There is a tide in the affairs of men /Which, taken at the flood, leads
on to fortune.” – Julius Caesar
Concluding the NAFTA negotiations in December 1992 was like a Christmas present for the high-tech community. In IBM’s Washington office, we were pleased with the reaction of our customs compliance organization – the IPODC in Boulder. One person in our office saw in the NAFTA agreement a key opportunity, especially the Common External Tariff (CXT) provisions. That person, Doug Worth, was one of our senior executives (indeed, he would soon become the head of IBM’s Governmental Programs office).
Even before Congress passed the NAFTA implementation legislation, Doug indicated that he wanted to capitalize on the liberalization momentum established in the new trade policy sphere. He walked into my office and asked, “Why can’t we expand the NAFTA CXT concept and get other countries to agree to it?”
He asked me to develop ideas about how to make it happen, but to be honest, I had serious doubts about the idea. As I considered it after he left my office, I started thinking about what key questions would have to be addressed –
What new countries should be considered?
What product coverage should be considered?
What negotiation medium could be used in which to push the idea?
How would we position our new idea with government and industry leaders?
Would the CXT approach of NAFTA work?
Who would our champion be in the U.S. government?
Who would our allies be in the private sector?
And of course, what barriers might block it from happening?
Initial Thoughts on Countries of Interest
In essence, most of the world’s largest marketplaces for high-tech products were addressed in one way or another by the three major agreements already – the U.S. Japan Semiconductor Agreement, CUSFTA, and NAFTA. But we knew that there were more target potentials to consider – most notably the member countries of the Organization for Economic Cooperation and Development, or OECD. It might be a surprise in today’s economy, but in the early 1990s, the global market for high-tech hardware was dominated by a few industrialized nations.
Roughly during the period 1990-1994, OECD countries accounted for about 85% to 90% of total global trade volume (including both imports and exports) in computers, peripherals, semiconductors, and computer parts (Source: OECD). The United States and Japan were by far the largest markets in this trade with the European Union (then called the European Community) appearing to trail closely with 25% to 30% of global volume, but that was deceptive, since the lion’s share of that trade was intra-European. Extra-European trade represented instead only about 12% to 15% of global trade share, still a sizable chunk.
As we looked at these numbers, we realized that the OECD was effectively a closed-loop for technology trade. This wasn’t surprising. The PC revolution had taken root in corporate America, Japan, and Western Europe. Similar to what we saw regarding intra-European trade volumes, the same could be said of intra-OECD volumes. Computer parts made in any of those countries were frequently shipped to other OECD countries for assembly into more integrated or final products. This was particularly true for companies like IBM. Compaq, and Apple. Developing and non-OECD countries had not experienced this level of technology absorption into their consumer economies yet and large-scale server infrastructures were yet to justify the import of finished computer products.
Admittedly, I had no idea about how to answer many of the questions mentioned above, but a truly aggressive idea – which I believed Doug would consider was too aggressive – was starting to take shape in my mind. For the lack of a better term, we called it the International Computer Agreement, or ICA. The idea was to capitalize on the trend toward liberalizing trade on high-tech that we had seen in recent trade agreements. Eliminate tariffs on computer products within the OECD countries.
Before going further, it’s important to recall a fundamental principle in international trade called Most-Favored-Nation (MFN) treatment. Simply stated, MFN guarantees equal and non-discriminatory trade terms between nations. This means that if one country grants a special trade advantage to another, it must automatically and unconditionally extend the same treatment to all other trading partners that hold MFN status.
There are just a couple of notable exceptions to the MFN principle: (1) countries may form regional trade blocs or customs unions such as the European Union where members can offer even lower tariffs to each other without being required to give those rates to the result of the world; and (2) wealthier countries may grant preferential market access or lower tariff to developing countries to help enable their growth.
With MFN treatment in mind, the aggressive idea presented to Doug was that his desire for further liberalization beyond the recent trade agreements might be met by extending freer trade concepts to ALL OECD COUNTRIES via a new agreement called the ICA.
