After reading the Intel case (Intel Whale in a swimming pool), write 5 paragraphs on noting what is a major lesson that other countries can learn from Costa Rica in terms of how they were able to attract and convince Intel to invest in their country. Your answer should include two elements:
- Element 1: A major lesson that other countries can learn from Costa Rica in terms of how they can attract a company like Intel.
- Element 2: References from below (include page numbers) with evidence/data/arguments that helped you arrive at your opinion
Whale in a Swimming Pool
That's how Intel Corp. once thought it would seem if it built a $300 million computer chip plant in tiny Costa Rica. And therein lies a tale...
By Debora Spar Harvard Business School, USA
On November 13, 1996, the US computer industry giant Intel Corp. announced plans to construct a $300 million semiconductor assembly and test plant (ATP) in Costa Rica, a plant that was to have up to 2,000 employees and within two to three years be fully manned and staffed by Costa Ricans. The announcement came as a triumph to Costa Rican authorities, and also aroused considerable interest in the broader foreign investment community.
With annual revenues of more than $25 billion, Intel is one of the world's largest and most profitable corporations. Costa Rica, meanwhile, is a tiny country, with a population of 3.5 million and only limited development in electronics and other high technology sectors. So why, and how, did Intel choose Costa Rica?
Costa Rica (like many developing countries) explicitly targeted the electronics sector as an area of high potential growth. Like other countries, it also set its sights on increasing flows of foreign investment to the country and created an investment promotion agency to attract and convince potential investors. But whereas other countries have seen industry-specific promotion efforts stagnate or stumble, Costa Rica appears to have succeeded quite brilliantly.
Thus the Intel investment provides a compelling platform to discuss how countries can gradually improve their climate for foreign investment and the design of their investment promotion strategies. Intel's decision to invest in Costa Rica serves as a rich example of how a small country with no domestic market can still lure a world-class, high technology firm.
Decision to Expand In 1971, the Intel Corporation introduced the 4004 chip, the world's first microprocessor. This introduction revolutionized the computer industry and set off the phenomenal growth that both the industry and Intel have enjoyed ever since. In 1995, the worldwide production of computers was valued at $237 billion, up 13.5% from the 1994 total. Semiconductor sales alone were valued at $123 billion and were predicted to continue growing at 20% a year between 1995 and 2000. These growth figures are driven by the diverse and increasingly pervasive applications of microprocessors, the advanced and complex semiconductors that form the core of Intel's business.
Essentially, microprocessors are the "brains" that drive most electronic computing functions. They are integral to the function of mainframe computers, personal computers, wireless communications, and a host of consumer electronics products. Because of their growing applicability, microprocessors are generally regarded as a key component of industrial growth —
and a symbol of economic and industrial prowess. Thus, although firms from the US and Japan continue to dominate the industry, the governments of Korea, Taiwan (China), and Singapore
have all actively supported the development of their home-grown semiconductor companies, and China, Ireland, Israel, and Malaysia have eagerly pursued investment from leading foreign firms. For developing countries, the semiconductor industry carries perhaps the ultimate promise of positive externalities — of jobs and technological innovation and the kind of long- term returns that have made Intel one of the most profitable companies in the world. Intel's distinctive position in the semiconductor industry has led it to pioneer an equally distinctive strategy for operations and investment. Essentially, the strategy is driven by cutting-edge technology and blistering speed. Every nine months or so, Intel builds a new plant. Nearly all of these plants are constructed to meet future, rather than existing, demand. As CEO Craig Barrett acknowledges, "We build factories two years in advance of needing them, before we have the products to run in them, and before we know the industry's going to grow."
Such optimism is rational, even required, in the fast-paced semiconductor market. For this is an industry where producers reap the bulk of their profits early, usually in the first six months following a product's introduction. During that time, manufacturers can charge up to $1,000 per chip. After six months, however, lower-cost imitations tend to exert significant downward pressure on prices, customarily pushing them toward around $200. For Intel, this basic cycle implies a constant need to innovate, and to ramp up production capacity as quickly as possible for each new generation of processor.
This logic of expansion has also led Intel to develop an impressive string of overseas facilities. By the time it began contemplating what would become an investment in Costa Rica, the company already had wafer fabrication plants in Ireland and Israel, and assembly and test plants in Malaysia, China, and the Philippines.
Early in 1996, executives at Intel decided to research sites for a new assembly and test plant. They convened a team of functional experts, composed primarily (though not entirely) of people who had significant experience with site selection.
