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.
Evening Session October Il. 1989 DR. PRESS: Good evening, ladies and gentlemen. Thank you for joining us to discuss this important topic. As everyone here this evening realizes, the health- or distress, as some may see it of U.S. industrial R&D today is going to affect American competitiveness in the future. With the wave of corporate restructuring that has swept over the American corporate scene, there is a sense of timeliness in this symposium, and I hope that with your help we will clarify some of the issues regarding its impact, particularly on research and development. But rather than introduce the subject of the symposium, which I will leave to tonight's speaker, who has studied the issue in depth and who, I am sure, will have many provocative points to make, I would like to say a few words about the National Research Council's Academy Industry Program. As you know, the National Research Council is the operating arm of the National Academy of Sciences, the National Academy of Engineering and the Institute of Medicine. As most of you probably know, most of the National Research Council's work is generated by requests from the federal government. Back in 1983, we in the National Academy of Sciences, the Academy of Engineering and the Institute of Medicine decided to expand our outreach efforts, beyond our traditional constituencies in government and academia, to industry, particularly those in corporations, whose products result from advanced research and development. We therefore established the Academy Industry Program, which is the National Research Council's official liaison to American industry. In 1
2 CORPOMTE RESTRUCTURING addition to putting on programs like the one tonight and tomorrow, the AIP regularly disseminates relevant NRC studies to corporate executives and keeps them informed generally about our activities. In addition, the program receives support from approximately 70 mem- ber companies each year, which together with support we receive from phil- anthropic foundations allows us to initiate studies on our own and reduce our dependence on the federal government. That money enables us to ask questions and raise issues, perhaps, before they appear on the federal radar screens. So, I welcome you to the Academy complex and to the Academy Industry Program, in particular those guests who may not have been here before. Before introducing tonight's speaker, I want to take this opportunity to acknowledge the participation of Mac Booth, President and CEO of Polaroid, in this symposium. Polaroid is and has been one of the mainstays of the Academy Industry Program and we appreciate that. I also want to publicly thank Stu Eizenstat, who will moderate the symposium tomorrow, for his help in putting this all together. We are particularly fortunate tonight to have as our keynote speaker, Joseph Grundfest. Mr. Grundfest has been a Commissioner on the Se- curities and Exchange Commission (SEC) since 1985. He has shared responsibility for regulating this nation's securities markets during what has obviously been their most tumultuous period in recent memory, with the crash, the insider trading scandals and the growth of sophisticated institutional investment instruments. His background helped prepare him for this position. He is both an attorney and an economist, and prior to becoming a Commissioner at the SEC, he was a senior economist on the President's Council of Economic Advisors, with responsibility for legal and regulatory matters. He has practiced law and has worked at both the Rand Corporation and the Brookings Institution. Mr. Grundfest is author or co-author of numerous research reports and publications. His works deal with a range of topics, including competition for corporate control, insider trading, international securities regulation, regulation of markets subject to kick-back schemes, the economics and reg- ulation of broadcasting and the role of citizen participation in administrative proceedings. Mr. Grundfest's education is equally varied. He received a BA in eco- nomics from Yale and a master's degree in econometrics from the London School of Economics. In addition, he earned a law degree from Stan- ford in 1978. Among his honors has been a National Science Foundation Fellowship and I know those are highly competitive. When Mr. Grundfest leaves the Securities and Exchange Commission
EVENING SESSION, OCTOBER 11, 1989 3 sometime next year, he will go to Stanford, where he will teach at the law school. Also, next year, Little, Brown will publish a collection of his essays and speeches under the title, Perestroika on Wall Street. With that, it gives me great pleasure to introduce him to you. Mr. Grundfest. The following is the prepared text of Commissioner Grundiest's talk. The views expressed are those of Commissioner Grundfest and do not nec- essarily represent those of the U.S. Securities and Exchange Commission, of other commissioners, or of the Commission's staff. It is an honor and privilege to be invited to address this distinguished gathering at the National Academy of Sciences. The privilege is particularly great because it affords an opportunity to consider a topic of substantial