Somewhat simplified, this was the overall approach:
Within OECD - 0% under Proposed ICA
Outside OECD - Existing MFN national rate
Computers - Within OECD: 0%; Outside OECD - Existing MFN national rate
Computer Parts - Within OECD: 0%; Outside OECD - Existing MFN national rate
Peripherals - Within OECD: 0%; Outside OECD - Existing MFN national rate
Semiconductors - Within OECD: 0%; Outside OECD - Existing MFN national rate
And, this is what the ICA would change –
United States - 3.9% Computer tariff goes to 0% for OECD imports
Canada - Already largely at 0%, so little additional change
Japan - Computers already effectively at 0%, some semiconductor tariffs remained sees some additional sectoral liberalization
Mexico - Significant computer tariffs see major reduction for OECD suppliers
European Community - Some significant electronics/semiconductor tariffs see major reductions for OECD suppliers
Korea - Semiconductor tariffs about 10% see major reduction for OECD suppliers
Considering the idea’s scope, I thought Doug would suggest scaling it back to be more manageable. The ICA would be a preferential trade agreement that would be sector-focused and permitted under MFN treatment – but an unusually large one.
Surprise: His response was – “Go for it !”
Initial Thoughts on Process
Up to this point, we have followed a series of trade-liberalization initiatives that helped open markets for the rapidly expanding information technology industry. Most were regional or bilateral in nature. But during the same period, something much larger was taking place: the transformation of the global trading system itself.
The General Agreement on Tariffs and Trade (GATT), created in 1947, had been remarkably successful in establishing rules for international trade and progressively reducing tariffs on manufactured goods. But by the 1980s, the nature of the global economy was changing. Trade was increasingly about services as well as goods; intellectual property was becoming an essential element of international commerce; and governments were using regulations, standards, and other measures that could restrict trade even when conventional tariffs were low.
The Uruguay Round of multilateral trade negotiations, launched in 1986, was an attempt to address this much broader world economy. The negotiations eventually produced agreements covering services, intellectual property, agriculture, textiles, dispute settlement, and numerous other subjects. Most importantly for the story that follows, they also created a new institution—the World Trade Organization (WTO)—to administer the resulting system. The Marrakesh Agreement establishing the WTO was signed on April 15, 1994, and the WTO came into existence on January 1, 1995.
Meanwhile, the United States Congress had been following the negotiations closely and passed the Uruguay Round Agreements Act (URAA) on December 8, three weeks before the WTO agreements were scheduled to take effect. As part of the Act, the Office of the U.S. Trade Representative was statutorily codified as the lead federal body responsible for all WTO representations, multilateral panels, and everyday policy coordination.
IBM and other high-tech and non-high-tech companies had been very much engaged in lobbying Congress for enactment of the URAA, because there were so many issues of strategic importance to their economic performance and success – such as trade in services, intellectual property protection, tariffs, and others. We understood that Congress had locked in the authority for USTR to be the lead agency responsible for the WTO, but a technical question remained open in our minds. Did USTR have the authority to negotiate agreements above and beyond the concessions offered in the Uruguay Round?
The answer was yes. Under section 111(b) of the implementing legislation, the President was given authority to proclaim tariff changes resulting from subsequent WTO negotiations involving categories of tariffs that had been the subject of duty elimination or harmonization talks during the Uruguay Round. And with this, as we will see shortly, Congress had left a door open without knowing how important it would become.
Wait. This was happening concurrently with the regional trade agreements discussed earlier. The difference was that the Uruguay Round was a global initiative. So what was this “zero-for-zero” approach? In essence, it referred to a specialized, reciprocal method of tariff elimination where participating countries agreed to reduce to zero on both sides within specific industrial sectors. Those principal participants in these discussions were the so-called “Quad” – the United States, the European Community, Japan, and Canada. The approach itself worked inside a two-pronged framework:
Reciprocity: Traditional bargaining in GATT boiled down to “I will lower my tariff on X if you lower yours on Y.” Under the zero-for-zero approach, negotiations changed to “Let’s cut tariffs to zero for a particular sector, and everybody will benefit.”
MFN Application: Even though these discussions took place among a subset of primary supplying and importing countries, the duty-free treatments that were agreed to were extended to all GATT members on a Most-Favored-Nation basis.
A number of industry and manufacturing sectors where developed economies held a strong comparative advantage were successfully addressed by this approach – pharmaceuticals and medical equipment, construction and agricultural equipment, steel, furniture, paper and paper and wood products, and others.
But why not information technology products?
The United States had entered the Uruguay Round with an exceptionally ambitious proposal for electronics. That position was heavily influenced by the American IT industry’s view that tariffs throughout the global supply chain were increasingly difficult to justify. Computers depended on components; components depended on semiconductors; semiconductors depended on sophisticated manufacturing equipment. The supply chain was not simply important; it was essential. A tariff imposed anywhere along the supply chain could significantly increase costs throughout the system.
