__---- ----- ---.-a- r~~r”ru,u‘, 6 WlAlCt: GAO Testimony - . 32iease on National Security Review of Two T-, I _- -5-A -*lery Expected Foreign Acquisitions in the a: 3:jC a.m. EDT Semiconductor Sector ;U'e:! ce s day Z:ne 13, 1990 Statement of Allan I. Mendelowitz Director, International Trade, Energy, and Finance Issues Before the Subcommittee on Commerce, Consumer Protection and Competitiveness House Committee on Energy and Commerce House of Representatives . GAO/T-NSIA3-90-47 GAO Pono 160 (12/W) Mr. Chairman and Members of the Subcommittee: I am pleased to testify today before this Subcommittee. At your request, we examined the review process of the Committee on Foreign Investment in the United States (CFIUS) for two foreign investments.1 Also, as requested, we are providing comments on provisions in proposed legislation relating to GAO access to certain kinds of government data on foreign investment. For the two acquisitions involving the semiconductor industry, we sought to clarify why CFIUS proceeded with the 45-day investigation stage in one case, as provided for in the Exon-Florio Amendment to the Defense Production Act, but, for the later case, ended its consideration after the initial 300day period. The two investments are (1) the acquisition of Monsanto's silicon division by Huels, AG, of West Germany, which CFIUS investigated in early 1989 (but the President decided not to intervene) and (2) the proposed acquisition of Union Carbide Chemicals and Plastics Company (UCcXP) by Komatsu Electronic Metals Company (KEM) of Japan, which CFIUS considered in March and April 1990 but did not investigate. kFIUS agencies are the Departments of Treasury (chair), State, Commerce, and Defense, the U.S. Trade Representative, the Council of Economic Advisors, the Attorney General, and the Director of the Office of Management and Budget. Other agencies participate, as appropriate. 1 These two acquisitions involve U.S. fims that produce the most basic elements in the "food chain" of component suppliers to the semiconductor industry and uitimately to defense electronics systems. UCC&P is unique as the technological leader in producing ultra-high-purity polysilicon, which DOD anticipates it will use in missile guidance systems, infrared sensors, and electrically powered drives for submarines. Monsanto's silicon division was the last major U.S. merchant producer of silicon wafers. Although it was not a supplier directly to the Defense Department, it was to be a key supplier to Sematech, the U.S. business-government consortium supported by Congress to strengthen manufacturing technology in the U.S. semiconductor industry. In the case of Huels' acquisition of Monsanto, we found that CFIUS' decision to investigate was based on the fact that, at the end of the initial 300day consideration period, some CFIUS agencies had questions or concerns that still needed to be fully examined . according to the Exon-Florio Amendment's requirements. This acquisition was the first CFIUS case to proceed to the 450day investigation phase. In the UCC&P-Komatsu case, the Department of Defense (DOD) position that U.S. laws are adequate to assure continued supply was pivotal to the CFIUS decision to end the case at the conclusion of the 30-day period. We note that both cases raise broader questions regarding the preservation of U.S. technological capabilities in sectors critical 2 to national security, and that these questions need to be addressed at a higher policy-making level and in a broader context than the case-by-case approach afforded by CFIUS. EXON-FLORIO AMENDMENT REOUIREMENTS AS you know, the 1988 Exon-Florio Amendment to the Defense Production Act gave the President new authority to investigate and block or suspend foreign investments that threaten to impair national security. Although the amendment did not define "national security," Congress did note in the accompanying conference report that this phrase is to be interpreted broadly and without lim itation to particular industries. The President's authority to block an investment is more narrowly defined. To exercise this authority, the President must find that (1) credible evidence exists that the foreign interest m ight take action that threatens to impair U.S. security and (2) provisions of law, other than the International Emergency Economic Powers Act, do not provide adequate authority to protect the national security. The amendment specifies the maximum time periods for each stage of the review process, allowing 30 days to determine whether to initiate an investigation, 45 days to complete an investigation, and a final 15 days for the President to act. Virtually ill CFIUS 3 cases now are initiated when parties to foreign investment transactions notify the Treasury Department. AS of early June, 1990, 10 proposed foreign investments have been selected as warranting the 45-day investigation stage, out of a total of about 375 investments considered by CFIUS under the Exon- Florio Amendment. Of the 7 investigations completed so far, the President.decided to block one by ordering a Chinese firm to divest its interest in a U.S. aircraft