No U.S. company or industry is immune from the impact of decisions made in Washington. But many corporate executives still act as if politics is an exercise in crisis management—something to worry about after trouble comes.
Ignoring Washington until you need it may have worked 25 years ago. Today it is a prescription for failure. Divisions within the executive branch, the erosion of party discipline on Capitol Hill, and the sprawling power of congressional staffs have rendered obsolete once reliable channels of access and influence. Increased political competition among businesses further complicates the situation. In the early 1960s, fewer than 150 companies maintained Washington offices. Today the number is close to 700. Some 2,500 companies have other forms of Washington representation, and more than 15,000 lobbyists work the halls of Congress and the regulatory agencies.
Executives must bring to politics the same long-term perspectives they apply to marketing and investment decisions. This is especially true for companies or industries whose size and reach don’t provide a natural base of influence. Manufacturers of steel, automobiles, or textiles expect a sympathetic hearing in Washington—too many jobs in too many congressional districts are at stake. But politics is very different for smaller industries that don’t affect millions of people, don’t have employees spread across the country, or can’t draw on the clout of national labor unions.
Like a superior business strategy to niche players in a crowded market, intelligent and carefully executed political strategies are crucial for smaller industries to succeed in Washington. The challenge is to create and sustain political advantage—to develop a stable and constructive relationship that makes Washington an ally in the battle for global competitiveness. The process of building political advantage requires high-level executive attention and action. If government is important to an industry’s competitive future, political activism must be a business priority.
The political effectiveness of the semiconductor industry shows what it takes to succeed. The design and manufacture of semiconductors is a small industry by any standard. Merchant suppliers (companies that sell chips to other companies) employ fewer than 115,000 U.S. workers and generate revenues of $12 billion—just over 10% of the revenues of General Motors. Employment is focused geographically; more than half the industry’s work force lives in Arizona, California, and Texas. And most of the top semiconductor companies were founded by self-assured entrepreneurs dedicated to the virtues of small enterprise and corporate self-reliance. For much of their early history, these driven industrialists resisted outside alliances of any sort, especially with Washington. These are hardly the ingredients from which political juggernauts are fashioned.
Yet semiconductor companies have compiled one of the most impressive political performance records of the 1980s. In 1983, the industry established an ambitious agenda for government action: more favorable tax treatment, relaxed antitrust regulation of joint research and development, greater protection of chip designs, and several tough policies to promote fair trade. The industry had realized these objectives and more by the end of 1987. (See the insert, “Big Results for a Small Industry.”)
The track record has been most striking on the trade front. The semiconductor industry has nursed serious grievances about Japanese trade practices since the mid-1970s. Once they decided to act, U.S. companies needed less than two years to convince Washington to take aggressive action on their behalf.
The formal campaign began in June 1985, when the Semiconductor Industry Association (SIA) filed for relief under Section 301 of the Trade Act of 1974. Section 301 authorizes the president to penalize countries that deny U.S. products fair access to their markets. Fourteen months later, the United States and Japan signed an accord that gave the industry the assistance it sought. The Japanese agreed to open their markets to foreign semiconductors and to stop selling chips below cost. In April 1987, concluding that Japan was violating the accord, the Reagan administration approved $300 million in punitive sanctions—the first such penalties against Japan since World War II. This series of measures is unprecedented for its swiftness, severity, and agreement with industry recommendations.
Why have semiconductor companies succeeded in Washington when other industries of comparable and greater size (telecommunications equipment, footwear, automobile parts) have not? There are no magic formulas for influencing government. Political environments change, public officials come and go, opponents and allies shift overnight. But the semiconductor industry’s experiences illuminate four general principles for making things happen in Washington.
1. Companies need a united front. Small industries must develop and maintain alliances among competitors, suppliers, and customers. Such ties expand the range of affected constituencies, increase the resources available for political action, and defuse potential sources of opposition.