The U.S. government’s Uruguay Round proposals were, of course, influenced by lobbying from numerous sectors of the American economy. Most industries wanted tariff cuts that would benefit their particular products. The information technology industry differed in one key respect: It increasingly viewed the entire technology production chain as an interconnected system.
For the information technology industry, U.S. negotiators sought input from three key groups –
The American Electronics Association (AEA) – the largest electronics trade association, representing thousands of companies. After surveying its membership where U.S. industry found severe market access barriers (both tariff and non-tariff barriers), AEA gave USTR a definitive list of computer hardware and networking peripherals that would require zero-tariff status.
The Computer and Business Equipment Manufacturers Association (CBEMA) – including large U.S. high tech companies like IBM, DEC, and HP. CBEMA’s approach was to work with U.S. negotiators on a line-by-line basis in the Harmonized System for the U.S. zero-for-zero proposal. They focused on ensuring that computer parts, subassemblies, and electronic assemblies were covered, knowing that in the rapid evolution of the information economy globally and its deeply integrated global supply chains, a tariff on a single component could harm the competitiveness of the final machine.
The Semiconductor Industry Association (SIA) – including microchip giants like Intel, AMD, and Texas Instruments. Their push was to ensure that microprocessors, integrated circuits, and memory chips were included in the U.S. offer in the Uruguay Round. Essential to their list of requests were the very large wafer-fabrication and chip-testing equipment used in making semiconductors. From their viewpoint, the U.S. had a dominant edge in semiconductor manufacturing equipment, so getting other countries to drop tariffs on these machines was not just a top priority – SIA felt it should be at the center of the U.S. offer.
In other words, the industry itself was already a matrix before international negotiations even started. The position of the three associations illustrated an important characteristic of the emerging IT economy. It was no longer possible to think about “computers,” “semiconductors,” or “electronics equipment” as stand-alone industries. They were increasingly components of a single, interconnected technology ecosystem. The industry was asking USTR for something much broader than simply eliminating the tariff on a particular finished product.
That was part of the problem – There were still key differences from country to country in tariff classification. A product that appeared to be an information technology product under one country’s tariff schedule could fall under a broader tariff category in another country – one that included products the country had no intention of liberalizing. As technology evolved and as we have seen, these classification problems became increasingly difficult to resolve through traditional tariff-line negotiations.
And there was a structural problem in the negotiations. The Uruguay Round largely took the approach of traditional tariff-line negotiations, while the IT industry was increasingly composed of products whose technology and functions were evolving faster than government bureaucracy could keep up.
As a consequence, the idea of IT inclusion in the final Uruguay Round zero-for-zero tariff reductions fell apart in the final hour of negotiations. But here’s the point: The failure was not because countries had rejected the idea of free trade in information technology. The failure was more a failure of the Round’s negotiating process than a rejection of the objective. And this failure actually created an opening. The broader objective that the IT industry sought was still possible–but it would require a fundamentally different negotiating mechanism.
Earlier, we described the initial liberalization initiatives in North America as putting the pieces of a puzzle in place. The International Computer Agreement (ICA) proposal would take that process much further. It would not simply add another piece to the puzzle. It would require us to connect pieces that had never before been thought of as belonging to the same puzzle.
And that was the real challenge.
A broader, more inclusive industry-wide approach to negotiations would require a meeting of minds among an almost unheard-of number of stakeholders from a multifaceted and loosely connected industry in the U.S., to industry groups on both the supply and demand side of the equation abroad, to tangentially related industry groups in various countries (many of whom opposed liberalization), to a number of Administration agencies in Washington, to similar agencies in other countries involved in the negotiations, to the United States Congress.
This was no longer a simple puzzle. It had become a multi-dimensional matrix.
Testing the Waters
Cultivating U.S. Government Support
We had to start somewhere, and frankly, the questions were somewhat daunting. Could we persuade the U.S. Government to take an idea that originated with two industry people seriously enough to become a negotiating initiative. Further, since GATT had already begun to consider how best to establish the new World Trade Organization, could we somehow wedge the ICA concept into those negotiations? And very key – since the WTO formulation talks were already well underway, and there were no planned discussions among trade ministers for other purposes, could we get an ICA approved fast enough at the international level? We did have a preliminary approach in mind based on a few key facts –
that serious discussions had taken place during the Uruguay Round about tariff relief for information technology products,
That we had had significant U.S. Government support during the Round for this position,
that the Office of the U.S. Trade Representative had authority to negotiate new trade agreements as long as USTR did not go beyond the parameters established for Uruguay Round negotiations,
that computers, computer parts, peripherals, semiconductors, and semiconductor equipment were a good target for our efforts, and
that we could demonstrate that revolutionary technology development and supply chain dependencies were major concerns of the OECD countries.
