Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Software-Related Offshoring1 Alfred Z. Spector Iâll begin with four rather simple observations about soft- In Scenario 1, although offshoring continues, the United ware. Although they are simple, I believe they are important States retains its dominant role in many segments of the soft- to any discussion of software-related offshoring: ware field as a result of the differentiation I mentioned above. In Scenario 3, I present a worrisome picture of accelerating 1. There is a global leveling of opportunity in the soft- migration of software jobs overseas as talent in the United ware field. States dries up, perhaps the result of an expectation-driven 2. The software field is very large, and its subcomponents downward spiral. In Scenario 2, I describe an intermediate include many diverse practices and skills. situation in which many traditional programming-related 3. Terrific opportunities for innovation in software re- jobs migrate, but high-value growth both in the field and main, and demand for software should increase as around its periphery is sufficient to sustain the industry in prices decline and innovation continues. Economists the United States. would say there is both high-price and high-innovation Finally, my primary conclusion is that we must attend elasticity of demand. to the talent available in the U.S. labor pool. It seems 4. As applications of information technology continue to self-evident that unless we have a sufficient number of increase rapidly, they provide increasing opportunities enormously talented individuals, whether U.S.-born or im- and reinforce the centrality of software to science, migrants, who have been given the best training in the world, engineering, and, indeed, society at large. we will gradually drift toward Scenario 3. I also briefly men- tion that we must vigilantly protect the laws and economic To illustrate some of these points, I will continue with a structures that encourage continuing investment in both new brief description of some offshoring activities by IBM, where research and novel businesses that generate the opportunities I was recently vice president of strategy and technology for on which we will depend. the IBM Software Group. Then I will talk about the impli- cations of offshoring for growth in the software field, the MESSAGE 1: WORLDWIDE LEVELING impact of diversity in the software industry on the potential competitiveness of different populations, the importance of To introduce my first observation about the global lev- the software industry to our society, andâfinallyâthat the eling of opportunity, Iâll tell you a personal story. In the United States cannot expect to dominate the software indus- summer of 1973, at the end of my freshman year at Harvard try they way it once did. College, I got a programming job at Harvard. I recall viv- I will then propose three possible scenarios for the future. idly one particular morning, after I had pried myself out of â Alfred A. Spector is a consultant and former vice president of strategy 1â The views expressed herein are the authorâs and not those of institutions and technology of IBM Software Group. with which he is affiliated. 195
196 THE OFFSHORING OF ENGINEERING bed early in the morning and headed off to the then-extant software production. Here are some aspects of the process, Harvard Aiken Computation Laboratory to write and debug although not all of the elements Iâve listed are applicable to code, I found myself the sole user of the Harvard PDP-10 all software production: research time-sharing computer. This was my reward for getting up before 5:00 a.m. â¢ conceptual work as a basis for deciding what can and The PDP-10 ran at roughly 400,000 instructions per sec- should be done ond. It had less than half a megabyte of memory and a few â¢ competitive analysis to determine how to succeed in megabytes of disk storage; we used small magnetic tapes, the market called DEC tapes, for all long-term storage. Even in that â¢ work on requirements as a basis for making a formal early era, the Harvard PDP-10 was networked with a few decision about what a program must do other computers via the ARPANET, the predecessor of the â¢ various perspectives for considering the design of a Internet. In fact, in my work I regularly accessed computers system: at MIT and Carnegie Mellon. At the time, I didnât know how âhe human interface t much the PDP-10 cost. However, I did a little research for âhe security of operation t this presentation, and I believe that it would have cost about âhe robustness of operation in the presence of faults t $2 million in todayâs dollars. â ther factors o In addition to pursuing my debugging work, I remember â¢ development of the high-level design of the major my mind wandering and contemplating my career options. modules and information structure of a program At the time, I was considering going into economics and â¢ the low-level design of individual modules journalism, but I was also thinking about computer science. â¢ coding That early morning I do remember explicitly thinking about â¢ porting to alternative platforms the comparative advantage I had as a student at a U.S. uni- â¢ formal and informal verification versity, capitalized by my very own PDP-10 (at least at 5:00 â¢ testing of components, modules, and systems a.m.), and thinking about all the folks in Europe who had â¢ evaluation and tuning of performance minimal computational access (note that Europeans used to â¢ intellectual property protection and licensing do much less hands-on computer science because of this). â¢ development of documentation/information and At the time, I never even considered India and China as hav- national language support ing any software capabilities. I believe my unquestionable â¢ packaging and delivery comparative advantages impacted my career choice. â¢ project management Almost 35 years later, the contrast is clear. Modern com- putation and networking