2. Government allies are essential. Friends inside the government are as important as corporate allies. An industry should identify and cultivate executive branch agencies and members of Congress with stakes in its agenda. This means crafting positions that will appeal to targeted officials and factoring their interests and agendas into the industry’s own political calculations. An industry should also identify potential adversaries in government and take steps to minimize their impact.
3. CEOs have a special role to play. A visible, persuasive, accessible CEO is often more effective than dozens of hired political guns. Top executives can overcome “barriers to entry” in Washington. In the early stages of political activism, gaining access to public officials and raising industry visibility can be especially difficult. Lobbyists often struggle to get time with a senator or a cabinet member. Business leaders can get to them more easily. When CEOs make promises, they can deliver; lobbyists are only messengers. Politicians know this.
4. Political action by company executives is more effective than trade association efforts. Trade associations can play important roles in setting agendas, monitoring political developments, and establishing contacts in Washington. But company managers—senior executives, middle managers, plant supervisors—should be on the front lines. Managers understand the details and subtleties of their industry; it’s their livelihood. And precisely because it’s their livelihood, they may be more driven to get results.
The modern semiconductor industry was born in 1959 with the invention of the integrated circuit. The industry has experienced three phases of political engagement. The first 15 years were marked by an arm’s-length relationship with government. U.S. companies were global technology leaders in all segments of the chip market, industry revenues exploded, and many millionaires were born. During this expansion period, NASA and the Pentagon were important customers, but chip makers wanted the government’s business—not its meddling.
This attitude began to change in the mid-1970s with the first stirrings of foreign competition. During the 1975 recession, U.S. semiconductor companies scaled back plans to add production capacity much as they had in previous recessions. Then, when the economy rallied, they were caught short of capacity. This time the Japanese rushed in to meet excess demand. When Japanese success in commodity chips threatened to put certain U.S. producers out of business, a few of them turned to Washington for help. But their proposals for government intervention fell on deaf ears.
Gradually, as Japanese competition seemed to verge on domination, more chip makers began to look to Washington. U.S. companies accounted for 55% of world semiconductor production in 1978; by 1987, their share had dropped to 44%. During this same period, Japanese market share rose from 28% to 50%. Alarmed by these trends, semiconductor companies made political activism a priority.
Trade policy has been an issue of particular urgency. Its importance was rooted in one of the industry’s most striking characteristics—that production costs for most products declined by 30% for every doubling of cumulative volume. This is because semiconductor manufacturing lines frequently turned out more defective than sound chips. With new products, yields were often as low as 25%, even for the best companies. As products matured, however, yields would run as high as 90%.
The need to raise yields led companies to manufacture high-volume products that could act as “technology drivers.” It was generally believed that skills learned in manufacturing large quantities of a simple product could be transferred to more complicated, higher value-added devices and help “drive” the company down a steep learning curve. Dynamic random access memory (DRAM) chips, a 1971 American invention, were the most widely used technology driver for many years. Other technology drivers included static random access memory (SRAM) chips and erasable programmable read only memory (EPROM) chips.
During the past ten years, however, U.S. companies have lost market share in each of these critical products. In 1975, for example, U.S. companies accounted for 90% of world DRAM shipments. By 1987, their share was just over 20%, an erosion suffered almost exclusively at the hands of the Japanese.
U.S. merchant producers have articulated two basic objections to Japanese practices. First, for decades U.S. companies have been unable to increase their 10% share of the Japanese market, despite wide swings in the dollar’s value and their consistent success in Europe against the Japanese. De facto exclusion from Japan, which by 1986 had become the world’s largest market for semiconductors, put American manufacturers at a serious competitive disadvantage in technology drivers. Second, U.S. companies claimed that the Japanese were selling commodity chips in the United States below production costs. This alleged dumping not only cost U.S. companies market share, it also put tremendous pressure on industry profits. U.S. merchant producers lost a staggering $2 billion in 1985 and 1986. This dismal outcome was largely the result of a global industry downturn (Japanese producers also lost $2 billion), but dumping in the U.S. market made a bad situation worse.