Doug and I decided to test the waters on whether we could get U.S. Government endorsement for our International Computer Agreement idea. Our choice to go to the government first may seem strange at first. We hadn’t even broached the idea with IBM’s CEO, John Akers. Nor had we discussed it with other industry representatives – why not get a number of companies behind the idea first? And as Doug showed by his desire to push forward aggressively after the success of negotiating the CXT for NAFTA, I shouldn’t have been surprised by his aggressiveness once we had outlined the ICA concept. Indeed, he said in essence, “We need to fast-track this.” Why, in other words, not go straight to the negotiators themselves; i.e., USTR?
Any of those approaches could have made sense, but Doug’s instincts were to go straight to the source decision-making (the government). Critical to our success, we felt, was identifying key advocates in the government to help advance our case. So we started with two individuals who knew our industry, its dependence on supply chain certainty, and who could suggest how to proceed. We went to John Simpson of Treasury first, stressing that the Common External Tariff (CXT) in the NAFTA agreement was a major step forward in having public policy recognize the realities of global supply chain dynamics. We noted that when considering tariffs, asking, “What tariff treatment makes sense for the entire class of IT products,” was far more practical and preferable than the traditional question, “What tariff should each country impose?” John, as one of the principal advocates of the CXT in NAFTA, understood the point well.
So, we told him about our idea. We noted that we had confirmed that Congress had granted negotiating authority to the Administration for items that we discussed in the context of the Uruguay Round. We told him that we needed Administration support for not only a renewed effort to get tariffs reduced on the computer sector, but also suggested that this could be done by expanding the CXT idea to include all OECD countries. It was the first time that we had, in contacts with the U.S. Government, formally suggested the idea of an International Computer Agreement (ICA).
John offered five comments: (1) We needed to be specific on the products that we proposed to be covered under an ICA, so we should confirm our suggested product list with the Customs Service; (2) The most logical place to raise the issue was in the context of the new negotiations in GATT to create a World Trade Organization; (3) We were suggesting sector-specific talks and that would require a shift from line-item tariff negotiations thinking among trade ministers to tariff elimination for an entire sector; (4) We would have to demonstrate that we had widespread industry support for the idea; and, (5) We were coming to the Administration shortly after the Uruguay Round and U.S. legislative implementation, and from his observations, there was considerable fatigue to overcome in USTR and elsewhere in the Administration.
In retrospect, John had given us our marching orders. And he asked that we keep him in the loop.
Before talking to Customs, we decided to take his advice and explore whether we could get industry support for the idea. In a sense, we started close to home. Washington, DC is a virtual alphabet soup when it comes to trade and industry associations, and that is especially true of high-tech associations. Virtually across the street from our office on K Street, we knew the Computer and Business Equipment Manufacturers Association (CBEMA) very well. IBM chaired a number of CBEMA committees, including the international trade committee. So we approached Rhett Dawson, president of CBEMA. Rhett was in the process of rebranding the association which would then go by its new name – the Information Technology Industry Council (ITI). He felt that the ICA would be a great way to establish a new identity for the ICA, and he asked Meghan Rainey of his team to work with us.
Meghan and I decided that Matt Rohde of Customs was the best place to start. We laid out the ICA concept as Doug and I had done with SImpson. Matt’s reaction mirrored Simpson’s and thought the ICA was worthy of support, assuming we could encourage USTR to overcome its institutional fatigue due to the recent Congressional Uruguay Round approval process. He suggested that we talk to Dorothy Dwoskin. She had recently been appointed to be the Assistant U.S. Trade Representative for World Trade Organization (WTO) and Multilateral Affairs.
The Start of a Global Coalition
Meghan and I asked Matt to join us, because we knew that he and Dorothy had worked closely during the Uruguay Round. For the first time, we experienced pushback, or more accurately, muted skepticism. Dorothy had doubts about whether a push for the ICA would be successful, largely due to the EU’s opposition to accept proposals to eliminate computer and semiconductor tariffs during the Round. But she didn’t discourage our effort either. She laid out two pre-conditions that we must meet before we could present the concept to the U.S. Trade Representative –
We must get broad U.S. computer industry support behind us. That meant the ITI companies for sure, but preferably, companies from other (and mostly much larger) industry associations as well.