parts manufacturer. In 2 of the 7 completed investigations, the proposed foreign investments were withdrawn. CRITERIA FOR INITIATING THE INVESTIGATION PHASE The amendment did not specify what criteria are to be used in deciding whether to initiate an investigation, nor are these criteria specified in the proposed regulations. These regulations were published by CFIUS in July 1989 but have not yet been made final. CFIUS participants state that each proposed foreign investment is considered individually, on a case-by-case basis, without defined criteria but in light of the amendment's explicitly required Presidential findings. In general, the move to initiate the 450day investigation period is not to be interpreted as prejudging the case; it is normally characterized as a further step in the fact-finding process, which 4 may be needed when important questions remain unanswered at the end of the initial 300day consideration period. The Exon-Florio Amendment does not specify how many CFIUS member agencies are needed to support the initiation of an investigation. However, under CFIUS' present operating rule for initiating investigations, a minimum of three agencies is needed. According to Treasury, when two agencies request an investigation, Treasury or the U.S. Trade Representative Will normally join as the needed third party. Treasury, Commerce, and Defense officials also noted that initiating an investigation can have costs in terms of potential business effects and demands on the President's time, in addition to the direct costs of an investigation. Of course, CFIUS investigations may also have implications fc: :e U.S.' overall relationships with other countries. We note that the UCC&P/Xomatsu case was being discussed the same week that the United States and Japan were involved in intensive talks under the Structural Impediments Initiative. These talks included a,discussion of the U.S commitment to maintaining an open U.S. investment climate. 3g ISSUES RE T The two investment cases we were requested to examine illustrate the manner in which CFIUS considers the three key elements: the 5 link to national security; credible evidence of a possible threat to U.S. security; and the adequacy of other laws to protect national security. We note that the types of issues presented in ,these cases have also arisen in other CFIUS cases. Link to National Security Komatsu's acquisition of UCCCP required evaluating the importance of ultra-high-purity polysilicon to national security. During the initial 300day period, certain questions about this link were raised that did not have clear answers, namely -- what future military and commercial applications are possible for ultra-high-purity polysilicon; -- to what extent VT ultra-high-purity polysilicon technology be central to U.S. firms' ability to participate in future generations of microelectronics, i.e., to what extent will ultra-high-purity polysilicon be a technology dri.vert -- to what extent is UCC&P's technology still a product of a Komatsu license for the original process and what technology is unique to UCC&P; and 6 -- to what extent Komatsu would be gaining access to technology it did not already have? Ultra high-purity polysilicon, the fundamental material for several technologies included on the Defense Department's Critical Technologies List, is an essential material for use in the Strategic Defense Initiative and other weapon systems. Although DOD did not have an immediate need for such polysilicon, it anticipated its use in semiconductors used in high power switchink ' devices and more sensitive infrared detector materials. CFIUS agencies recognized UCC&P as the U.S. firm capable of producing the highest purity polysilicon. UCC&P had completed two contracts for DOD but had no current DOD contracts. One other U.S. firm had bean awarded a DOD contract to produce ultra-high-purity polysilicon, but it failed to meet the specification with regard to phosphorus Although commercial applications for ultra-high-purity polysilicon were unknown, their future importance was not ruled out. CFIUS members knew that UCC&P had originally licensed technology from Komatsu but UCCSIPalso had a patent pending for major changes it had made to that technology, which Komatsu would gain through the acquisition. It was not clear whether CFIUS agencies independently verified how the technology involved in the pending patent related to U.S. security interests; one CFIUS member 7 stated that Komatsu,may already possess in Japan the technology covered by the pending patent. Credible Evidence The requirement of the Exon-Florio Amendment for llcredfble evidence" of a threat by the foreign investor to take action that might impair U.S. security implicitly calls for an examination of the past behavior of the acquiring.firm. In the o'ne case in which the President moved to compel divestiture of a completed acquisition, confidential information regarding past activities of a foreign firm of a Communist country was cited as "credible evidence." In cases involving allied countries, it is inherently more difficult for a CFIUS agency to argue that the