It’s easy to see why most semiconductor executives eventually agreed that the trade situation and their collective market position would deteriorate further without aggressive intervention from Washington. The challenge became translating the goal of fair trade into the reality of government policy.
A crucial choice for executives trying to build political advantage is whether to advance the interests of their own companies or to support industrywide initiatives. Political activism is expensive, so only big corporations can consider a company-based agenda. The startup investment for a bare-bones Washington operation—an office, a secretary, and a full-time lobbyist—typically runs about $1 million.
Even for large companies, industrywide campaigns are often preferable for issues where rivals share common political interests. For small companies, forging common political ground is a necessity. A politician evaluating a proposal from business usually searches for answers to several basic questions: What will the proposal cost? What sectors of the economy benefit? What sectors lose? No politician wants to help one industry if it means antagonizing three others. That’s why coalitions are so important from a strategic perspective. Just as powerful suppliers and customers squeeze profit margins in the marketplace, unrestrained political competition among rivals, suppliers, and customers usually reduces everyone’s influence.
The defeat of several early proposals for government assistance taught semiconductor industry leaders the importance of building coalitions. During the first 20 years of the industry’s existence, it had no organization to represent its interests in Washington. Some companies belonged to large electronics trade associations, but these groups’ agendas did not reflect semiconductor industry priorities. So in 1977, five leading merchant producers of high-volume chips—Advanced Micro Devices (AMD), Fairchild, Intel, Motorola, and National Semiconductor—founded the Semiconductor Industry Association.
The SIA and individual companies floated several initiatives against Japan during the late 1970s, and political leaders did what political leaders invariably do. They sought the reactions of other players in the industry as well as outside constituencies. In this case, the key outside constituency was the biggest buyers of chips—giant computer builders like Hewlett-Packard and Digital Equipment. The users’ primary concern was maintaining reasonable prices and flexible supplies. Invariably, they opposed SIA initiatives, which effectively doomed the proposals. A string of defeats convinced semiconductor leaders that they had to expand their base of business support.
The process of building coalitions took two forms. Merchant producers cultivated active support from companies and industry associations that had opposed previous SIA initiatives. They also worked to discourage active opposition. If a company or association could not endorse a semiconductor proposal, it might be persuaded at least not to undermine it.
One important step in the coalition-building process was expanding the industry’s trade association membership. SIA leaders had been working to increase membership since 1977, but in the early 1980s they made a particular effort to enroll the large captive producers of chips. (Captive producers build chips primarily for use in their own products.) IBM joined the SIA in December 1980, and during the next few years, so did other big captive producers. By 1985, the SIA had 48 members with combined revenues of more than $100 billion. The organization included some large merchant producers like Texas Instruments that had initially resisted joining the group, smaller companies building custom chips, major chip buyers like Control Data and NCR (most of which also produced their own chips), and giant captive producers like IBM and Digital Equipment.
The most important new member, of course, was IBM, the world’s largest manufacturer and consumer of semiconductors. IBM never played a leadership role on trade issues, but the company’s envoy conveyed to the SIA what positions IBM could and could not live with. Moreover, IBM’s membership gave the SIA a level of credibility and visibility in Washington that it had never before enjoyed.
Collecting in one organization large and small merchant producers, captive producers, and important users also created a means by which each industry segment could develop a better understanding of other segments’ interests and priorities. Debates within the SIA were, in effect, a negotiating process that produced consensus on particular initiatives. Even if a member company was unenthusiastic about a final SIA proposal, it was unlikely to sabotage the initiative in Washington. It had participated in the internal debates and understood how and why the SIA had reached the position.
The evolution of the Section 301 petition is a case in point. Agreement on the terms of the complaint took three years of bargaining among SIA members. A critical issue was whether trade sanctions should be proposed and, if so, how severe they should be. Some large merchant companies favored an outright embargo on certain commodity chips, or at least the imposition of high tariffs and stiff fines, unless Japan met a timetable for increased U.S. access to its markets. Large users like IBM and Hewlett-Packard rejected this proposal outright. Merchant suppliers realized that user opposition would pose huge political obstacles, as it had with previous initiatives, so IBM’s and other companies’ objections carried great weight in the SIA.