national concern: the relationship between America's research and devel- opment efforts and the growing wave of restructuring activity now sweeping our corporate sector. It probably comes as little surprise that hordes of critics stand ready to condemn corporate restructuring in the United States as bad for workers, harmful to local communities, damaging to U.S. international competitive- ness, and threatening to the financial stability of the U.S. economy.) The most damaging allegation against corporate restructuring may, however, be the charge that it stifles corporate R&D and forces management to adopt irrationally myopic strategies.2 1b my mind, this allegation is the most se- rious potential indictment because, if true, it suggests that restructuring is eroding America's industry at the point where it is most vulnerable: at its knowledge base. The reality of today's marketplace is that a firm's knowledge capital measured in terms of its know-how, technical expertise, trade secrets, patents, productive processes, and accumulated R&D efforts is often far more important than its physical capital measured in terms of bricks and mortar, lathes, delivery trucks, or power plants.3 More and more, capital ~ See, e.g., M. Lipton, Corporate Governance in the Age of Finance Corporatism, 136 U. Pa. ~ Rev. 1 (1987~; L" Lowenstein, Management Buyouts, 85 Colum. In Rev. 730 (1985~; LBO's: Friend or Foe of Industrial Research.7, 31 Research & Development 13 (April 1989~; WhatAre LBO's Doing to R&D?, Chemical Week, Feb. 15, 1989, at 26; Thurow, US. Can't Compete if Finance Continues as the Master of Industry, LEAS Times, Nov. 17, 1985, Pt. 5, at 3. 2See, e.g., J. Stein, Takeover Threats and Managerial A1yopu~, 96 J. Polit. Econ. 61 (1988~; P. Drucker, A Crisis of Capitalism, Wall St. J., Sept. 30, 1986, at 32, colt 3; "Long-Tem~n Band- wagon Hot, Pensions ~ Investments Age, May 1, 1989. ("Business executives have argued that short-term pressures placed on many mangers force . . . [executives] to take a short-term view.") 3See, e.g., B. Hall and R. Hayoshi, Research and Development as an Investment, Nat. Burl Econ. Res. Working Paper No. 2973, May 1989, at 2, 33.
4 CORPORATE RESTRUCTURING exists as software that is invisible to the eye rather than as hardware that we can touch or feel. This "knowledge capital" is not easily inventoried by accountants or valued by appraisers. In many situations, it is not even easily described. Yet, while knowledge capital cannot easily be measured, without it our economy surely will not thrive. Unfortunately, when it comes to evaluating our stock of knowledge capital the indications for the U.S. economy are quite depressing. At the most elementary level, our school children do not read, write, or add as well as school children in the countries that are our strongest economic competitors.4 The test scores of entering college freshmen are in a long-term decline, having peaked in 1963 and having recovered only slightly from a 1980 trough.5 The adult literacy rate is a national disgrace.6 The educational situation in our country is in such shambles that many companies have to provide remedial literacy and arithmetic training on the job just so they can have a work force competent to operate an efficient production process.7 The time has passed for the production line worker who knows only how to turn a wrench or drive a screw. Technology will not create high- paying, high-quality jobs for a nation of illiterates who cannot add. Until we turn this situation around, and start educating a work force smart enough to master the demands of modern production technology, it is hard to understand how we can expect to make progress against competitors who can read operating manuals that we can't and perform statistical analyses that we don't understand. Viewed from the shop floor, that's the cold reality of the world we face today. Viewed from the corporate boardroom, our comparative situation may not be much better. Although corporate spending on R&D climbed steadily at an inflation-adjusted rate of 5.8 percent per year in the decade preceding 1986, aggregate corporate spending on R&D has slowed substantially since 4 See A. McLaughlin, Education and Work: TheMissingLu~k, N.Y. Times, Sept. 25, 1989, at A19, cot. 1.; Vagelos, The Sony State of Science Education, Scientific American, Oct. 1989 at 128. 5See L Feinberg, Student's Scores Drop in Test of Verbal Skills, Wash. Post, Sept. 12, 1989, at A18. See also Apptitude Test Scores Drop for Women and Minonties, LA. Times, Sept. 11, 1989, at A2, colt 3. 6 See ~ McLauglin, Education and Hark, supra n. 4, ("Between 20 million and 40 million adults today have literary problems'); S. Knell, An Investment in Hump BettermentAdult Literacy, Chi. Tribune, Sept. 18, 1989, at C13 ("There are an estimated So million to 30 million [illiterate] adults in the United Stately; M. Spencer, Why Won't Jury Learn? Look in the Mirror, L.A. Times, Sept. 9, 1989, at Part 9, page 1, cot 1 ("adult illiterary [is] running at a 20% rate"~. 7See J. Berger, Skills ~ Jobs: The Classroom Mismatch, N.Y. Times, Sept. 26, 1989, at A1, colt 2; C. Sklzycki; The Company as Educator: Rums Teach Workers to Read, Write, wish. Post, Sept. 22,1989, at G1.
