some kind of congressional veto of export-licensing decisions. Sometimes these take the form of an effort to void in advance a particular sale that members fear might otherwise be consummated. Sponsors of these efforts usually try to make it appear that, were it not for their vigilance, the executive branch would sell the national security down the river. This happened in 1977, for example, when a "Dear Colleague" letter was circulated to all members of the House alleging that the Commerce Department was about to license the sale of an advanced Cyber-76 computer to the Soviet Union. Investigation by the Subcommittee on International Economic Policy and Trade, however, uncovered no inclination on the part of any responsible executive branch official to license the sale. The license was eventually denied on its merits, not because of congressional pressure. The congressional opponents of the sale mistook administrative due process, under which that particular application was granted the same review as any other, for an intention to export the computer.
Proposals also surface for a general congressional veto authority over license approvals with respect to communist countries. (Interestingly, no one has come up with a proposal for congressional reversal of license denials.) Proposals of this kind are misguided. The Commerce Department reviews thousands of such applications every year, granting most, rejecting some. All but the most routine applications are subjected to exhaustive review. Under the law, no export with potential military significance can go forward without the concurrence of the Department of Defense, unless the President himself overrules the Department. As a practical matter, any one of several agencies is able to block an export to a communist country.
It is difficult to see what Congress could add to the review process, especially since Congress lacks both the time and the expertise necessary to review these thousands of highly technical applications. The practical effect (and, one is tempted to say, the purpose) of injecting Congress into the day-to-day licensing process would be to politicize and further encumber a process that already has great difficulty producing technically sound decisions efficiently.
Trying to second-guess specific licensing decisions is the wrong way for Congress to go. If we really want rational decisions, what we need to do is work to increase the rationality of the process that produces those decisions, rather than further burdening an already glacially slow process with congressional override provisions.
If the 96th Congress is to improve the export-licensing process, it must deal with three underlying and related problems: the inefficiency of the process, the presumption of license denial that is built into it, and the existence of an entrenched bureaucracy making export-licensing decisions which is not accountable to those most affected.
The current process is so unwieldy one wonders that it produces any decisions at all. The Commerce Department's Office of Export Administration expects to receive some 70,000 license applications in 1979. That works out to nearly 300 per working day. Backlogs are increasing as licensing officials slowly lose the battle with the paper. Too many commodities are subject to control, right down to microprocessors, which are generally available throughout the Western world and which a Soviet embassy official could purchase for $15 from Radio Shack and carry out of the country in his pocket. Too many agencies must sign off on too many licensing decisions, and there are no effective limits on how much time they can take to reach a determination. Indeed, the system is so structured that there is a built-in incentive for those who oppose exports to the East to use delay as a tactic. So long as an application is in the bottom of a bureaucrat's in-box, nothing can happen. There are no effective procedures for forcing a decision. The result is that too many applications, rather than being dealt with rationally, simply languish in the bureaucracy until the customer cancels the order and gives it to one of our foreign competitors.
The basic factor that accounts for the number of export license applications is the presumption of denial that underlies the system, harking back to the tradition of the broad embargo policy of the cold war years. Indeed, the government's legal premise is that it retains theoretical control over all exports, that all exports are expressly prohibited unless licensed.
As a practical matter, the licensing requirement applied from the start to all technology exports to communist destinations. This is a broad category (as it is construed), and the burden of proof is on potential exporters to show that commodities should be taken off the control list because of obsolescence or foreign availability. Once an item is on the list, it stays there until someone successfully makes a case for taking it off. Meanwhile, new products and technological advances are controlled as they are developed, so the scope of the controls constantly increases. Until the Bucy Report, never since 1949 had the government started a "zero-based" effort, in effect, to throw away the old list and construct a new one based on up-to-date conceptions of strategic significance.
It seems bizarre that, in a supposedly free enterprise society, exports of civilian goods could be governed by a doctrine which asserts that they are prohibited unless approved. This doctrine may be justified in the case of so-called Munitions List items: arms, ammunition, implements of war, and related technical data. But with respect to categories of civilian products, the burden of proof should be reversed. A businessman should be presumed to have the right to export a particular type of civilian product to a particular destination unless the government intervenes and determines that for some overriding reason of public policy that type of export should be controlled. The burden of proof should be on the government. This issue is of more than theoretical significance, because only if we reverse the presumption of denial will we be able to achieve a system lean enough to function effectively and efficiently.
