Back in 1987, Cadence Design Systems became the first electronic design automation (EDA) vendor to open an R&D center in India, where it now has 300 developers. Since 2000 it has opened two R&D centers in China and one in Russia, employing a total of 100 developers. Cadence says that none of the centers replace existing U.S. jobs. As business demands, it will increase R&D jobs both onshore and off, but more jobs will be created offshore.

Cadence isn't alone. Electronics companies have done R&D offshore for years and are doing so more and more, for many reasons. Doing R&D in India or China reduces costs by at least one-third, but that is only part of the story, executives argue. It gives them access to skilled workers at a time when the United States is producing fewer engineers, creates 24-hour development cycles and places R&D resources close to current or future markets.

Those are the rewards. The offshore-R&D strategy also has risks. Haphazard execution offshore can lead to lower productivity and put intellectual property at risk. Back at home, engineers' morale suffers every time new offshore jobs are created while U.S. unemployment remains high. There's also a growing public outcry over the loss of U.S. jobs due to the offshore trend.

Given that R&D has been going offshore for years, why the public furor now? Because R&D is caught in a perfect storm created by many pressure systems: the broader trend of all kinds of jobs moving offshore, the jobless recovery, record unemployment for electronics engineers, workers suspicious of management's intentions, election year politics and media coverage of all of the above.

Despite headlines suggesting otherwise, executives argue, the majority of R&D will stay in the U.S. for the foreseeable future. They see offshore R&D as a way to stretch budgets, improve overall competitiveness and preserve—not eliminate—jobs in the U.S. "If we take more relatively mature programs and work on them offshore, then we have money to make R&D investments here," says Jaswinder Ahuja, Cadence's corporate vice president of central operations.

Not everyone believes that rationale. Bill Reed, president of the tiny but feisty American Engineering Association (1,000 members, half in electronics), argues that shipping R&D offshore is a slippery slope. "We're losing our high tech, and there's nothing to replace it." Critics such as Reed lump R&D in with the broader trend of manufacturing, call center, IT service and other jobs moving offshore solely to cut costs.

However, company executives argue, R&D is different. "R&D is not a pure cost-reduction play," says Ahuja. So far, that argument has been hard to sell.

Remember the good old days when displaced workers found safer and more rewarding jobs in the service and knowledge sectors after their factory jobs moved offshore? Some executives naively believe that the current public furor will subside once the economy produces a spate of new jobs the way it used to.

A few executives, including Intel CEO Craig Barrett, believe that this offshore wave is fundamentally different, a strategic inflection point full of political uncertainty. Here's the gist of Barrett's case, as he explained it to the San Jose Mercury News in December.

"China is making an effort to show that its laws are in compliance with treaties. But we still see a flood of unauthorized merchandise coming from China."—Alan Kindred, Buchalter Nemer Fields & Younger

The three most prevalent popular nations for offshore R&D—India, China and Russia—have strong educational systems and three billion people. If only 10 percent are highly educated, that's 300 million—more than the entire U.S. population. In earlier offshore waves, the U.S. refocused its more highly educated workforce farther out on the knowledge curve. That's no longer possible, given the level educational playing field and the numbers. The worldwide competition for knowledge jobs—including R&D—will only increase.

Individual companies, Barrett continues, will remain competitive if they take advantage of the offshore workforce. Although he often reiterates Intel's commitment to its 50,000 U.S. workers, he also says that the company has no obligation to suffer financially just to employ Americans. If the U.S. as a nation wants to remain competitive, he argues, it must do things to keep workers on the front of the knowledge curve, such as improving education and funding more basic R&D. That, in turn, will also help business remain competitive.

Barrett thinks this inflection point is widely misunderstood, even in corporation-friendly Washington, D.C., where, he says, officials still believe that the U.S. economy will inevitably produce newer, better jobs. It's not actually inevitable, he argues.

Ravi Aron, assistant professor of business at the Wharton School at the University of Pennsylvania, largely agrees and adds that mass unemployment is possible unless the government bolsters educational opportunities for laid-off workers so even the highly educated can retrain for new fields— although no one knows exactly what those might be.