foreign firm may threaten national security. In the Union Carbide case, questions arose regarding allegations that Komatsu had engaged in anticompetitive behavior, to the detriment of U.S. firms. In October 1988, UCCCP filed a legal action alleging that Komatsu and six other Japanese firms had formed a buyers' cartel (the High Purity Silicon Issues Study Group) as early as 1983 to manipulate world polysilicon prices to lock non-Japanese competitors out of the market for high-purity polysilicon. While not all working level staff were aware of the allegations in this legal case, officials at the Justice Department, Commerce, and the U.S. Trade Representative were 8 familiar with the UCC&P complaint. Indeed, UCC&P representatives had met with the Under Secretary of Commerce for International Trade in August 1989 to discuss (1) how the Study Group sponsored by Japan's Ministry of International Trade and Industry acted to transform the Japanese industry to the detriment of U.S. commerce and (2) the possibility of raising this in the Structural Impediments Initiative talks. However, officials at the Justice Department, Commerce, and the U.S. Trade Representative told us that they did no analysis of the lawsuit in the context of the Exon-Florio Amendment. Exploration of this allegation as behavior related to the credible evidence criterion was not considered relevant, and no other agencies requested that either the Department of Justice or the U.S. Trade Representative provide cFIUS members with background and analysis of the UCCCP allegations of anticompetitive behavior by Komatsu and the other Japanese polysilicon producers. CFIUS members accepted, without independently verifying, the information provided by Komatsu and ucC&P related to Komatsu's participation in the study group. The terms of the acquisition of UCC&P by Komatsu are to include the termination of the legal case. Questions had also arisen regarding Komatsufs willingness to provide timely supplies to U.S. firms and to DOD. Apparently, at least one U.S. firm had experienced previous problems in obtaining timely supplies from some Japanese firms. 9 Adecuacv of Other Laws . The principal concern expressed in the UCC&P case was assurity of supply to DOD, and so discussion of the adequacy of U.S. laws focused on the Defense Production Act. Under this act, the U.S. government can compel any U.S. -based firm to supply defense contractors before other customers. If a foreign-owned, U.S.-based firm were to withhold or delay supplies to defense contractors, the act can,be used to compel supply. However, other questions were * raised about the ability of the act to protect against a foreign- owned firm's decision to close down a U.S. factory or to change the firm's product line or research direction. AGENCY POSITIONS ON THE UNION CARBIDE CASE On April 3, 1990, CFIUS agencies met at the Assistant Secretary level to discuss whether to initiate an investigation. CFIUS member agencies looked to DOD to take the lead in raising national security concerns about proposed foreign investments. DOD and commerce staff attending this meeting had expected the DOD Assistant Secretary to request an investigation because three DOD units had advocated in discussion papers that an investigation be initiated to answer several questions. The DOD Assistant Secretary told us that he did not have strong views on the need for an investigation of the proposed Komatsu 10 investment and that he stated this at the CFIUS meeting. He had accepted that ultra-high-purity polysilicon was of national security interest, and that continued supply to DOD needed to be assured. In his view, the case presented a choice between two alternatives-- Union Carbide's closing down its operations in high purity polysilicon production or Union Carbide's acquisition by Komatsu. He stated that acquisition of UCC&P by Komatsu would be preferable to having Union Carbide discontinue its polysilicon production. . The Commerce Assistant Secretary told us he raised the issue of assured supply (1) in the case that the Komatsu-owned UCCfP facility remained in the United States and (2) in the case that the UCCCP facility was either closed down or moved offshore. Regarding the first case, the DOD Assistant Secretary told us that the Defense Production Act would be adequate to assure continued and timely supply, since the production facilities are located in the United States. Regarding the second case, he noted that, for sound business reasons, it was unlikely that Komatsu would close the UCCtP facility. The DOD Assistant Secretary further stated that, if Komatsu closed the U.S. facility, the 1983 Mutual Defense Assistance Agreement between the United States and Japan would enable the U.S. government to enlist the aid of the government of Japan to compel Komatsu to supply U.S. military needs. 