are four to five orders of magnitude Undoubtedly, many more activities could be added to better, cheaper, and more ubiquitous than they were then. this list. And most necessary information is on the Web. Take just one The second cut across the field, the application domain of example: MITâs plan to put most of its instructional materials software, influences development processes in many ways. on the Worldwide Web. Even machine translation is making Systems software (e.g., operating systems, database manage- some progress making information available in multiple ment systems, server infrastructure, middleware) that run languages. Thus leveling of opportunity is undeniable. continuously have different requirements, such as robustness This leads me back to my first observation. A U.S. student and scalability, than tools that are executed and re-executed going into the field of computer science today does not have periodically. Packaged applications that are sold to numer- as great a comparative advantage as a student even 10 years ous customers have different requirements (e.g., significant ago. This is not a reason to avoid computer science and soft- expertise requirements in the huge number of potential ap- ware, but it is important to recognize that the U.S. advantage plication domains) from programming tools, although they has decreased. must still be of use to a variety of customers within a par- ticular industry or problem domain. Custom applications for one or a few uses or customers may be considerably easier MESSAGE 2: A VERY DIVERSE FIELD OF ENDEAVOR to develop because they require less generality, and there The second point I want to make is that âsoftwareâ or âin- is, therefore, less of the combinatorial explosion that makes formation technologyâ is not one large, coherent, aggregated packaged software so expensive. These different types of profession, but is instead a very diverse field. This is partly applications also require significantly different production because it is a very big fieldâmore than a trillion dollars methodologies. are spent on software worldwide (in aggregate). To illustrate Even in each of these application areas, there are many this diversity we can look at four different âcutsâ across the approaches to developing software: variety of activities in and around software (Figure 1). The first cut considers software from the vantage point of â¢ The traditional waterfall method is a common baseline,
SOFTWARE-RELATED OFFSHORING 197 Domain 1 Domain 2 Formal Methods, including Verification Domain 1 Domain 2 Development Methodology FIGURE 1â The diversity of software activities. Spector Figure 1 Landscape view but as requirements and designs are refined, every step software-development organizations is not like designing a might need to be revisited requiring cycles in develop- semiconductor fab, for which one can create a design that ment processes. Because this is a fixed image (jpeg), wecloned in whatever locations have the most favorable can be cant exactly match the type face. â¢ Interest in modular assembly (e.g., web service-based and tried to match the original Software development is too We have covered the old type cost or regulatory structures. development or previously object-oriented tech- variable for that. niques), with its greatly reduced emphasis on new Finally, the last cut attempts to capture the important coding, is increasing. The newest incarnations are interactions with the world around software including other âmash ups,â connected groupings of reusable com- items that go into the software life cycle. For example, one ponents that provide a new function, often intended cannot undertake the automation of a medical procedure for a modest-sized audience. Simplicity of assembly without understanding the impacts of failure, FDA require- is the focus, and success is based on the existence of ments for proof of safety and efficacy, and much more. a massive, society-wide capital plant of increasingly Automation strategies, the creation of business models for modular components, such as maps, calendars, group supporting software, an understanding of the management bulletin boards and editors, etc. of holistic systems in which software will operate, and an â¢ Open-source techniques have been remarkably effec- increasing focus on risk and compliance management must tive for creating good software. To the amazement all be considered part and parcel of the software field. of some, volunteer groups in modest organizational I suspect that with more time and thought we could structures, often using many preexisting software fill out and enlarge this multidimensional matrix in each components, are proving adept at developing quality dimension. software. We can conclude, however, that software is a very diverse field. Thus, when we consider offshoring, we must remem- There is no agreed-upon standard methodology for creat- berâand this is my second messageâthat there is great ing software (and there may never be); so software devel- variability in software objectives, job types, and practices opment is not amenable to rigid standardization. Creating around the world. Thus, even if a population somewhere