Intel executives had already determined the precise contours of the planned investment. It was to be a plant of 400,000 square feet, employing up to 2,000 people to assemble and test the latest Pentium microprocessors. This type of plant, known as an ATP, is the second type of plant that constitutes Intel's manufacturing base. The first type, a fabrication plant (or fab) is where the heart of the microprocessor is produced. Essentially, fabs take thin layers of silicon, known as wafers, and use a highly advanced process of photo-lithography to etch layers of electronic circuitry on each eight-inch wafer. This process requires an ultraclean environment and staggering levels of capital and technical expertise. Once the fab process has been completed, the wafers are sent to the ATP. There, the wafers are thinned to reduce internal stress, then cut into anywhere from 300 to 500 individual chips, or integrated circuits. The chips are then mounted onto a lead frame and attached to thin gold wires that will eventually connect them with the other elements of the computer. In the final stage of the manufacturing process, the chips are encapsulated in either ceramic or plastic packaging and subjected to a rigorous series of tests.
To run the new ATP as cost effectively as possible, Intel knew it had to find a low-cost yet highly trainable workforce. It would also have to find a spot where highly qualified engineers were available, and where employee turnover could reasonably be kept to a minimum.
Intro to Intel Within the semiconductor industry, Intel's position is legendary. With 85% of global microprocessor sales, it is by far the world's leading producer. In 1997, Intel recorded revenues of $25.1 billion and net income of $6.9 billion. The company's average return to investors has been roughly 44% a year; its cutting-edge chips were selling in early 1997 at gross profit margins of nearly 60%.
Even more impressive, though, than Intel's numbers is its undeniable position as the industry's technological and strategic leader. Nearly since its founding in 1968, Intel has been the first to introduce the latest and fastest lines of microprocessors. Driven by cofounder Gordon Moore's well-known "law" that the power of these computer chips doubles every 18 months, Intel has consistently created successive generations of ever more powerful chips. Its competitors, meanwhile, simply follow Intel's engineering lead, imitating new designs (called "architectures" in the trade) without having to invest the massive research and development funds that support each Intel launch. These competitors then compete with Intel almost exclusively on price. And as this rivalry inexorably drives its own prices and margins down, Intel moves on to the next generation of processors.
In recent years, Intel has also worked aggressively on the demand side of its industry. Rather than risking a downturn in demand for its new, more powerful chips, Intel encourages other companies, such as Microsoft, to design software that requires faster processing capabilities. This encouragement creates a unique but profitable dynamic, described by one Intel executive as akin to "a wrestling match and a dance that occur simultaneously." Intel helps create the demand for its product and enables its partners to share in its extraordinary success. In 1996, the company spent $500 million to fund software start-ups and to prod the development of other potential users of new-generation microprocessors. This investment came on top of its own $5 billion expenditure for capital projects and research and development.
— Debora Spar
Costa Rican GDP: $9.0 billion Intel Annual Revenues: $25.1 billion
Underdog In these early days, Costa Rica was not seen as a particularly strong contender. It had gotten on the list, in fact, almost by accident. For two years, investment promotion agency CINDE (la Coalición Costarricense de Iniciativas para el Desarrollo) had been actively targeting and approaching large, US-based electronics firms. In the late 1980s, CINDE had decided to follow a focused strategy of investment promotion, marketing itself to a specific group of potential investors, rather than spreading its fairly limited resources across a hodgepodge of ambiguous leads. For several years, this focus had been textiles, but as Costa Rican wage levels rose and competition from lower-wage emerging markets mounted, CINDE abruptly shifted out of textiles
and concentrated instead on electronics. With a high level of technical knowledge in the country, relatively low labor costs (for this industry), and an abundance of bilingual workers, Costa Rica's advantages seemed to mesh well with the needs of the growing global electronics industry. And so, ever since 1993, CINDE had been assiduously courting Intel and the other "big fish" of the electronics industry.
In November 1995, Intel at last responded with interest and invited the current director of CINDE's New York office, Armando Heilbron, to its headquarters in Santa Clara, California. Almost immediately after hearing of Intel's interest, Enrique Egloff, CINDE's CEO in Costa Rica, assigned three investment officers exclusively to the project. It was several months after the first meeting in Santa Clara, and after CINDE staff sent Intel a detailed and extensive information package that Costa Rica made it onto Intel's long list of possible investment sites along with Argentina, Brazil, Chile, China, India, Indonesia, Korea, Mexico, Puerto Rico, Singapore, Taiwan, and Thailand.