Finally, this cumbersome and biased system is at the mercy of a licensing bureaucracy that has been in place for 30 years and has evidenced great difficulty in changing its approach to technology transfer issues. This bureaucracy has become accustomed to functioning in near-total secrecy without having to account for its decisions. Rarely is a meaningful explanation for a denial or a delay provided to the exporter. Unless he happens to be well connected in Washington, a prospective exporter seldom has an opportunity to sit down with licensing officials, present his case, rebut their arguments, and make sure they understand the equipment in question. Procedural safeguards such as those contained in the Administrative Procedure Act do not apply to the export-licensing process. A great deal of secrecy prevails; it is difficult even for committees of Congress, let alone the exporter, to find out the status of an application or, if it was denied, the reason for the denial.
There may be cases where secrecy is justified, but in general it is part of our democratic faith that the government produces better decisions if it is forced to justify those decisions publicly, and if it is in some sense accountable to those who are principally affected by those decisions. Accountability will lead to more rational licensing decisions, whether those decisions be approvals or denials.
As we have noted, our ability to control technology transfers to the East is limited by the availability of alternative technology from non-U.S. producers in the free world. Both the Export Administration Act of 1969 and the Mutual Defense Assistance Control Act of 1951 (Battle Act) stipulate that this country, in its efforts to control exports, should work in cooperation with other nations, and for 30 years there has been in operation a multilateral Coordinating Committee known by the acronym COCOM. Its members are the NATO countries (minus Iceland) plus Japan. This means, of course, that some technology producers, particularly Switzerland and Sweden, do not belong.
COCOM was set up in 1949, at the instigation of the United States, in an attempt to multilateralize what was then a virtual trade embargo of the Soviet Union and its "satellites." However, there is no "COCOM agreement"-no treaty, executive agreement or other constitutive document. COCOM has no enforcement powers over its members. Principally because of domestic political considerations in cooperating countries other than the United States, participation in COCOM is completely informal and, in some cases, not even officially recognized.
COCOM maintains a list of items which, by consent of the parties, are embargoed for export to communist destinations. Under COCOM procedures, items on the list can only be exported if the exporting country submits an "exception request" which is approved by all the parties. Understandings regarding the details of COCOM's operations and procedures are codified as footnotes to the list. All of this-the list, the footnotes, the exception requests, virtually everything having to do with COCOM-is secret. Secrecy is designed not only to protect sensitive national security information, but also, in our judgment, to protect the participating governments from domestic political opposition. Some European governments are sensitive to criticism from the Left for participating with the United States in anti-communist trade controls, and our own government is sensitive to charges that it quietly acquiesces in loose European and Japanese enforcement of the controls.
The secrecy also applies to the criteria, developed more than 20 years ago, that are employed by COCOM for including items on the embargo list. However, these have been characterized in public hearings as including: (1) whether the items constitute weapons or equipment for their production; (2) whether the items incorporate unique technological know-how of military significance; and (3) whether the items represent materials in deficient supply in relation to military potential in the communist countries.
COCOM has not worked very well, and for many of the same reasons that our own licensing system does not. A single case history, developed in 1977 House subcommittee hearings, is revealing.4
In 1976 the Cyril Bath Company, of Cleveland, a manufacturer of machine tools, received an invitation from Avtopromimport, a Soviet import agency, to bid on ten metal-forming presses. Although the end use was not stated, it was clear to the company from the technical specifications that the machines would be used to manufacture aircraft bodies. Cyril Bath submitted bids on all ten machines, but was awarded a contract to supply only one. The Russians told the president of the company that they were ordering the other nine machines from a French company, ACB-Loire, because they already had ACB presses in operation and were satisfied with their performance. ACB subsequently confirmed this. Cyril Bath applied for an export license, which our government denied on the grounds of strategic significance to the Soviet aeronautics industry. Only after a hearing by the Subcommittee on International Economic Policy and Trade did the government reverse itself on the grounds of foreign availability from France.
The United States then submitted the case to COCOM as an exception request, where it still sits, more than three years after Cyril Bath's original application! The French at first denied that they were manufacturing any such machinery, then admitted it but maintained that the machines were designed for automotive, not aeronautical, use. Indications are that the French machines have already been shipped, without ever having been brought before COCOM. But COCOM has refused to accede to the U.S. exception request so long as the French do not admit that they are supplying comparable machines.
The Cyril Bath case, while hopefully too extreme to be typical, illustrates the major problems with COCOM. Metal forming is not a new technology, and it is one the Soviets obviously already possess, as evidenced by their ability to manufacture advanced aircraft. The French have apparently decided that they will simply ignore COCOM controls on this technology. The United States has failed to make vigorous representations to the French government. Nor has the United States taken the lead to remove this technology from the COCOM list in recognition of the fact that COCOM is not willing to enforce controls on its export. We simply continue to apply the COCOM controls to ourselves, and let our partners sell. The Soviets get the equipment, the French get the sale, and the United States gets left out.