Opposition to the offshore trend has already triggered responses that could harm business, including a defeated bill in Congress that would have required China to float its currency, most likely raising the cost of doing business there. Wharton's Aron is bothered most by the ramifications of government intervention: Public outcry may spawn legislation that will hurt the electronics industry. "My fear is that the policy makers are going to make the wrong decisions," he says.

The offshore trend is an issue in the presidential campaign, with Democrats citing job losses, declining competitiveness and rising trade deficits. Some Republican congressmen have voiced similar concerns—this is not a strictly partisan issue.

Politicians are responding to the growing outcry from grassroots organizations such as Mad in the USA and a nascent technology union. Mad, a Waterbury, Conn.-based organization of small manufacturers, is in a loose coalition of similar groups that represent 10,000 small and medium-size businesses with 500,000 employees, says Mad cofounder Fred Tedesco, CEO of Pa-Ted Spring Co., a spring maker in Bristol, Conn. "If we can't do the R&D and create new products here, we're going to be second-rate." Tedesco was spurred to action, in part, because he learned that more than half of $5.7 billion in defense spending awarded to Connecticut companies was subcontracted offshore.

"If we take more relatively mature programs and work on them offshore, then we have money to make R&D investments here."—Jaswinder Ahuja, Cadence

"What's going on is at the expense of the working people who have built this industry and created the profits," says Marcus Courtney, president of the Washington Alliance of Technology Workers (WashTech), a Seattle local of the Communication Workers of America and the first high-tech union. "White-collar workers had a sense that they would always be taken care of. Now, I don't know how they'll be able to retain jobs, good wages and power without organizing."

Faced with growing criticism, companies are spinning the R&D story differently from the way the press portrays the broad offshore trend. Chip companies argue they are still firmly planted in the U.S. "Seventy-seven percent of capacity is here, even though 80 percent of the world chip market is outside the U.S.," says Sharon Hampton, a spokesperson for Texas Instruments, which opened the chip industry's first Indian software development center in 1987 and has R&D sites worldwide.

Most of Intel's R&D remains in the U.S., where it has invested more than $24 billion in R&D in the past three years, says spokesperson Chuck Mulloy. Most new R&D growth will occur outside the U.S., however, he admits. "Over the next three years, we expect to double the number in India," where it has 1,000 engineers. Intel set up its first offshore center—in Israel—30 years ago and does R&D in many places, including China and Russia.

Some industry executives are also voicing policy alternatives. The Computer Systems Policy Project (CSPP), led by Barrett, actively urges government not to adopt any laws that harm business and to invest more in education, basic R&D and infrastructure, which, it believes, will help create U.S. jobs. (For more on the CSPP, see "The Wrong Way to Innovate," page 15.)

The Electronic Industries Alliance (EIA) is also involved. "One of the big free-trade arguments used to be that a rising tide lifts all boats," says Storme Street, senior manager of government relations. "I don't think those arguments are good enough anymore. Legislators are hearing about highly skilled engineers who are out of work. They need to offer new solutions to these people." When legislators seek solutions, the EIA wants to be heard. So it is gathering input from its 2,500 member companies to produce a set of proposals it expects to publish in the spring. Street says the formation of WashTech last summer was a wakeup call. "That made us realize how [engineers] are reacting" to the loss of jobs.

Is unionization a real possibility? With fewer than 1,000 members between WashTech and its California counterpart, Tech Unite, it is not a likelihood yet. Not nearly as much as the morale problem among U.S. engineers.

Morale is low because engineers see companies creating R&D jobs offshore while there's record unemployment onshore. Unemployment among electronics engineers is higher than in the general population and more than twice that of other professionals (see the chart "Some Comparative Unemployment Rates," below). The figures seem to argue against one reason executives give for the increase in offshore R&D: There aren't enough engineers in the U.S.

"The engineer shortage is a red herring. We're at a historic high in unemployment for electrical engineers," says Ron Hira, assistant professor of public policy at Rochester Institute of Technology and chairperson of the R&D policy committee of the Institute of Electrical and Electronics Engineers-USA, which has more than 235,000 members. Hira knows many laid-off, midcareer engineers with doctoral degrees who can't find work.