198 THE OFFSHORING OF ENGINEERING becomes very, very good at one aspect of software, the design and engineering costs are the primary reasons we field is so diverse that it is unlikely that population, or any canât implement all of the requirements and make rapid population, will be very, very good at everything. If, when enough progress to meet grand challenges. Apparently, itâs we think âsoftwareâ we think only âcoding,â we miss the big not cost-effective to tackle some of these, or, presumably, picture. rational firms would do so. Clearly, itâs not the marginal cost of production (duplication or transmission) that is stoppingÂ us. MESSAGE 3: UNBOUNDED OPPORTUNITY There are also management and operational costs asso- My third message is that software offers unbounded op- ciated with deploying software. When we software people portunities. I emphatically disagree with those who say that generate new software, we often forget to take into account software opportunities are fading away and that the bloom the life-cycle costs of maintaining and managing it. Manage- is off the rose. Software is a synthetic discipline for creat- ment and operational costs are important factors that limit ing the logic to encode virtually anything! Software is also the impact of software. operational in the sense that it is a constructive synthesis that Finally, a lack of innovation is always an impediment. For generates useful entities that work and produce value. The many years, people think something cannot be done. Then target domain for software is broader and more varied than a breakthrough occurs, and we begin to do it. For example, for other fields of engineering. Software has applications in there was once a general consensus that we could not do âar- all areas of human interest and all human endeavors. tificial intelligence.â Then, all of a sudden, we had a winning I believe software people are ambitious. They (we) feel chess program and Internet-scale search engines. Maybe that those magnificent computational engines, called com- neither is perfect, but both are exceedingly good. If the initial puters, along with their storage, communication, and I/O entry in an application domain is successful, market forces capabilities, are capable of vast, nearly infinite brilliance. stimulate iterative enhancements that generate a long stream First, consider the most traditional space of software: en- of continuing advances. terprise computing. Even the most mundane application sys- Thus my third message is that software is a field that offers tem in any corporation has a backlog as far as you can see for tremendous opportunities for the future. The level of oppor- improvements, endless requirements that have not been met. tunity has not reached its peak. There is a tremendous amount This necessitates a continuing prioritization process. When I still to be done, and given that the major stumbling blocks managed software products at IBM, we always(!) had much are not the lack of opportunity, but limited cost-effective more to do than we could do with available resources. So we resources and talent (innovation), I submit lower prices and are not running out of even the most traditional work. greater innovation will spur more demand for peopleâthat More important, consider the plethora of uses for comput- is, there is high price and innovation elasticity of demand for ing that have been postulated. Very gradually, we are getting software people. to some of these, but we have a huge backlog. I submit that most of the âfantasiesâ about computer applications will MESSAGE 4: THE EXPANDING eventually come true. Whether playing chess, supporting SPHERE OF SOFTWARE autonomous robots, providing universal access to informa- tion, answering questions, or you name it, these and many My fourth and last point is that the field of software is not other uses for computing just keeping coming. just concerned with refining past achievements. The field is Some argue that Mooreâs law will meet its limits in the expanding (Figure 2). near term and slow innovation. Even if this were true, we In November 2004, I gave a talk entitled âResearch on have massive underutilized capacity today. Even if we hit a the Edge of the Expanding Sphereâ at Harvardâs Center brick wall tomorrow, it would not have a great impact. for Research on Computation and Society. The message of Second, although nothing grows to the stars, frequency that talk was that the software field clearly has a core, say, and density scaling are continuing (particularly the lat- inside a sphere. Today, that core includes, for example, the ter), and this growth will continue for a while, for many study of algorithms, compilers, operating systems, distrib- reasons. uted computing, et cetera. However, every year, the sphere Third, given the very low manufacturing cost of silicon gets bigger. devices, we can have as much processing power as we want For example, when I was a Ph.D. student in the 1970s at as long as we are willing to embrace parallelism. The worldâs Stanford, there were six elements to the field of computer fastest computers are already made up of tens of thousands science: algorithms, complexity theory, software, artificial of processing units, and there are no limits to their feasible intelligence, numerical methods, and architecture. Since expansion. Although exploiting parallelism is sometimes then, the field has grown incredibly in two ways: challenging, the challenge itself opens up fascinating opportunities. 1. The density of elements inside the sphere has in- So what are the factors that could limit growth? First, creased. There is more in operating systems today than
SOFTWARE-RELATED OFFSHORING 199 Growth in the sphere (density) Growth in domains of application â¢ Processor architecture and exploitation of parallelism â¢ Art â¢ Distributed systems â¢ Bio- and medical informatics â¢ Graphics â¢ Business process modeling and integration â¢ Information retrieval â¢ Computer-mediated human collaboration â¢ NLP and voice processing and social networking â¢ Networking â¢ Robotics â¢ Numerical methods â¢ Entertainment/gaming â¢ Operating sytems â¢ Sensor networks (e.g., empirical science) â¢ Programming languages â¢ Societal infrastructure â¢ Trust, security, malleability â¢ Transportation and telematics FIGURE 2â The software sphere is simultaneously denser and expanding. there was 30 years. There is a lot more in program- Sam Palmisano, has explained the many reasons IBM has ming. There is a lot more in artificial intelligence, the created an overseas technical presence. These include prox- study of algorithms, and so on. Spector imity to markets, the capability of understanding overseas Figure 2 2. The sphere has also expanded into new domains