The basic flaw in the COCOM process is that our allies in Western Europe and Japan have never agreed with our conception of it. Because of the greater dependence of their economies on exports, their historic patterns of trade with the East, their tendency to keep trade separate from political considerations, and the failure of fervent anti-communism to take hold in their societies, our allies have always wanted more trade with the East than we have wanted. For 30 years the United States has fought a rearguard action to maintain a broader embargo than our allies have been willing to accept. In the early days of COCOM we had a sanction: the Battle Act required a cutoff of foreign aid to any country refusing to cooperate with U.S. trade controls. So long as Marshall Plan aid was more important to the West European economies than trade with the East, we could secure compliance with a broad embargo. Once this ceased to be the case, we lost much of our leverage.
At our allies' insistence, the embargo was gradually narrowed from an economic to a strategic one-a basic change in criteria finally reflected, as noted above, in our own Export Administration Act of 1969. But our allies remain dissatisfied with the scope of the controls we seek to apply for strategic reasons. They argue that too many technologically obsolete products are on the COCOM list-products that are available in the Soviet Union anyway, and that can therefore safely be sold, to the economic benefit of the West, without damaging our security.
COCOM seems almost designed for evasion. The United States attaches great symbolic importance to the continued existence of COCOM as an expression of Western determination to maintain a common front vis-à-vis the East. Hence, we are prepared to wink at evasions of the controls by our allies because we recognize that too much pressure might cause COCOM to fall apart. Evasion is easy, because each participating country decides for itself whether a given export should be brought before COCOM for approval. And the evidence is that the other parties do not in fact enforce the COCOM controls as strictly as we do, with the result that they get business that we deny ourselves and the Soviets get the equipment anyway. We do not blow the whistle for fear of risking COCOM's collapse.
Reinforcing the mutual suspicions inherent in COCOM is the fact that, because of the inefficiency of our licensing process, our government is unable to deal expeditiously with the exception requests of our COCOM partners. Since they generally do not hold up our requests, they naturally resent it when we do it to them, and suspect that we engage in these delays for our own commercial advantage.
Investigations by the Subcommittee on International Economic Policy and Trade indicate that our COCOM partners do agree with the necessity of maintaining controls on the export of truly critical goods and technology to the East-the same general terms used in the Bucy Report. In this situation, the approach the United States should take seems clear. We should agree to negotiate with our allies a significant reduction in the scope of the COCOM list, so that it embodies only those goods and technologies we can all agree are critical to our security. We should reform our own licensing process so that we can respond quickly to their exception requests, and expect them to continue to do the same with ours. In return, we should insist on more vigorous and effective enforcement of the controls that remain. This might entail reconstituting COCOM on a more formal basis, including the creation of enforcement mechanisms.
It cannot be in our interest to seek to continue the COCOM arrangement as it currently functions. We have to be more accommodating to our allies if we expect to gain their cooperation for an effective multilateral export control system. Above all, we must resist proposals for an expansion of U.S. controls. To the extent that such proposals would require the United States to go it alone on export controls, they would be self-defeating and would have the effect of merely giving the trade to our competitors. More important, to the extent that they would require an attempt by the United States to reverse the trend toward a narrowing of the COCOM embargo, and to bludgeon our allies into accepting greater controls, they would have the effect of destroying whatever consensus does exist among the Western countries on controlling strategic exports to the East.
In short, in COCOM, as at home, we must learn to accommodate ourselves to the reality of a world with many technology producers not subject to our control. We must set priorities and concentrate on implementing our objectives efficiently. Only then can we expect greater cooperation from our allies.
The Export Administration Act of 1969 also authorizes the President to apply export controls "to the extent necessary to further significantly the foreign policy of the United States and to fulfill its international responsibilities." This broad authority in fact codifies a power long exercised by the executive branch-often under express congressional authority.
Controls on exports for other than direct national security reasons have been used for various purposes: to deny economic benefits to countries deemed hostile or threatening to the United States; to give concrete expression to a sense of moral outrage at the behavior of the would-be importing country; to comply with international obligations; to avoid disturbing relationships with other countries (e.g., to prevent an increase in the military capacity of the target country vis-à-vis an ally or friend of the United States); and-with particular recent emphasis-to discourage human rights violations or, at the least, to avoid facilitating them. Usually these purposes overlap, but often one seems to predominate.
Take first the most sweeping examples-the total embargoes applied during the 1920s and early 1930s against the Soviet Union, and from 1950 until 1971 against the People's Republic of China. Trade embargoes are still in effect against Cuba, Vietnam, Cambodia and North Korea, having been imposed by executive order principally under the emergency powers of the Trading with the Enemy Act of 1917, with supplementary authority provided by other legislation. And in 1978 a law was enacted imposing a trade embargo against Idi Amin's Uganda.