11 We asked the DOD Assistant Secretary to explain his basis for concluding that the Defense Production Act is adequate to assure supply from the proposed Komatsu-owned UCC&P facility. He cited the authorities provided in the act, noting that the act "provides the President with broad powers including authority to require priority performance of contracts, to allocate materials and facilities, and to enter into arrangements to guarantee sources of supply essential to national defense." However, he also noted that it is the Commerce Department that is responsible for implementing these broad authorities. When DOD did not recommend an investigation, the Commerce Department agreed to end the case at the conclusion of the initial 30-day period. Once DOD had stated that the Defense Production Act was adequate to assure a continued supply, no CFIUS agencies raised questions and this key element of the CFIUS process was not developed further. Neither the importance of ultra-high-purity polysilicon as a technology driver, nor the importance of the allegations about KomatsuVs anticompetitive behavior were discussed at the April 3, 1990, CFIUS meeting. CFIUS members told us that those matters either had been dealt with at the working level or had not been raised. 12 COMPARISONWITH MONSANTOCASE In late 1988 CFIUS began consideration of the West German firm Huels' proposed acquisition of Monsanto's silicon wafer division. At the end of the initial 300day period, CFIUS decided to initiate an investigation in order to examine further and resolve the questions raised in the case. l National Securitv Link In this case, the link to national security was somewhat indirect. Monsanto was not a direct supplier to DOD, but it was to be a key supplier to the newly created business consortium Sematech, which Congress had authorized in 1987 to promote the U.S.' ability to manufacture advanced technology semiconductors. Because other U.S. firms produced silicon wafers for their own internal use (as tgcaptivelg producers), the uniqueness of Monsanto's wafer technology also needed to be established. Credible Evidence CFIUS members found no credible evidence that Huels, as the acquiring firm, might take action threatening U.S. security. In addition, West Germany, as Huels' parent country, is a military ally. The case did raise questions about what could be cited as 13 credible evidence regarding the future intentions and behavior of foreign investors from allied countries. Once a foreign investment case becomes public, it can be awkward for an agency to argue that the foreign fim'S intentions are malevolent, in terms of l!threateningl* national security, particularly if the foreign firm makes a formal statement of its intentions to maintain U;S. operations. Huels, in fact, provided a letter assuring CFIUS that it intended to maintain Monsanto's operations, was committed to the U.S. market, and planned to invest $50 million for research and capital improvements. A senior CFIUS official told us that such letters of assurance have no legal standing and that CFIUS does not have the means to follow up or enforce them. The Treasury Department, which chairs the CFIUS, prefers not to request such assurances because they may be regarded as a type of performance requirement inconsistent with principles advocated by the U.S. government at the General Agreement on Tariffs and Trade. Provisions of Laws Regarding the requirement to prove that laws other than the International Emergency Economic Powers Act do not provide adequate authority to protect the national security, the Monsanto case had no such finding. 14 The Union Carbide and Monsanto cases highlight an apparent inconsistency in U.S. defense technology policy which has arisen in other CFIUS cases. On the one hand, the U.S. government has established a national goal, through Sematech, of developing U.S. capabilities in the semiconductor materials and equipment sector and has supported it with DOD funds. On the other hand, some U.S. companies with the most advanced semiconductor technologies have been acquired by competing foreign firms Without the U.S. government objecting. CFIUS' case-by-case approach does highlight certain foreign investment issues, but it is focused on developing information to meet the Exon-Florio Amendment's specific requirements. Although particular cases may not individually present a threat to national security, the overall decline of U.S. commercial competitiveness in some sectors that are key to defense technology leadership does raise broader concerns about preserving the U.S.-owned production base. An additional limitation of the CFIUS process is its reactive nature. By the time some foreign acquisitions come to government attention, there may be no other way to maintain production in the United States if the U.S. firm wants to discontinue operations and no U.S. buyers show interest. Although foreign investments can 15 bring new capital and technology to U.S.