The petition filed in June 1985 did not advocate punitive embargoes, tariffs, or other restrictions on chip supplies. It proposed two tough but constructive policies. First, it demanded a commitment that by the early 1990s the U.S. share of the Japanese semiconductor market increase commensurately with the U.S. position in the rest of the world. Penalties for noncompliance were left vague, although the SIA made it clear that it expected penalties if negotiations failed. Second, the petition suggested a monitoring system to ensure that Japanese companies were not selling chips below cost in the United States or elsewhere. All SIA members could live with these positions.
The process of building coalitions—or at least of neutralizing potential opposition—extended beyond industry borders. George Scalise, chief administrative officer of AMD and chairman of the SIA’s public policy committee, worked hard to secure an endorsement of the 301 suit from the American Electronics Association. The AEA, which has been a Washington presence for decades, represents some 2,800 companies with $305 billion in global sales. Scalise understood that strong opposition from the electronics group might doom the 301 action.
The AEA never fully embraced the SIA position, but it did issue a letter supporting the objectives of the trade case. Coordination with the AEA was also important after adoption of the August 1986 trade agreement. Prices for certain DRAMs quickly skyrocketed, which shook up small chip buyers. The SIA established a subcommittee within the AEA to address user concerns and dampen possible calls for repeal of the agreement.
Target Government Allies
A sound marketing strategy balances product positioning, distribution channels, and reasonable prices. A sound political strategy has similar balances. An industry must position a proposed course of action so as to appeal to the customer (White House officials, regulators, or members of Congress), target the most effective channels of influence (congressional committees, executive agencies, or the courts), and impose reasonable costs on taxpayers and other constituencies. Pushing the right buttons, identifying government allies, and working hard not to antagonize potential adversaries are all essential. Failure in any area usually dooms the entire initiative.
The semiconductor industry paid close attention to all three areas. The 301 complaint’s primary objective was opening foreign markets rather than closing U.S. doors. This positioning meant the trade action appealed simultaneously to “Japan bashers” and free traders. It certainly satisfied the White House, which liked the idea of appeasing protectionist sentiment on Capitol Hill by supporting a trade initiative that rejected quotas and tariffs. Representatives and senators with protectionist leanings were also satisfied. Even legislators whose districts included no semiconductor employment were attracted to the cause. They could go home and tell voters they were being tough on Japan, and not antagonize other corporate interests. In short, virtually every set of actors in Washington had something to gain by supporting the skillfully crafted industry position, and few had anything to lose.
As for distribution, the industry used a number of complementary political channels. A common strategy of smaller industries is to avoid “politicized” channels (Congress and the White House), where votes matter, and pursue “administrative” channels (the courts and regulatory agencies), where cases are won or lost on their merits. When the data are clear and convincing and the issue is a one-time problem, administrative remedies are usually adequate. If the objective is to make government an ongoing ally, however, companies and industries must tap political channels and find ways to build alliances with public officials.
A Section 301 petition is one of the more politicized avenues for relief under the trade laws. The U.S. trade representative rules on the merits of the case, and the president determines appropriate sanctions. But the law leaves much room for maneuver. The trade representative can postpone the decision for up to one year, and the president can choose not to act even in the face of a favorable ruling. So the semiconductor industry took steps to heighten the visibility of its trade complaint and the costs of complacency.
One tactic was for companies to file their own trade actions, even without SIA endorsement. Days after the SIA filed the 301, Micron Technology charged that Japanese companies were dumping 64K DRAMs. A few months later, three industry leaders (Intel, AMD, and National Semiconductor) accused the Japanese of dumping EPROMs. Even the Commerce Department, anxious to get tough with the Japanese, initiated its own dumping complaint on 256K and higher DRAMs in December. By the end of 1985, the U.S. government faced four separate complaints on unfair semiconductor trade practices, which put even more pressure on the administration and the Japanese to reach an accommodation.