Trade embargoes are typically justified at first as a device to deny resources to an offending regime, both to limit its capabilities and with the ultimate hope of bringing the regime down. However, the fact that embargoes invariably remain in effect long after the regime in question is firmly ensconced in power indicates that the real purpose has become-or always was-the basically symbolic one of dissociating the United States from the regime and expressing disapproval of its behavior. Because embargoes serve these symbolic purposes, both internationally and domestically, it is difficult to lift them without seeming to send an unwanted signal. This alone suggests that trade embargoes are too insensitive to changes in another country's behavior to be a very appropriate tool for influencing that behavior.
Such primarily symbolic action certainly has its place. In retrospect it is shameful that the United States did not impose an embargo on trade with Nazi Germany until after Pearl Harbor. But if we are to take such drastic action, it will be well for us to recognize that to do so may cost us something, in human or economic terms or both, and that there may come a time when we will decide that the benefits-psychological or otherwise-are no longer worth the cost. That was the case with the Soviet Union in 1933 and with mainland China in 1971.
The judgment involved is largely subjective and the considerations not easily quantified. Thus no one seriously quarrels with the continuing embargo on trade with Cambodia, and the stopping of trade with Uganda was popular in the Congress. But there are many, these authors included, who believe the embargoes against Cuba and Vietnam are not on balance in the interest of the United States.
Our sense of self-righteousness in imposing embargoes ought to be a little diminished when we stop to think that we tend to abandon the moral position when large quantities of potential trade are involved-i.e., when the target of the embargo is a great power. Conversely, we tend to maintain the moral position when the cost is not high.
The case of multilateral, and especially U.N.-decreed, embargoes raises different questions. That against Rhodesia, initially imposed in 1967 and strengthened in 1968, was in conformity with mandatory sanctions imposed by the U.N. Security Council and accordingly was required under the terms of the U.N. Charter. For many Americans sympathetic to the black African point of view, the action was morally sound, including the purpose of bringing down the illegal Smith regime or causing it to surrender. However, it seems clear that the embargo would not have been imposed had it not been for the Security Council's action. Even as it was, the Congress in the Byrd Amendment insisted on exempting chromium imports from the embargo from 1971 to 1977, and in 1978 tacked another amendment on the foreign military aid bill calling for an end to the embargo if the Smith regime took certain steps. Strong efforts will undoubtedly be made to lift or weaken the embargo early in the 96th Congress.
Multilateral embargoes obviously have a greater potential for being effective than do unilateral embargoes, but can prove extremely difficult and costly to sustain domestically. The Rhodesia embargo was imposed by this country pursuant to the decision of an international organization despite the absence of a strong domestic political consensus in favor of the embargo. Clearly the Smith regime finds considerable comfort in divisions within the United States and exploits those divisions to the hilt. And such divisions would surely be compounded if a multilateral embargo against South Africa were to be attempted.
At a very different level is the relatively small category of export controls designed to protect relationships with countries other than the proposed recipient. One such case, where the denial has been hotly protested by the manufacturer, involved the proposed export of machinery to Taiwan which could be used in the fabrication of missiles. It is quite obvious that an export license is denied in such cases, not because of any impact the exports would have on U.S. security, but because of the delicate balance involved in our overall China policy.
Another license denial, undertaken for a comparable reason, involved a sale to Libya, by the Oshkosh Truck Company of Wisconsin, of heavy trucks which the Egyptians objected could be used to carry tanks. Until the trucks were about to be shipped, the transaction did not require a validated (i.e., individual) license, but could proceed under what is known as a general license, which is a standing authority to ship a certain class of products. At the last minute, the Administration, responding no doubt to strong representations from the State Department, changed the rules so as to require a license for the trucks, and then denied the license. Understandably, the company appealed for help to Wisconsin's representatives on Capitol Hill. While the decision on the heavy trucks stood, a subsequent sale by the same company of different trucks considered less adaptable to military use was approved.
The main problem with this type of foreign policy control tends to be its unpredictability. Export controls become part of a baffling, constantly changing diplomatic signaling process which leaves exporters angry and confused-especially when the exporter is told that, for vague and unexplained foreign policy reasons, he cannot get a license for a product similar or identical to one he had previously shipped to the same customer in the same country. While large enterprises may be able to bear the costs imposed by this uncertainty, it surely has a dampening effect on the propensity of small businesses to export.