EVENING SESSION, OCTOBER 11, 1989 s then.8 When these figures are adjusted to reflect it&D/sales ratios and inflation, the recent slowdown in aggregate R&D expenditure may not appear quite as foreboding.9 However, once we compare U.S. research expenditures with our Japanese and West German competitors, the picture turns dim again. Nondefense corporate spending on R&D in the United States stood at about 1.9 percent of gross national product (GNP) as of year-end 1987.~° In Japan, the comparable measure of R&D spending ran at Z7 percent of GNP and in West Germany, it stood at 2.6 percent of GNP.~i Most ominous, perhaps, is the fact that Japanese commercial R&D expenditures are growing at a far faster rate than comparable U.S. expenditures. Thus, not only are we falling behind at the most elementary levels of our knowledge base, but if the trajectory of R&D expenditures as a percentage of GNP is a harbinger of future trends, then it appears that we may also be falling behind at the most advanced levels of our knowledge base as well. Simply put, we are in deep trouble. If we don't invest in our knowledge base both at the top and at the bottomwe are going to lose out in the battle for international economic competitiveness. The question is not if we will lose out; the only question is when. Faced with this serious predicament, wouldn't it be wonderful if we could find a quick, simple, and popular cure for at least part of this ~See, e.g., Testimony of Erich Bloch, Director, National Science Foundation, Before the House Ways and Means Committee, March 14, 1989, at 2 (hereinafter cited as "Bloch Testimony'); Business TaBcs a Better R&D Game Than It Plays, Bus. Wk., Aug. 21, 1989, at 20 (economists estimate that real outlays on research and development will increase by lees than 1% in 1989, compared with 1.3~o in 1988 and 3.7~o in 197; R. Cassidy,ResearchFundin~g~r1989 Won 't Even Reach $131 Billion, Research ~ Development, Jan. 1989, at 47; R. Winter, Research Sparing in US. to Slow in 19&9, Wall St. J., Dec. 21, 1988, at B3, colt 1. 9K Plamm, Indus~calResearch and CorparateRestructuri~g An O~vofSomeIssues, Septem- ber 1989, at 4-5. ("In the face of a declining sales base in 1985 and 1986, the relative size of the research efforts of American R&cD-performing companies increased. When sales picked up in 1987, the interest of these companies' research efforts decreased.") 10U.S. Department of Commerce, StansticalAbs~act of the United States, 1989 at 578, 1bble 973 ("Statistical Abstract"~; Missed Oppormniiies: R&D A Bigger Push u' Japan, Wall St. J., Nov.14, 1988, at R21, colt 3. See also Clark tic Malabre, Slow Rise u' Outlays for Research Impairs US. Competitive Edge, Wall St. J., Nov. 10, 1988, at A1, colt 6. For purposes of the present analysis I am focussing on nondefense RED because it is unlikely that restructuring activity has a meaning- ful effect on defense R&D spending levels. In addition, nondefense R&D expenditures are more directly related to the economy's international economic competitiveness than military R&D ex- penditures. See, e.g., Reich, The Quiet Path to Technological Pre-erruna~ce, Scientific American, Oct. 1989, 41, 44. ("Several factors impede technology transfer from military to commercial ap- plications . . . military R&D has become an inefficient means of generating commerical spin- o~s....") 1 Statistical Abstract, Dora n. 10; Missed Opportunities, supra n. 10. Missed Opp~u;nities, supra n. 10, at R22.
