Intangible Assets in a Knowledge Economy
In introducing the first session, which was charged with defining the role of intangible assets in the economy, workshop chair Kenneth Flamm (University of Texas, Austin) identified the questions he hoped would be answered: What are intangibles? What distinguishes tangibles and intangibles? How should these business inputs be defined and measured? And how do these vary across industries and firms? For example, is a dollar of investment on advertising and brand equity by Pepsodent the same as a dollar spent on advertising and brand equity by Intel? Is a dollar of research and development (R&D) spent on innovation at McDonald’s the same as a dollar of R&D at Microsoft? These are difficult questions, but presenters throughout the session enlightened aspects of each.
TRANSITION FROM THE INDUSTRIAL TO THE KNOWLEDGE ECONOMY
Irving Wladawsky-Berger began with a description of the recent evolution in the economic landscape brought about by the digital technology revolution. The distinguishing feature of this revolution, he noted, is the ever-increasing speed and ever-decreasing cost of computer components. If steam power was the driver of the industrial revolution, there is very little question that information technologies are driving the evolution of the knowledge economy.
Wladawsky-Berger pointed to the advent and proliferation of the Internet as a key point in the transition from the industrial to the knowledge economy. Before the Internet, technology change mainly referred to components, such as microprocessors, and the products that were built with them, like computers. That has begun to change drastically in the past 10-15 years; as a result of
advances related to the Internet, technology is being applied in many different areas. Wladawsky-Berger predicts that, in the knowledge economy, the bulk of the innovation will increasingly be “up the stack,” occurring in applications, services, business practices, and the workings of society generally, as opposed to the hardware, goods-producing side of the economy. The technology and product innovations are now “market-facing” systems that are complicated and have to be designed and managed in a way that reflects the essential differences between machines and products, on one hand, and people and services, on the other. Technology is being applied to help people performing services do them better, which involves complicated emergent applications and systems, rather than well-behaved deterministic machines that do what people tell them to do. Part of this transition involves a world in which, increasingly, much that is needed to make progress is intangible in nature.
Wladawsky-Berger posed the question, “How can we structure a conversation about something as cosmic as the industrial and the knowledge economies and the transition between them?” He proceeded to examine how the knowledge economy differs from the industrial economy on the basis of three concepts that were at the heart of 18th century economist Adam Smith’s writings.
Division of Labor
Adam Smith observed that, when markets are big enough to warrant worrying about efficiencies and productivity, then companies and people can achieve greater productivity through specialization (instead of doing the same custom jobs, as was the practice in agriculture). This involves breaking down a problem into its components, which is all about specializing and improving processes.
The consequences of division of labor are fairly obvious when examining a big industrial project, such as designing and building a new airplane. But how does division of labor apply to the knowledge economy? Does it apply to the knowledge economy and to more intangible types of activities? The answer, according to Wladawsky-Berger, is that, more and more, people are able to apply technology, engineering disciplines, information analyses, and collaborative capabilities to the design, management, and operations of a business—and this is evident in sectors from banking to health care. In fact, he pointed out, in a lot of cases in which things are not working well—such as the health care system—it is because these things have not been done to the extent that is needed.
The same principles of division of labor, specialization of processes, technology, and automation collaboration are still critical elements of economic efficiency, but now they must be applied to very complex systems. Wladawsky-Berger described how IBM has started a major drive toward services sciences, management, and engineering, which examines this question of how to apply technology up the stack to improve production processes of services. In the case of IBM, services account for about 55 percent of revenues and, in the case of
the U.S. economy, about 75 percent of gross domestic product (GDP). Without improvements in productivity and quality that have already occurred in manufacturing, a company like IBM would have difficulty earning profits and a nation like the United States would have trouble raising its standard of living. Indeed, development of intangible assets and innovation generally facilitates division of labor between big and small companies and across borders in the global economy.