Most government decisions on exports reflecting human rights considerations have involved exports financed by the Export-Import Bank or the bilateral or multilateral foreign aid programs, but in some cases such considerations have been paramount in connection with license denials for strictly commercial exports. For example, in 1978 crime control and detection equipment, previously controlled for export mainly to communist countries, was placed under licensing to all destinations except the NATO countries, Japan, Australia and New Zealand. Also in 1978, following the trials of Aleksandr Ginsburg and Anatoly Shcharansky and the harsh sentences imposed by the Soviet authorities, the President denied a license for the export of a Sperry Univac computer to the Soviet Union for use by TASS, the Soviet news agency. Both of these actions were taken in response to, or at least subsequent to, strong pressure from Capitol Hill.
The authors have no quarrel in principle with export controls for human rights purposes. However, it must be noted that there is probably no type of foreign policy control that sends businessmen up the wall faster. Besides the unpredictability of the controls (why are they applied to some human rights violators more than others?), businessmen argue that their unilateral nature merely guarantees that the business will go to other countries and the target country will still get the trade.
Most fundamentally, businessmen argue that trade controls are ineffective in changing other countries' behavior. (The Sperry Univac license denial, for example, was not just a protest against an egregious violation of human rights, but also represented a deliberate effort to discourage future such actions by the Soviet Union.) This brings us to the consideration of foreign policy controls that are intended to achieve a change in the behavior of the country affected.
This type of motivation is not often the sole factor in the denial of export licenses, but it may be one of the purposes underlying a particular set of export controls, especially if Export-Import Bank credits or other government assistance is involved. For example, where the law directs that no credits or economic aid other than to benefit the very poor be extended to governments that persistently engage in gross violations of human rights, the underlying implicit purpose is not so much to punish those governments as to persuade them to change their ways. The same purpose clearly underlay the 1974 Stevenson Amendment to the Export-Import Bank Act limiting credits to the Soviet Union to a total of $300 million, and the Jackson-Vanik provision of the 1974 Trade Act, which prohibited export credits and most-favored-nation tariff treatment for non-market-economy countries imposing unreasonable restrictions on the right of their citizens to emigrate.
It is not within the scope of this article to argue the effectiveness of the Jackson-Vanik and Stevenson Amendments. It is, however, pertinent to point out that, when we embark on foreign policy controls of this kind, we should try to answer the $64 question: what are the chances of success? The same question should, of course, be asked when the controls in question are broader than export limitations, such as total embargoes on trade, or are of a slightly different type, such as controls on investment. The Congress in 1979 will again be wrestling with proposals of this character: for example, bills to limit trade with countries that harbor international terrorists and to limit investment in South Africa.
With respect to these various categories of foreign policy controls, are there general principles which should apply and which the Congress should attempt to spell out in the coming legislation extending the Export Administration Act?
The essential point to recognize at the outset is that when we employ export controls for foreign policy purposes, we do so alone. No other COCOM country is willing to go along. Thus, to apply foreign policy controls is to apply unilateral controls. We do not want to argue that the United States should never act unilaterally. But if we do, surely it should be on the basis of the most careful consideration, with full regard for the costs.
We can start with the proposition stated at the beginning of this article: that the United States has a dangerously large trade deficit, and that the encouragement of U.S. exports must be a major national objective. Obvious as this proposition is, it often seems to be forgotten when emotionally based arguments are advanced for limiting, or continuing to limit, a particular kind of trade. The time is past, if it ever existed, when we could think of trade controls as a cost-free way of expressing disapproval of another country.
If exports are in the national interest, as indeed they are, then they should be controlled and limited only for clear and compelling reasons. As we have already suggested in connection with national security controls, the burden of proof should be on the proponents of controls. The presumption should always be that a company has the right to export, and that the U.S. government supports its efforts to export and will intervene against that right only for carefully designed and important reasons of public policy.
This presumption must be recognized particularly when the President or the Congress seeks to impose export controls of a new type or against an importing country not hitherto affected. Where such action is taken by law, then both the Congress and the President are involved and share the responsibility. But when the action proposed to be taken is within the authority of the executive branch, then it seems reasonable to give the Congress the right to veto that action by Concurrent Resolution. This is not at all the same as providing for a congressional veto of all licenses issued, as discussed above. The latter would affect, and interfere with, the efficient execution of a previously defined policy. When a new type of controls is involved, then the Congress should not be excluded from the decision.
Currently, a major reason that some people in the Administration consider export controls an attractive foreign policy tool is precisely that these controls are available to be used at will, without recourse to Congress. If a congressional veto provision in the law achieves nothing else, it will serve to encourage the executive to proceed cautiously with new export controls and to consult with the appropriate congressional committees before deciding to impose them.
A second general principle, which ought to be accepted without dissent, is that the government should be clear, both in the institution and in the exercise of foreign policy controls, on just what it is seeking to accomplish. If the objective is known, then the appropriate questions are more likely to be asked: What are the chances that the controls will achieve the desired result? At what cost? If they fail to do so, can they be lifted without embarrassment? Can the target country retaliate by denying us something we want? And it would be salutary for the government to be required to answer the following question about every control it proposes to apply: How will we know when the control has succeeded?