6 CORPORATE RESTRUCTURING problem? Wouldn't it be wonderful if we could establish that corporate restructuring has become a millstone around the neck of America's R&D efforts? Wouldn't it be wonderful if we could just stop all this restructuring and thereby restore at least part of America's R&D vitality? Yes, it would be wonderful, but it wouldn't be true. The best available evidence suggests that corporate restructuring has relatively little to do with our declining international position in the R&D race. Thus, even if we placed substantial constraints on takeovers, leveraged buyouts, spinoffs, stock buybacks, leveraged recapitalizations, and other forms of corporate restructuring, I doubt that we would accomplish much, if anything, in terms of restoring the vitality of America's R&D efforts. The facts that lead me to this conclusion are not pretty, but with your indulgence I'd like to review them in some detail. THE COST OF CAPITAL If corporate restructuring is not a major cause of the relative decline in U.S. R&D activity then what is? One need not search hard or long for the answer to this question because the primary culprit is quite clear: it is the cost of capital. The evidence is overwhelming that the cost of capital for R&D projects in the United States is far higher than it is in Japan or West Germany. Because our capital costs are higher, it Is more expensive to conduct R&D In the United States. Moreover, because our capital costs are higher, the projects we conduct must have faster payoff penods, and we cannot afford to undertake projects that are as Oslo as some projects conducted In Japan or West Germany. Just how much higher are our capital costs than Japan's or West Germany's? Recent estimates by staff of the Federal Reserve Bank of New York are truly frightening.~3 After adjusting for inflation, tax rates, and other factors, the New York Fed study suggests that in 1988, the average annual effective cost of capital in the United States for a benchmark R&D project was 20.3 percent. The cost of capital in West Germany for the same benchmark project was 14.8 percent, and in Japan it was only 8.7 percent.l4 Thus, capital for R&D purposes is now more than twice as expensive in the United States as in Japan. Unfortunately, the conclusion that U.S. capital costs are higher than foreign capital costs is borne out in several recent studies.l5 13R. N. McCauley and S. ~ Zimmer, ExplainingInternaiional Differences in the Cost of Capital, Federal Reserve Bank of New York, Quarterly Review 7, 16 (Summer 1989~. 14Id. 15B. D. Bernheim and J. Shoven, Taxation and the Cost of Capital: An International Compari- son in ~ E. Walker and N. A. Bloomfield, The Consumption Tax A Better Alternative? (1987),
EVENING SESSION, OCTOBER 11, 19&9 7 These cost of capital figures are consistent with calculations of required breakeven periods for capital projects in the United States and Japan. For example, a recent Stanford University study suggests that capital costs in the United States imply an average breakeven period for new investment projects of 5.7 years. In contrast, lower capital costs in Japan push the Japanese breakeven period out to 10.3 years. Under these circumstances, an eight-year research project that seems perfectly reasonable to a Japanese manager may be totally out of the question for a U.S. manager not because the American lacks the wisdom, vision, or will, but simply because Americans can't rationally bear tile cost of capital. The implications of these high capital costs were well explained by Carl Ledbetter, formerly the president of ETA, the supercomputer division of Control Data that was recently shut down in response to competition from Hitachi, Fujitsu, NEC, and others.~7 As Mr. Ledbetter put it, "If our capital costs had been lower, ETA could have survived." Touche. If our capital costs were half of what they are today, our R&D efforts would mushroom. With lower capital costs we too could afford to be much more patient in waiting for R&D efforts to pay off; and we too could afford to take risks that are just too far-fetched when you're facing a hurdle rate of return of 20 percent per year. The key then is to identify the cause of high capital costs in the U.S. markets and strike at those economic factors. The most recent evidence suggests that our capital costs are higher than Japan's predominantly be- cause our savings rate is lower and because we have less macroeconomic stability as reflected in the volatility of price levels and interest rates.