Capitol Hill was another focus. Congress has no direct role in complaints filed under the trade laws, but the SIA approached representatives and senators to support the 301 petition by lobbying the administration and drawing media attention to the issue of unfair trade practices. The association set a very specific objective: develop legislative support that was bipartisan, bicameral, and as geographically broad as possible. The result was the Congressional Support Group, a caucus composed of ten representatives and ten senators, ten Democrats and ten Republicans, from states including California, Missouri, Florida, and Pennsylvania. The legislators applied pressure on the White House by making telephone calls, lobbying members of the cabinet, and pushing the trade representative to become more involved in the trade dispute.
The SIA also persuaded 180 representatives and senators—including the entire delegations from several states with heavy semiconductor employment—to send letters (drafted by the SIA’s general counsel) to the administration. Meanwhile, the California delegation met with the Japanese ambassador to the United States to emphasize the gravity of the issue. By the time the White House decided to negotiate a semiconductor agreement, the SIA had made as many allies inside the government as it had developed among suppliers and customers.
Many political observers believe that the Pentagon figured prominently in the semiconductor dispute. They argue that microelectronics is so important to high-technology weapons and communications systems that the Defense Department must have been the natural ally of U.S. chip producers. How could the Pentagon tolerate Japanese domination of a commodity dubbed the “oil of the 1980s”?
The truth is, the Defense Department never weighed in strongly on semiconductor trade policy. One of the department’s long-standing priorities was to persuade Japan to increase its military spending; trade disputes wouldn’t advance this goal. The SIA understood this and kept its distance from the defense establishment until after trade issues were settled. Later, when it was looking for funding for its Sematech research consortium, the industry enlisted the Pentagon’s support. The fiscal 1988 defense appropriations bill included $100 million for Sematech.
Get CEOs Involved
Broad membership in an association or informal coalition helps build legitimacy in Washington. Effective political targeting further improves the odds of success. But who does the work?
Here again, semiconductor companies made the right moves at the right times. Even before they had forged an industry consensus on trade policy, they had worked hard to establish a beachhead in Washington. Like companies in so many other industries, chip makers turned to political activism only after troubling market trends had surfaced. Like the marketplace, however, politics usually rewards early movers. The first companies that enter political life make the campaign contributions, build the key congressional ties, and establish the reputations that translate into access and influence. Latecomers can find themselves at the end of a very long line.
Japan sells more chips… (percent share of world DRAM shipments)
The semiconductor industry recognized it needed a distinct approach. So it called on the charisma and persuasiveness of its senior executives, who proved to be a very valuable resource. Their technical achievements and business celebrity gave them great visibility on Capitol Hill. Chairmen, presidents, and other high-ranking executives traveled to Washington to meet with cabinet secretaries, senators, and members of Congress and to testify before congressional committees. Especially prominent from 1979 through 1981 were Robert Noyce, vice chairman of Intel, Charles Sporck, CEO of National Semiconductor, W. Jerry Sanders III, chairman of Advanced Micro Devices, and Motorola Chairman Robert Galvin.
These men opened many doors that might have been closed to lobbyists or lower ranking businesspeople. Noyce, for example, is something of a legend in the electronics world. A multimillionaire, coinventor of the integrated circuit, he was general manager of Fairchild Semiconductor during its rise in the 1960s and one of the founders of Intel. The Washington establishment wanted to get to know him as much as he wanted to develop political contacts. Noyce spent 20% of his time during the early 1980s on political action. He and his colleagues in effect softened up Washington for the formal campaign that began in 1985.
Politically engaged CEOs are an effective weapon against the ever-increasing Washington barriers to entry. They are also important for maintaining political ties. Even the most loyal customers need periodic attention from top brass; so do politicians. Noyce and his industry peers still visit Washington a few times a year, every year, to stay close to the players.