Related to the second is a third general principle: so far as possible, the reasons for a particular set of controls or for denying a license under those controls should be made known. No one quarrels with the proposition that the processes of government should in general be open rather than secret. In this case, as in other fields of government regulation, those adversely affected by the controls will be more likely to accept them if they can understand the reason for them. Moreover, in some situations (such as an embargo imposed to express moral outrage), the purpose of the controls will be furthered by publication of the reasons for them.
Inevitably, however, there will be other cases where the reason for certain controls or for the denial of licenses cannot be made explicit, even though the reason may be self-evident. Such a case may arise, for example, where the purpose is to bring about a change in the target country's behavior. In such situations, the target country's pride must be taken into account: the policy change will be easier to achieve if it can be made without loss of face.
Finally, if exports are to be encouraged, the system of controls, both for national security and for foreign policy reasons, should be made as predictable as possible. We recognize that unpredictability can be a diplomatic virtue, but that just points up the difficulty of using trade for diplomatic purposes. It is questionable how much is gained by sacrificing the development of stable trade relationships to the vicissitudes of diplomacy. The government should seek to avoid changing the rules in the middle of the game, thereby creating uncertainty in the minds of our exporters as to whether it is worth pursuing foreign markets.
Sometimes surprise is unavoidable; for example, in the Sperry-Univac computer case the sentences of Ginsburg and Shcharansky came after the company had negotiated the sale. But in other situations the U.S. government has simply changed its mind in midstream: this was true of the Oshkosh truck sale to Libya, and of the ban on exports to most countries of crime control and detection equipment, which went into effect before the regulations were even published and had the effect of voiding existing contracts. In both of these cases, the action taken was probably justified, but the timing and the procedure were unfortunate.
There will no doubt always be some unpredictability in the administration of foreign policy controls. And exporters have a responsibility to be more sensitive to situations where it is obvious controls may be applied and not to jump in without careful thought. But the Congress and the executive branch should try in the law and the regulations issued to create as much predictability as possible.
Many of the questions we have raised about foreign policy controls find practical expression in a major issue confronting the U.S. government: Is it in our interest to seek to slow down the development of Soviet oil production capabilities? The proposal to use export controls on oil production equipment as levers to influence Soviet foreign and domestic policy is principally associated with Samuel P. Huntington, who, while on the staff of the National Security Council, argued publicly for "a new approach of conditioned flexibility in which changes in the scope and character of U.S.-Soviet economic relations are linked to and conditioned by progress in the achievement of U.S. political and security objectives."5
While the general proposition has a certain plausibility, Huntington fails to treat adequately certain problems which arise when that proposition is applied to specific issues such as that of oil production technology. In this section we will try to explore some of the questions his position raises.
Last year, following the Ginsburg-Shcharansky sentences, the President considered reversing previous approvals of licenses issued in connection with the export by Dresser Industries of a drill bit manufacturing plant to the Soviet Union, but then decided not to do so. He also approved a license for one component of that sale not previously licensed, an electron beam welder. The President did, however, order an expansion of the types of oil production equipment exported to the Soviet Union subject to license, and called for a review by the National Security Council of all applications to export such equipment to the U.S.S.R.
The buzzwords "national security" are often used by those who criticized the President's decision to approve the Dresser sale, and who oppose any similar actions in the future. But it has not been convincingly argued that the exports in question would have had a telling effect on Soviet military capabilities. Indeed, no agency involved in the decision, including the Defense Department, made such a finding. This presumably reflects the Administration's judgment that the technology has no significant military application or is readily available elsewhere-and that the Soviet military establishment in any case faces no shortage of oil. It is clearly not limited by a shortage at the present time, and even in the long run it seems unlikely that the Soviet military-industrial complex, which presumably has first call on oil resources, would be the sector to feel the pinch if the U.S.S.R. suffered an energy shortfall.
What opponents of the Dresser sale really seem to be saying is that the exports should have been barred because of Soviet actions in Africa and the harsh treatment of dissidents, both as a protest and in the hope of influencing that behavior. The moral or practical effectiveness of such action as a protest is difficult to analyze. But if the main purpose is to secure a change in Soviet behavior, then various questions must be answered. For starters: To what extent can the United States effectively slow down the development of additional Soviet oil production capacity? What capabilities do the Soviets already possess? What technologies are available from competitors? What are the chances that we can secure the cooperation of those competitors in controlling these technologies?