The Intel team began with basic desk research, looking for obvious reasons to exclude countries from the list.
To be considered a serious contender, the country in question had to have positive economic conditions, an established and reliable political system, and a relatively transparent operating and legal environment. It also needed a sufficient supply of professional and technical operators and a nonunion work environment.
Countries also had to present a workable financial situation for Intel. This was driven in large part by the cost of labor and overhead, taxation rates, tariffs, customs fees, and the ease of capital repatriation. Because all the plant's products were intended for export, tariffs and customs fees were particularly important.
Given the time pressures under which Intel generally operates, the country also had to ensure that products coming from its plants could move efficiently from the plant to an international departure point and then expeditiously through customs and any other export procedures. Before investing in any country, Intel had to be assured of receiving all necessary permits within four to six months. Any delay in the permitting process could seriously compromise the project's very tight schedule.
On the basis of these rough criteria, team members slowly winnowed down their list. By the end of this process, the original list of 12 had been narrowed to four: Brazil, Chile, Costa Rica, and Mexico.
During the spring of 1996, Intel's site selection team went directly to visit the countries under consideration, seeking to glean an "insider's" perspective on business conditions and state practices within each potential host. Team members conducted lengthy interviews with consulting firms, government officials and other US corporate executives. They met with accounting and law firms, ran in-depth analyses of key factors such as work force capability and waste water infrastructure, and tried to solicit the opinions and experiences of other foreign investors.
Costa Rica received its first visit in April. The two-day tour began with an overview presentation by CINDE, which also coordinated many of the subsequent meetings. The Intel team then spoke with representatives from Citibank to inquire about the adequacy of the country's financial infrastructure, and with executives from international accounting firms KPMG Peat Marwick, Price Waterhouse, and Ernst and Young to examine the reliability and transparency of Costa Rica's legal and financial institutions. Closed-door meetings with enthusiastic executives of DSC Communications, the largest US electronics company with operations in Costa Rica, as well as with several other manufacturers, apparently gave the Intel representatives additional confidence in the country's general business climate and its capacity to process and convert weekly flows of several million US dollars.
Presidential Push During this preliminary visit, CINDE had also arranged for the site selection team to meet with José Rossi, Costa Rica's minister of foreign trade, as well as with José María Figueres, the country's president. Since November, Figueres had been kept informed of CINDE's interaction with Intel and had continually expressed interest in helping with the project. The young, Harvard-educated Figueres, in the middle of his presidential term, was keenly aware of the potential impact Intel could have in helping to lead the country's growth. He took a strong personal interest in relations with Intel, and was a critical element in Costa Rica's eventual success. During the initial visit, he spent two and a half hours with the Intel representatives, during which he pledged to "do whatever was necessary" to make Costa Rica competitive in the race. He was engaged, enthusiastic, and energetic in his pitch and responded directly to Intel's concerns. When the team expressed doubts about the quality of the work force and the adequacy of technically trained graduates in the country, Figueres suggested that the government could create an enhanced training program to meet Intel's needs.
In what would become a critical move, Figueres appointed Rossi to manage the Intel project for the Costa Rican government. CINDE would remain a key contact for Intel and a facilitator for any sub-sequent meetings or negotiations, but Rossi, a high-ranking and well-respected government official, would serve as the central point of coordination within the Costa Rican government. Rossi impressed the visitors, who appreciated his active involvement and understanding of the firm's business needs. Rossi himself had been a businessman, running a sizable family-owned holding company before he joined the Figueres administration. He recognized the importance of speed and the need for an expedited process and clear, consistent communication from the government.
Areas of Concern Over the next several months, Intel representatives visited Costa Rica every week. There were different representatives each time, and different concerns to address, but CINDE remained the lead agency throughout the visit and negotiation period, ushering Intel executives around the country and working to find, or create, mutually acceptable solutions to each of their concerns. By July, the Costa Ricans knew that Intel's short list was down to only two contenders, themselves and Mexico.
This last stage of the process was in many ways the most intense, both for Intel and for CINDE. During this time, several major concerns emerged. The first was simply Costa Rica's size,
especially compared to a behemoth such as Intel. Company representatives worried that Intel would overwhelm the country and would demand an unsustainable fraction of Costa Rica's total resources. Robert Perlman, Intel's director of finance, expressed concern that Intel's investing in Costa Rica was akin to "putting a whale in a swimming pool." Intel still saw other problems in Costa Rica: its physical and educational infrastructures were inadequate, and the financial terms of the proposed investment were less favorable than those being offered elsewhere. To close the deal, Costa Rica had to address these problems.