The Invisible Hand
Wladawsky-Berger turned next to a second Adam Smith principle—the notion of the invisible hand—wherein each participant in the economy must be concerned only about how to do his or her job well and earn money, and need not explicitly be concerned about the general economic interests of society. By working in one’s own self-interest, chances are that society will benefit. How, asked Wladawsky-Berger, does the concept of the invisible hand translate into the world today? Does it apply in the same way to the globally integrated enterprises and industries that characterize today’s economy, as it did to the baker or the factory owner in Smith’s day? He argued that, if ever there was a need to rely on the invisible hand to help guide economic processes, it is in these times of incredibly complex systems that people are increasingly building and living with. He added that, with the world changing at an increasing rate, events that businesses encounter that used to occur with very low probabilities in a kinder, gentler world have begun to happen with more frequency and with more cataclysmic effects.
Of course, things can be done to help better manage the emerging, increasingly unpredictable, and complex world (one need only look at the current financial crisis). Technology and innovation at all levels once again offer the greatest hope for solutions, and, by Wladawsky-Berger’s reckoning, businesses and markets need to do a far better job using information-based decision support, information-based management, predictive analysis, and predictive simulations to harness this promise. Work is ongoing to do just that, but these are the early stages of an extremely complicated knowledge economy, so there is still a long way to go.
The third principle discussed by Wladawsky-Berger—one that perhaps is less known about Adam Smith, as it derives from his Theory of Moral Sentiments (not The Wealth of Nations)—is the idea that human beings have an innate sense of sympathy for other human beings. This trait helps bind the community together and is what puts a break on people’s worst impulses. While the invisible hand says, “go do whatever you need to,” moral sentiment counterbalances with, “but if you go too far, the community will ostracize you and, by the way, they may decide not to do business with you.”
Along with the first two Smith concepts, how does moral sentiment apply in a global world? It is easier to see how moral sentiment was influential during the 17th century, when people dealt with each other face to face on a village or neighborhood scale, and even later at the town or city levels. But people on Wall Street managing mortgages can act in a “semi-psychotic” way with no sympathy whatsoever because they are detached from the people they are dealing with. They can sell products that they know are poor and make a lot of money, while not facing the kinds of brakes created by having to look somebody in the eye. So the question is, how does a system scale up sympathy—Adam Smith’s moral sentiment?
Wladawsky-Berger then noted that some of the most exciting Internet-based activities going on today involve transforming the Internet into a far more social network and collaborative platform to bring people together from around the world and let them work with each other. Beyond the social networks is the work being done in open-source communities with Linux and other such areas. These developments depend and are motivated by the concept that sympathy can be scaled. It has to be managed carefully to deal with members of the community who may not care, but the concepts do scale, and there are things being done to make it happen on some level.
Talent-Based Intangible Assets
Wladawsky-Berger closed by reiterating that, while information technology plays a critical role in the transition from the industrial to the knowledge economy, it should be viewed only as an enabler for innovation in business society and people’s personal lives. Some incredible things can be done to better leverage technology on these fronts, and they all involve talent. And if talent capital is the intangible asset that is most important in a knowledge economy, it must have a value in which one can invest and monetize. Wladawsky-Berger added that it is critical that these problems be framed properly, because the battle for talent is going to be the most important one in the economic world.
During open discussion, one participant noted that the potential to learn to improve performance, often thought of as something innate, is also important. Given this, the question was asked “Isn’t the key characteristic the learning itself and the content of knowledge?” For example, R&D extends from basic physics to designing the door handles on an electric car. The R&D category is an accounting convention, but the content is hugely heterogeneous. And one can expand on that for other kinds of knowledge. Wladawsky-Berger responded that “talent” was not the only attribute contributing to capture this knowledge effect. For example, what appears to be most important to his IBM clients is to help them build global enterprises and cope with the changing market and improve their ability to design processes; to become more efficient and robust; to gather information, analyze it and take action on it; and to use social networks and collaborative mechanisms
to help their employees work more effectively with each other and with the outside. A whole set of things exist that are critical to leadership in the knowledge economy. He noted that the key ingredient that cuts across all of them (aside from technology) is the need for talented people. Cutting-edge companies rely on employees who are very well educated. It is conceivable that 20 years worth of talent from current and earlier periods will have been embodied in the right tools, the right software. Of course, this is already happening today, and the talent—that is, the people creating these things—is the most important input. A major way to monetize the value of these human capital components is to translate it into products and tools and the like.