-based production facilities, foreign control means that decisions affecting research, product choice, and plant modernization can be made abroad. The Exon-Florio Amendment Serves as a Useful tool for a narrow range of circumstances. As we have noted in our report on national security concerns about foreign investmentl, however, CFIUS cannot be expected to provide answers to the more complex questions arising out of individual cases. These questions include (1) how much of the defense industrial base has been acquired by foreign- owned firms, (2) which industry sectors, technologies, or types of firms, if any, should be preserved for U.S. ownership, (3) why some U.S. companies have found it desirable to discontinue operations in certain high technology sectors, or (4) how to assess the direction and effects of technology transfers accompanying foreign acquisitions. These questions need to be addressed at a higher policy-making level and in a broader context than the case-by-case approach presently afforded by CFIUS. *Foreisn Investment: Analyzing National Securitv Concerns, GAO/NSIAD-90-94, Mar. 29, 1990. 16 COMMl?NTSON H.R. 4520, THE FOREIGN INVESTMENT POLICY IMPROQMENTS As requested, we are providing comments on H.R. 4520, the Foreign Investment Policy Improvements Act, introduced by Congressman Sharp, to improve the quality of data on foreign direct investment in the United States. My comments on the bill are primarily on those provisions which relate to GAO. We believe that GAO access to foreign direct investment data would be consistent with GAO's mission and appropriate and beneficial to legislative branch oversight in this area. The legislative branch needs to be fully informed about the nature, extent, and effects of foreign direct investment in the United States. Providing GAO access to these data would allow it to perform such evaluations for the Congress and to assess the adequacy of federal government data on foreign investment. We believe, however, that the bill is restrictive in limiting the purposes and times for which GAO shall have access to agency information. Section 8 provides that GAO shall have access to agency information, the disclosure of which is otherwise restricted, Ifto the extent necessary to issue the reports required by sections 3 and 4...." The GAO reports, to the Chairman and Vice Chairman of the Joint Economic Committee, would analyze and recommend changes in annual reports on foreign direct investment 17 issued by the Secretary of Commerce as well as make recommendation9 for improving executive branch policy coordination and data collection and reporting. By limiting GAO's access to this narrow purpose, GAO will continue to be handicapped by a lack of access to detailed Bureau of Economic Analysis (BBA) and Census data to carry out work for other congressional committees that have important responsibilities relative to foreign direct investment. Further, section 4 provides for GAO access only with respect to the first five annual Commerce reports unless later reports are requested by the Joint Economic Committee. Thus, the new GAO access might terminate or become intermittent after five years. GAO believes that the new access authority should be permanent and should not be limited to that needed to issue the annual reports required by the bill. We note that GAO has in place strict and rigorous programs to maintain the security-and confidentiality of information consistent with originating agency requirements. Confidential data is safeguarded by the procedures for gaining access, handling, and enforcing regulations for unauthorized disclosure: and access is restricted on a need-to-know basis. GAO conducts many reviews of agency programs that necessitate access to confidential, proprietary, sensitive, and in many cases, highly classified information. We routinely have access to 18 national security information of the utmost sensitivity. We also routinely have access to proprietary information more similar in nature to the foreign direct investment data, such as that acquired by federal bank authorities in regulating bank activities. GAO has access to Internal Revenue Service tax data, including files that show names and social security numbers. There are occasions when GAO or IRS may prefer that GAO receive sanitized data, with names and social security numbers removed: but this is not a rule. GAO has access to IRS tax data pursuant to two provisions of law: 31 U.S.C. 713, which authorizes audits of the IRS, and Section 6103(f)(4)(A) and (i)(7) of the Internal Revenue Code, which authorizes the IRS to disclose tax returns and tax return information to designated GAO officers and employees. GAO does not have access to some IRS information, such as the formula used to derive the Discriminant Function scores for tax returns, which IRS uses to judge whether a tax return should be audited, and . certain law enforcement information. We also note that the annual Commerce report proposed in H.R. 4520 could be useful to identify investment trends and effects, by industry sectors or subsectors, on a timely basis. Such information may be more useful and timely than the BEA data, which is highly aggregated and sometimes dated due to the time needed to verify and compile the data. 19 This concludes my statement. I would be happy to try to answer any questions you may have. 20
National Security Review of Two Foreign Acquisitions in the Semiconductor Sector
Published by the Government Accountability Office on 1990-06-13.
Below is a raw (and likely hideous) rendition of the original report. (PDF)