~9 at 78. Bernheim and Shoven's estimates indicate that the cost of capital calculated at the av- erage interest and inflation rates for the 1980s, using 1985 tax codes, was 5.48 percent in the United States, 4.39 percent in West Germany, and only 2.76 percent in Japan. In addition, G. Hatsopoulis and S. Brooks, "The Gap in the Cost of Capital: Causes, Effects, and Remedies," in R. Landau and D. Jorgenson, Technology and Economics Policy (~1986) estimate that the cost of capital in the United States is almost three times higher than in Japan. For other estimates see A. Ando and A. Auerback, lhe Cost of Capital in the United States and Japan, 2 J. Japanese and Int. Economics 134 (1988~; A. Ando and A. Auerbach, "The Corporate Cost of Capital in Japan and the United States: A Comparison," in J. Shaven, Government Policy Towards Ir~soy ~ the United States and Japan (19~; Corcoran and Wallich, Lee Analyncal Economist: Me Cost of Capital, ScientiSc American, Oct. 1989, at 79. 16Bernheim and Shaven, Hera n. 15. See also L" Richman, How Capital Costs CrippleAmenca, Fortune, August 14,1989, at 50. 17Richman, supra n. 18, at 50. i8Richman, supra n. 16, at 50. i9See, e.g., McCauley and Zimmer, supra n. 13. ("Higher household savings in Japan and Ger- many and more sucessful policies for maintaining stable growth in Japan and stable prices in Germany have opended up the [capital cost] gap.')
8 CORPORATE RESTRUCTURING There is little or no support for the idea that corporate restructuring is a meaningful cause of high capital costs in the United States. Thus, even if we shut down all corporate restructuring activity we would still have a substantially higher cost of capital and we would still find ourselves falling behind Japan in the international R&D race. Moreover, if one wants to attribute causality, I think the better argument is that the high cost of capital in the United States causes both a decline in relative R&D intensity and an increase in corporate restructuring activity.20 Restructuring and R&D are thus codetermined by interest rates, and restructuring does not, in and of itself, determine R&D intensity. Put another way, increased restructuring activity does not cause a decline in R&D activity any more than an increase in R&D activity would cause a decline in restructuring. Addressing this problem will not be easy. ~ bring down the cost of capital we must increase our domestic savings rate and restore macroeco- nomic stability over a substantial period of years. That's nowhere near as exciting, dramatic, or simple as putting an end to corporate restructuring, but I'm afraid that this painful prescription is the only one likely to do meaningful good. THE ARGUMENT BEYOND CAPITAL COSTS Having argued that capital costs explain the lion's share of the R&D problem facing American industry, I could end right here and suggest that we stop blaming corporate restructuring for something that isn't its fault. While this approach has the virtue of being concise, it suffers from the vice of overlooking much additional evidence exploring the relationship between restructuring and R&D. ~ develop this perspective more fully, let's proceed on the obviously unrealistic assumption that capital costs have nothing to do with the R&D decision, and let's explore the evidence that focuses narrowly on the relationship between restructuring, stock market valuation, R&D activity, and the alleged myopia of America's capital markets. THE SECTORAL INCIDENCE OF RESTRUCTURING ACTIVITY If corporate restructuring activity truly has a major impact on aggregate R&D expenditures, we would expect to find that corporate restructuring occurs with some frequency in industries that engage in significant amounts 20High capital costs imply a greater scarcity value for capital. This higher price of capital sug- gests that the market will be less tolerant of managements that fail to earn adequate returns. Colporations that accumulate "free cash flow," ret, cash flow that could be reinvested outside the corporation more profitably than it can be reinvested in the corporation, are wasting valu- able capital and are potential subjects for restructuring. See, e.g., Jensen, He Avenge Cost of Free Cash Flow, 76 Amer. Econ. Rev. 490 (May 19863.