It is not clear how many jobs have gone offshore. "We don't have good numbers," Hira admits. In 2002 Forrester Research projected that 3.3 million U.S. service jobs would move overseas in the next 15 years (see the chart "How Many Jobs Will Move Offshore?" below). Actual numbers don't exist, and Congress has asked the General Accounting Office to figure out how many jobs have moved offshore.

Certain pronouncements by executives don't help morale at all. In January, Carly Fiorina, CEO of Hewlett-Packard, stated in Washington, D.C., that "there is no job that is America's God-given right anymore." Fiorina, a member of the CSPP, was trying to make the same point Barrett makes about job competition being global. That's not what Silicon Valley's unemployed heard. Her remarks drew angry letters to local newspapers, complaining that Fiorina, who made more than $10 million in 2002, had hit a raw nerve.

Balancing candor with sensitivity to workers isn't easy. "In today's context, just managing expectations and morale here in the U.S. is a big challenge," says Ahuja. "There is so much buzz in the media right now about offshoring of jobs that it creates a sense of gloom and doom around any global effort."

"We're at a historic high in unemployment for electrical engineers."—Ron Hira, IEEE-USA

The offshore issue often comes up at Cadence, he says. "It is not unusual for engineers to ask if, five to 10 years from now, there will be any R&D left in the U.S." He says Cadence's leading-edge R&D won't leave the U.S. in the near future, even as the company increases R&D offshore. About 75 percent of its R&D workforce is in the U.S. That proportion, but not the number of U.S. jobs, will drop over time, he says.

Workers need reassurance. "We need a constant dialogue with our engineers," he says. Cadence has stepped up efforts to communicate, counsel and train staff members to keep productivity high and to offer them education that will keep their skills at the leading edge. "I expect we'll be spending more on training," Ahuja says.

Ahuja is the only executive among two dozen interviewed to candidly discuss the morale issue. If companies continue to ignore the problem, they may find their engineers more inclined to join a union than they currently are.

Politics and worker backlash are the risks at home. On-site risks include unrealistic expectations about savings and how quickly offshore R&D can match U.S. productivity.

Experienced companies say that offshore R&D isn't fully productive overnight. "You should be doing this if you want to build a design center for the long term," says Samuel Fuller, vice president of R&D at Analog Devices, which launched its Indian R&D center in 1995. "If you think you can get it up and running in six months, you'll be disappointed."

Under pressure from their boards, most "CEOs don't have six months," says M.R. Rangaswami, managing director of Sandhill Group, a San Francisco consulting firm that spent last year interviewing dozens of software companies doing R&D in India. "A lot of companies are not thinking it through. When you focus only on the cost, you take shortcuts. The results are not optimal, and productivity is not as good." Most companies don't understand how difficult, time-consuming and costly this is, he says. (See the charts "What to Expect when Setting Up New Offshore R&D" and "What to Expect from Ongoing Offshore R&D," below.)

Analog Devices has spent several years building staff offshore. Its India center, with 250 engineers, first worked on portions of a digital signal processing (DSP) chip and only after several years took ownership of a DSP family, Fuller says. Engineering talent was easier to find than managers. A local management hire didn't work out, and Analog Devices had to send a senior executive—an Indian national—to run the operation for the first five years.

"Productivity can be a significant challenge when the offshore engineers are managed by U.S. personnel who don't understand cultural differences," says Debashish Sinha, managing director at NeoIT, a consulting firm in San Ramon, Calif. Chinese engineers agree to deadlines but often don't know how to use resources to meet them, he says. China is further complicated because English is not widely spoken, as it is in India. In China, companies need to hire local managers fluent in English.

Companies also suffer if "seat of the pants" describes their R&D processes. They need to put their own house in order before they set up offshore R&D, Rangaswami warns.