that markets, the availability of talent, lower costs, and so on. were once unrelated to software. Examples include In the past 10 years, IBMâs presence in India has increased e-commerce, social networking, bio-informatics, dramatically, especially in the last few years. Mr. Palmisano e-voting, and very many more. reported in June 2006 that IBM had 43,000 Indian employees out of a total of 340,000 employees.2 Thus India now has the So there are immense opportunities in what I refer to as second-largest IBM employee population in the world. CS + Xi, for many values of Xi. The Xi could be art, com- A tremendous variety of jobs are filled by Indians. Most putational biology, medical informatics, entertainment and are in IBMâs vast services business and involve custom ap- gaming, sensor networks, and so on. CS + Xi has an impact plication development, systems management, and call-center on all aspects of the economy and on the way we conduct automation. However, about 5 percent of the Indian employ- science as massive growth in these hybrid software-related ees produce packaged software for the circa $15 billion IBM activities continues. software business; this percentage is somewhat low because For example, our ability to measure all sorts of natural of the particular skill/experience requirements of that busi- phenomena with very low-cost sensor networks will continue ness. So, as I mentioned earlier, it is much more difficult to to revolutionize some aspects of engineering, but also of move certain software-related jobs overseas than others. science. Take another example, something in which all of us IBM has a somewhat larger number of software-group have an interest, the relationship of software to health care, employees in China, interestingly enough, which is a little a circa $2 trillion industry in the United States alone. Most bit inconsistent with the prevailing wisdom that India is people believe that many hundreds of billions of dollars are the developing IT powerhouse. Iâm not sure of the reason wasted because of a lack of good information technology for this, but quality English is not as much a requirement support, and that health care suffers as well. This topic, in the development of packaged software products as it is because of its scale and urgent need for creative solutions, in service-related jobs that involve direct interaction with will undoubtedly generate significant incremental demand customers. for software. My list of examples could continue. Based on my experience in the software business, I can explain the factors that influence a decision about where to locate employees: IBM AND GLOBALIZATION As the recent vice president for technology and strategy of the IBM Software Group, I should say a bit about IBMâs re- 2â In early 2007, after this talk was given, the number had risen cent moves toward globalization. IBMâs chairman and CEO, to 53,000.
200 THE OFFSHORING OF ENGINEERING â¢ Talent, experience, and maturity of the teams. Soft- lations have comparative advantages. One can differentiate to ware has sufficiently high margins that talent, quality, gain comparative advantage in many waysâtalent, experi- experience, and maturity can mean more than costs ence, capitalization, location, trust, risk, and so on. Take just per hour. one example, the need for trustworthy systems as software â¢ Organizational capability, including managerial lead- moves into life-critical domains. ership andâimportantlyâtechnical leadership. The I believe the application of software to other fields and lack of leadership tends to be the most difficult impedi- vice versa will be increasingly important to opportunities ment to growing teams in new locales. for differentiated innovation. A situation may require not â¢ Capability in a wide range of software activities, in- just software talent, but also multidisciplinary critical mass cluding interfacing with project management, custom- (Messages 2 and 4). ers, sales teams, and finance teams. These activities are Just because of its centrality in so many fields, computer not ancillary to software; they are a core part of the science and software are important. The ability to lead in IT business. development and IT applications continues to be important â¢ Co-location with a market. This is related to the previ- for our security and our economy (Message 4). ous point but bears repetition. Finally, global leveling means that Americans cannot take â¢ And, very important, lower labor costs. software leadership for granted (Message 1). Let me describe one situation in which IBM moved some SCENARIOS FOR THE FUTURE software development from England to India fairly recently. IBM had a few tens of people doing somewhat repetitive, but To crystallize my, and perhaps your, thinking on the im- still high-skill, high-profile, Java-related work in England. It pact of offshoring, Iâve developed three admittedly overly took about a year-and-a-half to make the transition to India, simplistic scenarios. during which time the English and Indian teams had to work together closely. The move was successful, with much of the Scenario 1 work now being done in India. To the best of my knowledge, the English team was not unhappy because members felt Certain activities, such as testing, integration testing, in- there would be new, more exciting work to replace what ternationalization (to make software ready for use in many they had been doing. That is, the move freed up the talent countries), and coding are much less expensive because of in England to do things that would generate more revenue offshoring. Nevertheless, elasticity of demand is still high, growth and employment. so lots of opportunities remain for talent in the United States. Overall, in the position I held at IBM, my biggest worry Dollars saved by the reduced costs of offshoring of certain was always about leadership. The same need for talented activities are available for higher value activities that encour- leadership