EVENING SESSION, OCTOBER 11, 19~ 9 of R&D. The data, however, fail to support this hypothesis. As explained by Professor Lawrence Summers of Harvard University, Governor Dukakis's chief economic adviser during his recent presidential campaign, "impost LBOs occur in mature industries that do not spend a lot on research, so there is not yet much evidence to support claims that R&D is severely cramped by LBOs."2t Indeed, most restructuring takes place in industries that are not R&D intensive, so the point extends beyond simple LBO transactions.22 This observation can be made much more simply: Bloomingdales didn't do much R&D before its takeover and it hasn't done much R&D since. The same holds true for takeovers, buyouts, and restructurings involving companies like Beatrice, United Airlines, Stop N' Shop, Allied Stores, Avin, Storer Cable, Columbia Pictures, Burlington Industries, and hundreds of others engaged in non-A&D intense lines of business. Indeed, an examination of all takeovers in 1988 suggests that more than 75 percent of the dollar value of merger and acquisition (M&A) transactions occurred in industries such as retailing, food products, broadcasting, and insurance industries in which R&D is not perceived as a major competitive factor.23 Measures aimed at hobbling takeovers and buyouts in the name of protecting America's R&D effort are therefore clearly overbroad because they would impose substantial restrictions on transactions that have nothing to do with R&D. Indeed, Professor Frank Lichtenberg of Columbia Uni- versitr has suggested that managers and administrators who typically lose jobs following restructurings may be raising false alarms about the general effects of takeovers on R&D.24 Thus, the R&D argument may make good public relations in the campaign against restructuring, but it is totally irrel- evant to the large majont~r of restructurings to which the argument might be applied. R&D INTENSITY OF RESTRUCTURING TARGETS But suppose we pursue the argument further and eliminate from consideration companies lee Bloomingdales where R&D isn't a meaningful issue. What then do we find? This question can be posed in two ways. First, what difference is there in the R&D intensifier of restructuring targets prior 21L. Summers LBO Debt and Taxes, Across the Board, Vol. XXVI, No. 4, April 1989, at 53, 54. 22See, e.g., B. Hall, Leveraged Buyouts, Corporate Debt, and R&D Investment: Is There Any C<7n- nection? (work in progress, version of Sept. 1989) at 6. 2 31988 Profile; MergerActrvity by Indus~yArea, Mergers and Acquisitions, at 54 (May/June 1989). 24Testimony of Frank R. Lichtenberg in hearings on "Corporate Restructuring and Its Ejects on R&D" before the Science, Research, and Technology Subcommittee of the House Committee on Science, Space, and Technology, July 13, 1989, at 2.
10 CORPORATE RESTRUCTURING to restructuring activitydo they conduct more or less R&D than their industry peers? Second, what difference is there in the R&D intensity of firms that have been restructureddo they conduct more or less R&D after the restructuring than before? Posed either way, the answer to the question is fascinating and again fails to support the argument that restructuring is a primary factor hobbling U.S. R&D activity. Let's look first at the data describing the prerestructuring R&D inten- sity of firms that conduct a meaningful amount of R&D. A recent study by economists at the Securities and Exchange Commission's (SEC) Office of Economic Analysis found that on average, "takeover targets undertake less R&D than non-targets in the same industry."25 This result is hardly novel and reaffirms earlier consistent findings at the SEC,26 at Harvard,27 and at the National Bureau of Economic Research that also show that much of the takeover activity in the U.S. market was directed "toward firms and industries that were relatively less R&D intensive and had a weaker technological base."28 The implications of this result are quite significant. If companies that engage in above average amounts of R&D are setting themselves up to be targets of restructuring activity then we should find that, within industries, takeover activity is targeted at highly it&D-intensive firms. Instead, the data support exactly the opposite conclusion and it appears that industry laggards who fail to do as much R&D as their counterparts are more likely to be takeover targets than the industry's takeover leaders. Thus, the image of America's R&D leaders as being under the restructuring gun appears to be at odds with at least one large and inconsistent fact: The companies feeling the heat from restructuring activity appear to be those that have failed to invest as much in R&D as their industry counterparts not the other way around. R&D CHANGES AFI ER RESTRUCTURING Having established that the bulk of restructuring activist occurs in industry segments that are not R&D intensive, and that the targets of restructuring activity tend to do less R&D than their industry peers, a 25L. Meulbroek et al., Takeover Threats and Research & Development: Testing Stein's Model of Managerial Myopia, at 6 (1989), J. Polit. Econ. (forthcoming). 26Securities and Exchange Commission, Office of the Chief Economist, Institutional Ownership, Tender Offers, and Long Term Investment, 1985. 27S. Addanki, Innovation and Mergers in US. Manufac~ringF1rms: A First Look, Department of Economics, Harvard University, 1985 (cited in Hall, supra n. Z). 28B. Hall, The Effects of Takeover Activity on Corporate Research and Development at 93, cited in A. Auerbach, Corporate Takeovers: Causes and Consequences (19883.