According to one participant, “competence” is another way to characterize the essential intangible asset described in the presentation. Wladawsky-Berger responded that competence is part of the story, but that great collaborative leaders are not just born. A superior manager or organizational leader—a financial person concerned about the impact of his or her decisions on the world—may require the right personality, but many of the skills can be taught. One hopes, he added, that academia contributes to this learning process.
DEFINING INTANGIBLES FOR MEASUREMENT PURPOSES
The measurement of intangible assets as productive inputs involves developing an operational framework that embeds a number of complex methodological issues. Charles Hulten explained that, for macroeconomic analysis, the structural shift that has occurred in business and in the world economy presents a real challenge, in part because most conceptual thinking on growth theory and accounting has heretofore applied to an environment in which the production of goods was the critical process to understand. However, in a world in which fewer and fewer physical goods are being made, at least in this country, it is essential to understand what modern companies really do. Hulten cited Apple Computer, Inc., as an example of corporate success attributable to acumen in design, technical innovation, and marketing—aspects of a business that embody significant integration of intangible assets. The challenge ahead is to try to encapsulate this shift in the nature of production in a set of measures, as well as to develop a parallel theory that sheds light on the structure of the modern economy.
Traditional Views of Intangibles
There has been no lack of interest in the topic of intangibles over time, although the primary focus has often been limited to R&D. This area of research can be traced back to scholars in the 1960s, such as Zvi Griliches, and, in some respects (as discussed by Wladawsky-Berger), as far back as Adam Smith. Beyond R&D, intangibles involve marketing, worker training, and the entire set of coinvestments and surrounding processes of which R&D is only a part. Much
of the literature that exists has taken a part-by-part perspective in studying these components. Hulten stated that one of the challenges is to find a way to integrate these various parts. In order to progress in this direction, better approaches to taxonomy and classification are needed.
According to Hulten, in the view of traditional accounting and of formal economic growth theory, much of what is now discussed in terms of intangibles was not considered investment in a company’s future. Marketing, innovation, and so forth were basically seen as current expenses. These kinds of inputs were treated in a similar manner to any type of material that was used up and that did not generate lasting effects.
The key point is that there was no output associated with the production of these intangibles. Thus, for example, when Apple busily prepared for its future by inventing the iPod and doing all the things necessary to make it a successful innovation, none of that counted toward (GDP) investment by the firm. As discussed below, the Bureau of Economic Analysis (BEA) is now moving to correct this by capitalizing R&D. Hulten pointed out that some intangibles do appear in part in firm accounts as a residual between the market value of the company and the reported book value.
Hulten presented some of the findings that have emerged from his research and that of his colleagues. Corrado, Hulten, and Sichel (2005, hereafter CHS) looked at the relationship between what firms report in terms of their capital base—the equity base—and the book value over time. It is a backward-looking metric of what firms invested and what the market says they are worth. The wedge between the two, the market-to-book gap, is a measure of something that has not been explained. It could be the volatility and irrationality of the markets, but, Hulten posited, the gap is probably a little too big to chalk up to that. Beginning with changes in financial accounting standards that took place in 2001, companies have been asked, when they acquire another company, to fill in as much of that gap as possible with specific line items. And, according to Hulten, this has not been an easy task. He spoke with one executive who had complained how difficult this is.