And more worrisome to Meghan and me (and Doug Worth), we must secure industry support in Europe, Japan, and Canada. That meant organization on a multinational scale, something that we really didn’t have much experience with.
The original ICA concept evolved from expanding the NAFTA/CXT model to the OECD, but the scope of our effort kept expanding as we encountered the realities of international trade negotiations at the same time that the scope of the global technology revolution was altering the shape of the high-tech industry itself. Importantly, it addressed a question: Why shouldn’t an industry that wanted a trading system suited to the needs of revolutionary technology and interlinked supply chains pull together for such an effort?
An easy question to ask, and at first, it looked like it would be easy to organize the effort. The American side was relatively easy. Using ITI as our home base, we met separately with four other key computer companies – HP, Digital Equipment Corporation (DEC), Compaq and Apple. Plenty of questions, but in the end, all of them enthusiastically endorsed the ICA.
Not so easy when we addressed Dwoskin’s “pre-condition” that we get industry support in the EU, Japan, and Canada – and it would require a different flavor of “alphabet soup” organizations of business abroad.
European Cooperation & Obstacles: In Europe, our initial contact was with Eurobit, a very strong industry trade group that represented the interests of European information and communications technology manufacturers. And it turned out that one ITI member company, DEC, had a good working relationship with the Eurobit president, Guenther Moeller. We anticipated that Moeller would have some reservations. During the Uruguay Round, another association called the European Electronic Components Association (EECA) had lobbied successfully to retain semiconductor tariffs. Nevertheless, Moeller agreed to support our efforts and even to try to open doors to EECA to get its support.
Japan: Bear in mind that under the ICA proposal, absolutely nothing in terms of tariff concessions would be asked of Japan. We thought, however, that the industry group in Tokyo, called JEIDA, might be a problem. We feared – and the fear was proven to be well-founded – that JEIDA would be concerned that old habits in trade negotiations might resurface. Put shortly, JEIDA and the Ministry of International Trade and Industry (MITI) would oppose the ICA if ANY tariff concessions might be requested by USTR or the Europeans. After all, Japan had eliminated all ICA-covered product tariffs already. Why should they give up more than what they had already given up?
The Trilateral Industry Alliance: During this time frame (1994-1995) and disappointed with the Uruguay Round’s failure to completely liberalize electronics industry tariffs, the three industry groups we had contacted – ITI, JEIDA, and Eurobit – created an informal “Trilateral Alliance.” As representatives of a global industry, the leaders of these associations strongly believed that the ongoing discussions in GATT to create the WTO represented our best and last chance to get trade policy to reflect the requirements of a rapidly changing industry sector.
Meghan Rainey and I attended their first organizing meeting in Brussels where we presented the ICA idea, asking for the associations’ formal endorsement. We hoped to take that back to Dorothy Dwoskin. As we had anticipated, JEIDA was concerned that the ICA could pose a threat to the tariff liberalizations that had occurred under the U.S-Japan Semiconductor Agreement. After reassuring the JEIDA delegation of ITI’s position supporting theirs, JEIDA agreed to support the agreement in writing as long as ITI and the U.S. computer companies would ensure (to the extent we could) that USTR would demand any concessions from MITI.
The Birth of the Information Technology Agreement (ITA)
And then, the big surprise.
We understood from our discussions prior to the alliance meeting that Eurobit would support the ICA. But unexpectedly, Guenther Moeller told us that he believed that our proposal wasn’t being aggressive enough – the product scope was too narrow. The ITA should cover ALL Information Technology products, not just computers. Specifically, the position of Eurobit was that the proposed agreement should include semiconductors and telecommunication as well.
Meghan and I resisted this, aware that semiconductors, due to the positions of EECA during the Uruguay Round, had killed the zero-for-zero tariff proposals. After lengthy discussion (and debate), Moeller told us that Eurobit’s position was “take it or leave it.” And Moeller surprised us one more time. He said that in order to take into account the broader product coverage, we should call our initiative the “Information Technology Agreement.”
Doug Worth and I had developed the ICA idea and taken it to the coalition. The three associations had taken it under consideration and after deliberations decided to broaden its scope. And from that emerged a new broader proposal under a different name to take to Dorothy Dwoskin and the USTR, Mickey Kantor.
Continue to Section 3: Building the Coalition, Building the Consensus →