Objective information with regard to the first three questions is limited. Opponents of trade in this equipment naturally downgrade Soviet capabilities and foreign availability, while proponents take the opposite view. The answer to the last question, however, seems clear: while there may be some conceivable circumstances in which our allies would join us in controlling trade with the Soviets for political purposes, we should be under no illusion that they will support attempts to link trade in oil production equipment to Soviet dissident trials or Africa policy. They will not. On this issue, the United States stands alone.
One commonly forgotten factor in this debate is thus that, even if U.S. technology is the best in some respects, the Soviets do not face a U.S.-or-nothing situation. Second-best technology is available, and the cost we can inflict on the Soviets for their behavior must be measured as the difference between that technology and what the United States has, not as the difference between what the Soviets already have and what the United States has.
If we determine that we have a lever in our oil production technology, another basic question arises: How badly do the Soviets want this technology, and what political price might they be prepared to pay for it? This question seems particularly ignored in discussions of this issue. There may well be concessions that we can obtain from the Soviets in exchange for this technology, but there are undoubtedly some that we cannot. Huntington's statement that "American oil and gas equipment would continue to flow into the Soviet Union" if "the Soviet Union would appropriately moderate some of its undesirable behavior" is not helpful.6 Proponents of controls to influence Soviet behavior fail to appreciate that they must define more precisely what changes they are seeking, and then calculate whether the Soviets could logically be expected to make them in exchange for whatever we have determined we can withhold from them. Based on the Soviet Union's long history of willingness to force its people to undergo tremendous sacrifices for the sake of preserving its internal system and foreign policy, the authors are skeptical that the Soviets would be willing to make fundamental changes merely in exchange for access to U.S. oil production technology.
A third set of questions has to do with the effects of a decision to impose export controls on oil production equipment on the likelihood that there will continue to be exports to control and consequently any leverage to exercise. It seems highly unlikely that American firms would continue to pursue the Soviet market if U.S. license applications were subject to denial on the basis of random occurrences in Soviet foreign or domestic policy, and even less likely that the Soviets would permit themselves to become dependent on a supply of U.S. technology which might be turned off in response to political actions. Yet the argument that we should deny licenses to export oil production equipment for the sake of leverage is based precisely on the assumption that both parties to the transaction will continue to pursue the business no matter what the U.S. government does. If they do not-if both the U.S. sellers and the Soviet buyers take their business elsewhere-there is no more leverage. That is the primary fallacy behind arguments for politically determined export denials: By seeking to gain leverage, you can lose it. If we want leverage, it would seem sensible to seek ways to achieve it that are not self-defeating.
Finally, a decision by the United States aimed at slowing up Soviet energy development would be perceived by the Soviets as a departure from our practice of the past decade of seeking to restrict the export only of products and technologies of more or less direct military significance, and of not seeking to weaken or retard the overall growth of the Soviet economy. The Soviets are certain to react to such a policy change. And if they subsequently see the United States failing to "appropriately moderate some of its undesirable behavior," what price will they decide to try to make us pay? And what card would we play in response to that? At what point would we move on to even more drastic forms of export controls-e.g., over grain sales-with all their attendant economic and political costs? It is important to look down the road and try to see where all this might lead. That does not necessarily mean that we should reject export controls used for leverage, only that we should be aware of the possible consequences and make sure that we are willing to pay the price. Our desire to show the Russians who is boss is understandable, but hardly a sufficient basis for making what could be dangerous changes in our foreign policy.
Overall, the arguments against using oil equipment controls for the political purposes which Huntington proposes seem to us persuasive. Surely the burden of proof rests on those who would employ such controls to demonstrate better than they have that the controls would be effective.
In conclusion, a brief word on our trade relations with mainland China may be in order, especially in view of President Carter's welcome decision to recognize realities and extend full diplomatic recognition to the Peking regime as the government of "China." This action raises many trade issues, most notably whether we should have granted China most-favored-nation treatment either within the terms of the Jackson-Vanik Amendment or some other basis. While that issue falls outside the scope of this article, several export control matters must also be confronted as we work out a new relationship with the People's Republic.
Unfortunately, because of the high degree of secrecy pertaining to export control matters both within the U.S. government and in COCOM, very little of our export control policy toward China is a matter of public record. From 1950 to 1971, as noted above, the United States maintained a trade embargo against mainland China. In the early 1950s, at U.S. insistence, the Western countries applied harsher controls to China than to the Soviet Union, giving rise to a so-called "China differential." A separate organization, CHINCOM, and a longer control list were maintained for China. After the Korean War the China differential became harder for the United States to sell, and it collapsed in 1957 when the British unilaterally abandoned it and the other COCOM members quickly went along. CHINCOM and the separate China list were abandoned.