Part of the difficulty has to do with the lack of conventions in reporting and classification systems. Hulten discussed how classification systems could be modernized to make it possible to move from a more traditional view that ignores own-firm-produced intangibles toward one that accounts for them more explicitly. One alternative for structuring the taxonomy, outlined in Lev (2001), distinguishes three classes of intangibles by their structural characteristics: (1) innovation-related intangibles, such as research, products, and so forth; (2) human resources, such as developing and retaining talent; and (3) pure organizational intangibles, such as management schemes and capacities, use of information technology, and business models in general.
This ambitious taxonomy involves broader coverage than anything that could be implemented in a short period of time. Hulten also pointed out that intangibles that actually appear in a firm’s annual report might involve blends of these conceptual categories. These, he said, are ingredients of the dish, not the dish itself. He then suggested that one way to think about what the dish might look like is to examine the functional characteristics by valuing the output of the tangibles. One prominent strand of research in the late 1990s used the stock market value of companies as a starting point. Hall (e.g., 2001) wrote several papers on this but rejected some of the ideas after the stock market crash following the dot-com bust.1
Another way of proceeding would be to model the classification structure along the lines recommended by the Financial Accounting Standards Board (FASB) for companies acquiring intangibles; this involves forecasting the income streams associated with various identifiable intangibles, such as copyrights, customer lists, and so forth. Business Week is doing something similar in its study on the valuation of brands (Kiley, 2007). The study estimates that the top 100 companies account for about $1 trillion in brand equity, about two-thirds of which is in U.S. companies. Hulten also noted the securitization option. Various organizations, such as Ocean Tomo, discussed later in this report, are attempting to create markets for identifiable intangible assets. This is an interesting and perhaps promising method that could be applied to certain types of intangibles, but perhaps not for all categories.
The final method presented by Hulten on how to proceed involves valuing intangibles according to their input cost. As he noted, this is largely what is done now on the tangible capital side. When a building or automobile is acquired, the value of that investment is recorded according to how much is paid in the transaction; there is no underlying assumption that it must be worth at least that much to the company. This same thinking could apply to intangibles as well and is the insight that led the CHS team in this research direction; Leonard Nakamura (e.g., Nakamura, 2008) has also done work in this area, as have researchers at BEA and the United Nations System of National Accounts. As a first cut at the problem, this is probably the way to go; later, other information may be used as a correlate to enrich the analysis. The problem, as Hulten put it, is one of trying to put a reasonable set of principles around a broad set of intangible asset types. He and his colleagues are encouraged by the results produced from treating intangibles analogously with tangible capital—that is, as resources that could be used in the production of current or future consumption and hence conceptually satisfy the definition of investment.
Unique Measurement Difficulties for Intangible Assets
Significant conceptual differences between tangible and intangible assets exist that have measurement implications, not the least of which is the potentially very long gestation lags—up to 14 years, for example, for a pharmaceutical patent life. Another is that intangibles, like R&D, are not continuous inputs to production. If Apple doubles its R&D budget, it will not necessarily double its production of iPods. This involves a different concept of the transformation of inputs into outputs that deviates from a conventional production function; it is a more general transformation process in which these intangibles are not fixed costs, but continuous inputs in other dimensions.
Also, with tangible assets, one can observe on a plant floor or in a parking lot the various machines and figure out what vintage models they are and depreciate accordingly. One can inspect what has survived from the past. Intangibles, in contrast, are generally produced within the company, but there is no market transaction. This presents a measurement complication and may be the principal reason that accountants have been reluctant to try to capitalize them; there is too much latitude for discretion and all that it implies. But it also leaves the statistician and the growth accountant with the difficult task of trying to figure out what these assets are really worth.
An example of the difficulty of establishing a link between intangible asset value and age is illustrated in an article by Jonathan Rauch (Atlantic Monthly, July/August 2008) describing the Chevrolet Volt electric car program. According to the article, the enabling idea for the research came from a program that had been discarded because the electric charge for the car was inadequate to allow it to drive very far. General Motors had the idea to use the gasoline engine to recharge the battery—an old idea, but one that would not be seen “in the plant.” It would have been difficult to observe this input.