Infrastructure Intel's problems with Costa Rica's infrastructure lay primarily in the transportation sector. To meet the demands of its market, Intel's new facility would be designed to use inputs from any fabrication plant in the world and to send products, by air, to any customer. While Costa Rica's location was very attractive in this regard, with access to California or Texas in under three hours, Intel was worried about the frequency of flights and the capacity of San José's airport.
Although the overall size of Intel's shipments would be small (about 18 tons/week), they needed to be divided into many batches and sent on several different flights. This was due to the insurance requirements that surrounded the transport of a cargo that was, quite literally, worth its weight in gold. Thus the volume capacity of flights into and out of San José was not as important to Intel as the number of flights and their destinations. And here Costa Rica fell short. While there were a number of daily direct flights to Los Angeles, Houston and Miami, there was only limited direct access to Europe and no direct access to the Far East.
Roads were also a source of some concern. The Intel site will be located close to the country's main international airport, along the highway linking the center of San José with the airport. The problem did not lie with the quality of the highway, which is more than adequate; rather the access to the main road from the planned location was indirect and convoluted.
As it became clear that transportation issues could stymie the investment, CINDE and President Figueres urged the Ministry of Transportation to find some way of accommodating Intel's needs. And they did. After gathering information and meeting with Intel, the Ministry agreed to grant more licenses to foreign carriers if it were necessary to ensure an adequate number of flights. It also apparently accelerated plans for a new cargo terminal, slated to open in May 1997. On roads, the two sides struck an easy compromise. Intel donated some of its own prime land to create an access road for its facility, while the Ministry agreed to improve access to the highway by constructing an overpass ramp, and to coordinate traffic patterns and public transportation schedules to make sure suppliers and employees had easy access to the facility.
Energy proved more troublesome. Even though Intel's projected share of total energy consumption fell from original estimates of 30% to a much more manageable 5%, the plant still demanded its own substation. And this substation had to be built and funded by ICE, Costa Rica's state-owned electric company. Initially, ICE estimated that construction on the substation could not even begin for a year and a half — clearly an unsustainable schedule for Intel. To expedite matters, Intel eventually agreed to cede all of the required land to ICE and to provide funding (through an undisclosed loan arrangement) for the additional power lines and substation. It also agreed to fund a second substation to serve a neighboring industrial park.
Meanwhile, Intel negotiated heavily with the minister of energy to secure better rates for the Intel plant. Existing rate structures included only two rates, residential and industrial, leaving Intel with prices of around $0.07 to $0.09 per kilowatt hour (kWh). For an energy-intensive facility such as an ATP, the difference between this rate and the $0.02/kWh that Mexico offered put Costa Rica at a significant disadvantage. So the Ministry worked with ICE and its National Regulatory Authority to develop a two-tier industrial rate structure, giving larger users like Intel more favorable pricing. Under the new agreement, still pending final approval, the cost of power will drop to an average of $0.05/kWh for any users consuming over 12 megawatts.
Financial Incentives Costa Rica's standard investment incentives and tax policies under the free zone system are extremely attractive, offering investors such as Intel a full exemption from taxes on profits for the first eight years of operation, and a 50% exemption for the next four years. However, at the time of the negotiations, Costa Rica still levied a 1% tax on the total assets of free zone developers. Intel was extremely unhappy about this tax, since its total cost to the company would be substantial: roughly $3 million for its proposed $300 million facility. Complicating matters was the fact that the tax law had temporarily lapsed, leaving some uncertainty as to its application, especially since Intel intended to build on a site not previously developed as a free zone.
The government decided to seek an interpretation by the attorney general, who, bearing in mind the country's new objectives of attracting high technology, capital-intensive industries, concluded that the tax did not apply to companies under free zone status.
Education Intel's most pressing concern, and Costa Rica's most interesting concessions, came in the area of education. Although education levels in Costa Rica were already substantially above the norm for developing nations, the country did not have the education infrastructure to support Intel's personnel needs — for example, to train the 800 technicians the plant would require. Both Intel and Costa Rica knew that this gap had the potential to be a deal breaker.