However, the new policy of what today might be called "even-handedness" soon created new pressures. The doctrine that COCOM members should not export to any communist country any product that would increase the military capabilities of the weakest such country meant that all were treated at the level of the Chinese. Inevitably, as Soviet technological and military capabilities increased, COCOM members argued that they should be able to sell to the U.S.S.R. goods which the Soviets had already demonstrated a capability to produce, even though the same products could not be sold to the Chinese because they would add to China's more limited capabilities. Given what we know about how COCOM operates, one assumes that some informal accommodation was made to this point of view, but secrecy prevails.
However, the argument that because China was so weak anything would contribute to its military capabilities, was turned around in 1975 when the British proposed to license the sale by Rolls Royce, Ltd., of supersonic military aircraft engines and production technology to the People's Republic. While the British arguments on behalf of this sale are not publicly known, they must have argued that, since the engines would contribute to Chinese capabilities vis-à-vis the Soviet Union, this would not affect Western security. The United States objected to this sale, but the British went ahead anyway.
By the end of 1978, it was clear that, following the British precedent, a "reverse China differential" was being applied. Although the United States was not itself selling arms to China, it was acquiescing in the sale of such arms by West European countries. What role COCOM has been playing in these transactions, if any, is not clear. (Since U.S. controls on oil production equipment for foreign policy purposes are being applied only to the Soviet Union, the differential between our treatment of that country and China for export control purposes is increased.) Earlier in the year there had been open talk by the President's National Security Adviser of "playing the China card" against the Soviet Union, in part by loosening up exports to China while tightening up on those to the Soviets, although one heard less of that by the end of the year. By the time that normalization was announced, the Administration was at pains to emphasize that its policy toward the Soviet Union and the People's Republic was one of even-handedness.
Clearly, many things need to be sorted out in our export control policy toward China. From the point of view of our security, it would seem that we could sell many military and dual-use items to China which would have the effect of building up Chinese military capabilities without necessarily affecting U.S. security directly. From the point of view of our foreign policy, however, it is not clear that we would want to do so.
Our own feeling is that an expansion of trade with China, including trade in advanced technology, is to be encouraged, creating as it would possibilities for significant improvement in the relations between our two countries. However, a policy of using export controls for the explicit purpose of building up China at the expense of the Soviet Union would be dangerous. Those who advocate this approach have apparently not thought deeply enough about whether it really would serve any good purpose for the United States to heighten Soviet fears of the Chinese or increase Soviet apprehensions concerning U.S.-China normalization. Instead, we should be able to communicate to both the U.S.S.R. and China that we are not attempting to increase the threat each poses to the other. Above all, we must remember that the Soviet Union remains the world's only other superpower-the only country in the world capable of destroying us. Maintaining good relations with the Soviet Union must be our paramount objective. While that will not and should not deter us from normalizing with the Chinese, it would be foolish in doing so deliberately to insult the Soviets.
This discussion of export controls for foreign policy purposes has posed more questions than it has resolved. But bringing some of the problems up for close examination has been one of our main intentions. The U.S. government itself is only tentatively feeling its way with this relatively novel use of export controls.
We believe that the American public as well as our policymakers must keep in mind certain principles: that trade is in itself good for us; that there are limits to our influence; that export controls, like all aspects of foreign policy, must be as open and as accountable as possible; that we have to be clear about our objectives and, in trying to shape a policy, must avoid simple answers; and that our trade policy should be an expression of what is good and not what is vindictive in us. With those guidelines, the Congress, the executive branch, and the American people should be able to evolve an export policy that supports and contributes to a constructive American foreign policy.
2 An Analysis of Export Control of U.S. Technology-A DOD Perspective. A Report of the Defense Science Board Task Force on Export of U.S. Technology, Washington, D.C.: Office of the Director of Defense Research and Engineering, Department of Defense, February 4, 1976.
3 The Department of Defense maintains a list of criteria employed by the Department to make this judgment. They are: (1) Is the item appropriate (in quantity, quality, demonstrable need, design, etc.) to the stated civil end use? (2) Is there any evidence that the stated end-user is engaged in military or military support activities to which this item could be applied? (3) How difficult would it be to divert this item to military purposes? (4) Could such diversion be carried out without detection? (5) Is there evidence of a serious deficiency in the military sector which this item, if diverted, would fill? (6) Is technology of military significance, which is not already available, extractable from this item?
4 For a complete account of this case, see Export Licensing: Foreign Availability of Stretch Forming Presses, Hearing before the Subcommittee on International Economic Policy and Trade of the House Committee on International Relations, 95th Cong., 1st Sess., November 4, 1977, Washington: GPO, 1977.
5 Samuel P. Huntington, "Trade, Technology, and Leverage: Economic Diplomacy," Foreign Policy, Fall 1978, p. 67.
6 Ibid., p. 77.