"The R&D design process needs to be well documented so the work can be partitioned," says Anita Manwani, vice president and general manager for global sourcing at Agilent Technologies, who opened HP's software center in India in 1989. "Work needs to be partitioned the same way, whether three engineers are side by side in cubicles or on three continents." In 2001 Agilent launched its Indian site, where it has 60 design engineers; it also does R&D in China and elsewhere.

Indian and Chinese engineers are well trained but lack experience. Because its foundry partners are in China and Taiwan, SST, a Sunnyvale, Calif.-based fabless chip company, does process R&D in those countries, says CEO Bing Yeh. He says that Chinese engineers are still relatively new to the industry and that it will be years before they are ready to tackle advanced IC designs.

Howard Yang, vice president and general manager of a 100-engineer center in Shanghai for Integrated Device Technology (IDT), which designs chips for telecom systems, had to teach new hires CMOS design when he launched the center in 1997. "It was work by day, teach by night."

Offshore R&D centers often require U.S. managers to spend more time on the phone and at the site than they expected, says Rangaswami. As a result, cost savings are not as much as anticipated, although significant—certainly not those achieved by offshore call centers. "Raw labor costs for junior engineers can be 60 to 90 percent cheaper," he says (see the chart "Average Annual Salary Requirements for an Engineer," below). With infrastructure, travel, management time and the need to bring local hires to the U.S. periodically, a 40 percent savings is more realistic. That's only in the beginning, though: Indian engineers expect annual increases of 15 percent or more and are known to jump to competitors that also have R&D in India if they don't get them.

Analog Devices' Fuller says that the biggest salary savings are for junior engineers and that then the gap narrows all the way up the ladder. "By the time you get to vice presidents, there's not much difference." The message is clear: Don't do offshore R&D just to save money.

IP can be at risk wherever it is developed, but more so offshore. "Companies are not as concerned as they should be," says Rangaswami. "The smaller ones are especially naive." In his study of more than 50 software companies doing development in India, none admitted having had IP stolen, but he knows it happens.

India has a British court system and IP laws, but justice is slow, says Alan Kindred, an attorney with Buchalter Nemer Fields & Younger. "Generally they are too slow to deal with serious violations of IP rights." China recently joined the World Trade Organization and adopted Western-style IP laws but has little experience in enforcing them, Kindred says. China "is making an effort to show that its laws are in compliance with treaties. But we still see a flood of unauthorized merchandise coming from China."

Asians don't have the same respect for IP as Americans and Europeans, says Yeh. "When we did work in Taiwan fabs, the engineers seemed to discuss our process technology with each other, no matter what they were working on. Taiwan has improved, but China has a ways to go."

Savvy companies use all the measures they use in the U.S. to protect IP, and then some. Among the extras: prohibiting floppy disks or CDs in the workplace and providing no drives for them, prohibiting laptops in the workplace, partitioning databases and keeping the development network disconnected from the Internet.

Yang tells a story that suggests hope for China. IDT has strict measures in place, but last year one engineer found a way to copy a design file to show a potential new employer—an IDT competitor. IDT's e-mail monitoring software picked up that he had sent the file. "We immediately called the police," says Yang. "Ten to 20 cops quickly came over and took the guy away. Several police went to the other company and searched one of its computers, where they found the file and deleted it." The hapless fellow was fired, held by the police for a week and released on one-year probation.

Most companies have devised strategies for managing productivity offshore and protecting their IP. Some recognize sagging morale at home and deal with it. Few, however, grasp that the offshore R&D trend represents a strategic inflection point fraught with political peril.

The industry needs to fight harder to make its views known. Barrett's big concern is that this inflection point is not widely understood even within the electronics industry. He says that more executives must internalize this fact. Then, he says, "What is necessary is for all of us in the industry to speak with a common voice, and we haven't done that."

If electronics executives don't speak loudly with one voice, they may find politicians passing legislation that protects U.S. jobs but is detrimental to the industry as a whole. That would be the biggest risk of all.

Who's in danger of causing more damage in the offshoring uproar—business, workers, or government? Send your thoughts to feedback@eb-mag.com.

Bill Roberts, a Silicon Valley-based EB contributing writer, thinks a move offshore to Tahiti would improve both his morale and his productivity.