Another important measurement difficulty discussed by Hulten concerns the strong nonrival characteristics often present in knowledge capital—patents are an exception. When intangibles are discussed as something symmetric with tangible investment, one is dealing only with the commercialized part of the investment, because a process of diffusion ultimately takes place for any major set of ideas that develops. Knowledge diffuses to other companies, to other industries, and to other countries, either directly copied or through spillovers; at this point, it benefits society, not as an investment, but as an idea that penetrates the business culture and is reflected in increased productivity and lower costs. Thus, a disconnect exists between the private return to intangibles and their general social value (their public return). Hulten suggested that the broader impact is more relevant from the point of view of public policy, and the private return is most relevant for establishing dollar metrics.
CHS Methodology and Results
Hulten summarized some of the conclusions of the CHS research to implement an input-based model, acknowledging along the way the limitations and question marks associated with the approach. In that research, the authors looked at software and innovative property, which amounts to a broader definition of intangibles than scientific R&D only. In constructing a list of items to be accounted for, he acknowledged an element of arbitrariness in drawing the boundary. For example, they included only about 40 percent of advertising and had to make assumptions about other categories as well. Hulten suspects that, if anything, the research underestimated the size of more categories than it overestimated. For example, CHS include only advertising and not marketing, which tends to be a much broader expenditure. At IBM, the advertising budget is only one-sixth of marketing; this is probably roughly true of the pharmaceutical industry as well.
Although sensitivity analysis has been left for the next generation of research on the topic, CHS were able to show that including intangibles makes a measurement difference. Hulten noted that scientific R&D (along the lines defined by the National Science Foundation, NSF, for its survey) is not particularly significant—and to think that R&D equals intangibles would be a mistake. By their estimates, in 2003, R&D accounted for just under $200 billion of the over $1 trillion spent on intangibles, which is greater than for business–fixed investment. The spending levels are also higher than they are for many of the inputs looked at as indicators of innovation activity—such as computer investment, software investment, and conventional R&D. So these, in and of themselves, do not provide a full picture of what is happening with business processes.
The CHS analysis also provides some evidence on the extent to which measured GDP gets a boost from including investment in intangibles. Hulten again emphasized the idea that research on the topic is needed not only to measure inputs into production more accurately, as an enabler of making things, but also to pick up the augmentation on GDP itself. These intangible resources are being used to build the future just as tangible assets are. It follow that they should be counted as GDP—and, in fact, such an approach makes a large difference in final estimates.
CHS estimated that the stock of intangibles amounted to about $3.6 trillion in 2003. Again, because of relatively high rates of depreciation assumed in the paper (compared with similar work in the literature), Hulten suggests that this may be an underestimate of their actual levels. The depreciation issue is an important question, though, and one that BEA is already working hard on.
Independent of the depreciation issue, ignoring intangibles leads to more than a level error in statistics. It is not that measured GDP will be wrong by 10 percent now and 8 points in the past. Business intangibles are growing as a fraction of business output. There are also huge private expenditures—on education, for example—that are another large source of investment for the economy. In this realm alone, the growing presence of these factors changes the dynamic path that researchers are analyzing.
Hulten then returned to the book-to-market-value gap issue. Taking the strategy developed in CHS and using Compustat data, he estimated the value of intangibles at cost. He analyzed individual companies at the level of their income statements and balance sheets to ask the question, “How much of the gap could we explain?” The answer turned out to be “a lot” for a sample of research-active firms. Although the picture is incomplete—it is hard to derive from Compustat data what other firms would look like, because so many data points are missing—the bottom line is that much of the gap can be filled. One can go from explaining one-quarter to one-third of the gap under the pure-equity view of the world (the prevailing view in accounting) to 80 or 90 percent. The conclusion that is warranted from the research is that good progress can be made to understand much of corporate valuation by including intangibles. Investors and corporate managers ought to be interested in this information as well, or some more accurate version of it, and it might be something that they themselves could be encouraged to produce.