was also important in offshore software re- age growth in overall output and employment. U.S. innova- searchâand was a persistent problem throughout the decade tion, employment, and economic contributions increase. or so when I visited IBMâs newer research sites. I think this scenario is not only possible, but is also the most likely to be realized. When I was at IBM, if develop- ment had been more cost effective, more development would IMPLICATIONS have been done, much of it naturally in the United States. I have discussed four observations relating to (1) the To make this scenario even more comforting for the long global leveling of opportunity in software, (2) the great term, certain coding, testing, and design activities would variety of objectives, job types, and practices in software, remain in the United States to ensure that American universi- (3) the high elasticity (price and innovation) of demand, and ties, labs, and corporations retained sufficient skill and train- (4) the interaction and mutual impact of software and com- ing capabilities and to prevent insidious âtechnical hollowing puter science on more and more fields of human endeavor out.â The United States is likely to retain some jobs across (my shorthand for this last point is Xi (CS + Xi)). the spectrum for two reasons: (1) if all members of a team The most important implication is that there are vast op- are co-located, the work goes faster; and (2) overseas cost portunities in software. The technology provides sufficient benefits tend to decrease as workforces there gain skills and benefits to ensure employment for many populations, with no experience (note the significant wage inflation for talented obvious limits. This has not been true in other areas or other Indian software professionals). U.S. industries, where there has been significant degradation. Perhaps, for example, only so much innovation occurred in Scenario 2 the steel industry over the years, and there is only so much demand. I do not believe there are similar limits in software In this scenario, more and more employment in the central (Message 3). sphere moves offshore. But, the software field continues to Second, because of the variability in the field, some popu- change fast enough to generate new subdisciplines or, if you
SOFTWARE-RELATED OFFSHORING 201 will, âsuperdisciplinesâ based on hybrids of software and As I consider what we should do, my obvious conclusion other endeavors. Although the United States may have lost is that we should attend to our future workforce. We must competitiveness in significant aspects of the core of informa- have a creative workforce that has high value compared to tion technology, the countryâs attention has turned to topics others around the world and that can keep us on the leading related to CS+Xi thereby providing continuing opportunities. edge of high-value opportunities. As long as high value is created in these hybrid activities, this Ensuring that we have this workforce will require both is a good outcome. in-depth and interdisciplinary education. I think we donât yet An analogy is to think of our jobs arranged in a pyra- fully understand the requirements and advantages of interdis- mid. As certain jobs at the bottom of the pyramid migrate ciplinary education. With our flexible institutions, the United overseas, the outcome is fine as long as we can move to the States may be better than most at âinterdisciplinarity,â but, top of the pyramid, which of course keeps growing higher. when teaching people about computer software and the fields Scenario 2 is optimistic and possible with a field as open in which we need software applications, interdisciplinary ended as software and CS+Xi. education will require a great deal of careful thought and planning. It is one thing to argue for a better educated populace. It Scenario 3 is something far different to suggest exactly what we should This scenario is pessimistic. When certain activities move do: incentives, curricula, organizational structures, ethnic offshore, our students and funding agencies take this to mean and geographical diversity, and so on. It will take some very that opportunities in software have dried up. As a result, U.S. deep thinking to get these things right. I believe it is time talent dries up, creating a downward spiral. Although there we revisited these topics in far greater depth than we have may still be elasticity of demand for innovation, we no longer so far! have the capability to innovate. Given the centrality of soft- When we think of our future workforce, we must also ware to everything in our lives, this has profound, negative think about immigration. There have been discussions in implications throughout the country. I consider this scenario academic circles about how difficulties in getting into the a risk. United States have reduced the immigrant graduate popula- tion. Although we want native-born Americans to go into science and engineering, we cannot afford to lose the cre- CONCLUSION ative, entrepreneurial immigrants who are integral to our Given that software is central to so much in our lives, talent poolâand who have done so much for our country I believe IT is a crucial fulcrum for American prosper- throughout our history. ity. I think leadership in aspects of software, particularly the Finally, we must retain economic incentives to encourage most innovative aspects, is important for the United States. people to pursue an education, to work hard and be creative, This does not mean we must dominate all elements of and to accomplish great things. We must have the right software, which is fortunate, because we cannot dominate laws to enforce business ethics and honesty, but we must software as completely as we did in the past. I also do not not go overboard in a way the drives the locus of industry think software leadership is incompatible with significant off shore. offshoring. However, we must remain strong in areas of differentiated value.