Put Managers on Front Lines
Of course, top executives can’t do the whole job. Over the long term, companies face another critical choice: whether to rely on their trade association’s staff and paid lobbyists or to involve other corporate managers. The nature of the agenda determines the right answer.
When an industry is involved in maintaining the political status quo—staying in touch with Capitol Hill, conducting routine public relations, monitoring the regulatory agencies—the best strategy is to use trade association professionals. Large groups like the National Association of Manufacturers and the Food Marketing Institute exist mainly to keep tabs on Washington and influence technical regulations and bills. It makes sense for large association staffs to run these activities. As the issues become more urgent, however, and the political agenda moves from reacting to initiating new policies, managers should become involved directly. The more consequential the issue, the higher the level of manager needed.
…and buys more too (semiconductor consumption in millions of dollars)
The semiconductor industry never had the luxury of maintaining the political status quo. By the time it turned its attention to Washington, the trade situation was urgent. So the companies agreed to dedicate money and some of their own people to government relations. Time and again, managers from the large manufacturers came to Washington to support the trade initiative. Chip producers and users, including AMD, Digital Equipment, General Instruments, Harris Corporation, Intel, International Rectifier, Motorola, National Semiconductor, Rockwell International, and Texas Instruments committed executives to help build congressional support, visit trade officials, generate press releases, and for other work to build political bridges. AMD’s George Scalise devoted 25% of his time to managing the trade case.
Washington lawyer Alan Wolff, a former deputy special trade representative, supplemented these industry lobbying efforts. The combination of experienced outside counsel and hands-on executive involvement proved powerful. Wolff provided the SIA with government contacts and a good feel for what agencies would be open to what positions. But many of the foot soldiers in the campaign were company managers whose expertise and commitment made them very persuasive with public officials.
This pattern of direct executive action explains why the SIA has remained such a lean organization. Despite its ambitious political agenda, the SIA employs only six professionals and two government affairs executives. Its annual budget runs less than $1 million, and the association doesn’t have a Washington office. The Japanese, on the other hand, reportedly spent between $30 and $50 million lobbying against the trade initiatives.
In less than a decade, the semiconductor industry has built a constructive and nonadversarial relationship with Congress and the executive branch. These ties have produced a series of policy initiatives which, while not ideal, have had overwhelmingly positive consequences for the industry. U.S. semiconductor companies still have serious competitive problems, including quality and manufacturing costs. But government intervention has improved short-term profitability and cash flows and long-term prospects for research and development.
The work of the semiconductor industry in Washington is not over. For government to be a reliable ally and partner, the relationship must be stable and ongoing. A shotgun approach to politics—get what you want and don’t return until you need something else—simply won’t suffice. So the industry remains mobilized. The SIA and member companies continue to monitor Japanese chip prices and foreign access to the Japanese market and to supply the data to Washington. This vigilance helps explain why the Reagan administration imposed the sanctions in April 1987. Literally hundreds of violations of similar agreements have occurred over the past 30 years, but no administration has taken action as severe as the semiconductor sanctions.
Most American companies continue to have an adversarial relationship with Washington. Business leaders complain about government bureaucracy and inefficiencies. Public officials complain about special interest groups maneuvering for their own side deals, making it impossible to fashion coherent national policy. Even in the semiconductor case, business-government relations have not always been smooth. Some companies have worked to undermine SIA proposals; certain government officials have tried to obstruct semiconductor policies.
We must build more constructive business-government relationships in the United States, which means not always assuming that government is the enemy. Many public officials are eager to help business in the battle for global competitiveness, especially when business frames its proposals in a way that appeals to the officials’ own agendas and priorities. The process of building political advantage may be especially challenging for smaller companies and industries, but it is far from impossible, especially if they focus their efforts. Semiconductor companies did not try to shape the entire range of policies affecting them. They targeted certain issues that were central to their future and where their case was strongest.
Not every industry can expect to duplicate the semiconductor producers’ success in Washington. But if U.S. companies are serious about regaining world competitiveness, the semiconductor experience must become the rule, not the exception.