Pointing toward the next presentations on the program, Hulten reiterated that intangibles are important for measuring total factor productivity. CHS showed this, and there is evidence dating back to the literature of the 1960s and even before, but it has a defeatist tone along the lines of “Yes, this stuff should be in there, but what can you do?” Investment in intangibles, however, is something that both researchers and policy makers should attend to, since it determines a lot of what transpires in the economy and, in particular, it shifts our sense of what drives growth. There is, Hulten concluded, lots of work ahead in order to develop a more complete understanding of both the private and social returns from intangible investment.
During open discussion, Flamm pointed out that advertising and brand equity expose the empirical difficulties of trying to measure intangibles. For example, when Macy’s runs an ad in the newspaper, a part of the function of that ad is building the Macy’s brand image, but the firm is also communicating information about prices or product characteristics, which has a very transitory value and which clearly is not a capital investment in any sense. In this case, there is a combination of investment in an intangible and a current expenditure in producing sales. He wondered how these components might be disentangled and whether the method for doing so would be the same for retail stores as for computer manufacturers or for advertising in some other industry.
This is where a lot of conceptual spadework needs to be done, Hulten responded. The default assumption seems to have been, at least in many areas, that advertising was primarily transitory. In areas where there is a lot of technical innovation, that may be a mistake, according to Hulten. Wladawsky-Berger added the point that maybe Macy’s advertising in The New York Times and its sponsorship of the Fourth of July fireworks and the Thanksgiving parade could be reasonably separated, as those things are quite distinct; and maybe that is at the root of short-term versus brand-building activities and actions.
Hulten noted that a lot of new pharmaceutical drugs emerge as a result of intensive R&D over a long period of time by small biotech companies. However, in the later phases of the development, these companies tend to want to partner with a larger pharmaceutical firm, in part, to help with the approval process by the Food and Drug Administration and also to use the marketing muscle and sales expertise of these companies. Typically, when drugs become successful, they do not suddenly jump to a huge market share. Their success is more often the result of serious effort, and so these calculations get factored in at an earlier stage in the innovation. He cautioned against the too-easy assumption that all such expenditures should be considered transitory.
During open discussion, Senator Bingaman was asked his view about the pace and the outlook for improved financial accounting in terms of disclosure and transparency, given that it is from the business sector that the statistical agencies would ultimately look to for source data. He responded that the likelihood of any action depends on whether a strong justification can be produced for requiring changes in accounting in the area of intangibles. If, he said, there is a real purpose served by it, then support can be generated for making those kinds of changes. To the extent that it is solely an academic inquiry that does not have clear policy implications, either for the company or for investors or for public policy issues, then it is more difficult.
Another participant asked whether the senator, as a member of the finance committee, could assess the outlook for the creation of tax incentives for intangibles. He noted that various members of Congress had been dealing with the R&D tax credit for a long time, but that no one could seem to make progress on an investment or knowledge tax credit for worker training, although the idea has been around for at least a decade and a half. He was also asked whether there was any hope of making some of these things, like the R&D tax credit, permanent.
Bingaman replied that the only way to make the R&D tax credit permanent would be as part of a larger reform effort. Whether the new administration will ever have this as a top priority was unknown at the time of the workshop. He added that there is a tendency to think that it will happen, but the reality is that the current deficit situation is such that the nation has to find a way to generate more revenue; rather than just raising taxes, politicians typically like to do that as part of a tax reform package. He expressed the view that the chances of a substantial tax reform might be reasonably good in the next Congress, and maybe some of these things could be done as part of that.
He noted that the reason many tax credits are temporary in the first place is because of the Budget Act of 1974, which, by putting limits on the size of the deficit that can be run each year, causes Congress to pass short-term tax extensions instead of making them permanent. The Joint Tax Committee takes the position that the effect on cost projections of making them permanent is substantial, and no one wants to have to factor that in. The president does not want to factor it into the budget he sends to Congress, and Congress does not want to factor it into the budget it passes.