Well aware of this threat, CINDE and the government quickly launched a program to ameliorate Intel's concerns. A team consisting of Intel human resources staff, CINDE staff, the minister of education, the minister of science and technology, and officials from national institutions of higher education was formed to identify the gaps in Costa Rica's educational system and to submit guidelines for improvement.
The team spent considerable time matching the detailed personnel requirements from Intel against the curricula of the country's technical high schools and advanced training programs. In addition, a group of four professors from the Costa Rican Institute of Technology (ITCR) and two teachers from local technical high schools made a six-week trip to Intel facilities in Arizona, New Mexico, and California. By speaking at great length to operators and technicians at the plants, they sought to understand precisely the education and skills required to support an Intel work force.
Following this review, the team submitted a detailed and extensive list of recommendations to the Ministry of Education:
ITCR would make this program available to either technical high school or academic high school graduates to update their technical skills and physics/chemistry competency on an as-needed basis. Graduates of the certificate program and qualified graduates of technical high schools would also be able to enter an additional one-year program designed jointly by Intel and ITCR. Initially, these programs would focus on semiconductor manufacturing, although they could be extended over time to include other career tracks as well.
ITCR would provide intensive language training courses in Spanish for expatriates from Manila and the US, and English training to the first group of 50 technicians hired in Costa Rica. This is independent of the degree program and will be done directly on contract with Intel. Urged by CINDE and President Figueres, the Ministry of Education approved all of the team's recommendations. ITCR began almost immediately to implement the new curriculum.
In the end, Mexico, which reportedly was the front runner, was compromised by its currency crisis and a system of mandatory union rules. For Intel, union-free in all its manufacturing facilities, the presence of a Mexican union might have created a significant culture clash within the plant and even within the company. Mexican authorities offered to make an exception to the rules for Intel, as it had for other major multinational investors. Yet their very offer made Intel wary of the way business policy was formulated in Mexico and ironically helped, in part, to eliminate Mexico from consideration.
Intel's Choice And thus, on November 13, 1996, Intel announced its decision to build its next assembly and test plant in Costa Rica. As is customary with Intel, the announcement was conditional: it declared that the project would be located in the chosen country only if the government delivered on the provisions of an agreed-upon contract. In Costa Rica's case, these provisions included the completion of Intel's registration in an authorized free trade zone, the awarding of a series of environmental and construction permits, and a government commitment to enhance technical curricula and training facilities at several institutions for students studying electronics.
For the next several months, various ministries, CINDE, and Intel worked to prepare the relevant documentation and finalize the arrangements of their deal. In April of 1997, construction on the new ATP began, and by early 1998 it was in operation, projecting $700 million in exports by the end of the year.
The title of the article (Whale in a swimming pool) refers to Costa Rica as the swimming pool and Intel as the Whale. This is understandable because the annual earnings of Intel is almost three times than the Costa Rican GDP (Costa Rican GDP: $9.0 billion Intel Annual Revenues: $25.1 billion).
Based on personal reading and re-reading, the reasons why Costa Rica was favored by Intel more than any other country is due to these 13 institutionalized policies:
1.) Availability of Trainable workforce,
2.) A functional Investment Promotion Agency,
3.) Stable economic and political system free from labor problems,
4.) Easy transit and access of Raw materials,
5.) Shortened steps and procedure for the grant of Business Application and permits,
6.) Strong, reliable and transparent Financial Infrastructure,
7.) More foreign airline companies to ferry in Intel's raw materials,
8.) Cheaper power cost preferably $0.02/kWh,
9.) Attractive tax incentives,
10.) Industry-Academe linkage (Academe meeting the needs of the industry),
11.) Availability of Multilingual technicians,
12.) Union-free operation in all Intel intermediaries, and
13.) Faithful contract agreement with the government (Government must deliver the goods as promised.).
The absence of any of these 13 items will disqualify a country from being chosen.
1.) Availability of Trainable workforce
To run the new ATP as cost effectively as possible, Intel knew it had to find a low-cost yet highly trainable workforce. It would also have to find a spot where highly qualified engineers were available, and where employee turnover could reasonably be kept to a minimum. (See Lines 74-76)
2.) A functional Investment Promotion Agency
For two years, investment promotion agency CINDE (la Coalición Costarricense de Iniciativas para el Desarrollo) had been actively targeting and approaching large, US-based electronics firms. In the ...
This discussion specified why Costa Rica was favored by Intel as their investment hub.