EVENING SESSION, OCTOBER 11, 1989 11 meaningful question nonetheless remains: What happens to R&D at those companies that actually do R&D and that are subject to restructuring? At this point we encounter a hailstorm of anecdotal evidence suggesting that R&D expenditures are viciously slashed in the wake of corporate restructuring efforts.29 For example, one economist claims that "one of the things that gets squeezed [in a restructuring] is R&D, because that's an investment in the future.... Whatever costs are postponable are likely to go by the boards."30 Parsing the evidence on this score is quite an interesting exercise because even if there is a postrestructuring decline in it&D/sales ratios, that decline could reflect greater efficiency resulting from economies of scale. A decline in R&D expenditures could also reflect a decision to kill R&D projects that have become white elephants. However, before pursuing these avenues of inquiry, let's take a step back and explore the evidence about postrestructuring R&D expenditures. Do R&D expenditures increase, decrease, or stay about the same following a corporate restructuring? The evidence on this score is more mixed than for the other points I have discussed. However, based on current research, I would characterize the data as either supporting the "no difference" conclusion or as too uncertain to support any conclusion. Lichtenberg and Siegel, for example, found that "the average R&D intensity of firms involved in LBO's increased at least as much from 1978 to 1986 as did the average R&D intensity of all firms responding to the iNational Science Foundation] NSF/Census survey of industrial R&D."3i In earlier research the same authors found that R&D employment does not change following restructuring, even though there is a substantial decline In nonproduction employment, that is, of managers and administrators who work at corporate headquarters.32 Similarly, Bronwyn Hall examined all takeovers of publicly traded manufacturing firms between 1976 and 1986 and concluded that the data "provide very little evidence that acquisitions cause a reduction in R&D spending." In the aggregate the firms involved in mergers were in no way different in their pre- and postmerger R&D performance from those not 29 See, e.g., Statement of Dr. Julie Fox Gorte, project director, Office of Technology Assessment, Before the Subcommittee on Science, Research, and Technology, Committee on Science, Space, and Technology, U.S. House of Representatives, July 13, 1989, at 5-10. 30C. Sk~ycki, Impact on R&D is Newest Worry About LBO's, Wash. Post, Dec. 18, 1988, at H1, colt 3 (quoting Walter Adams, professor of economies at Michigan State University). C. SkIzycki, Impact onR&Dis Newest Wony About LBO's, Wash. Post, Dec. 18, 1988, at H1, colt 3 (quoting Walter Adams, Professor of Economics at Michigan State University). 3iF. Lichtenberg and D. Siegel, The Effects of Leveraged Buyouts on Productivity md Related Aspects of Firm Behavior, Nat. Burl Boon. Res., Working Paper No. 3022, June 19&9, summary. 32F. Lichtenberg and D. Siegel, The Effects of Takeovers on the employment and Wages of Central Of fire and Other Personnel, Nat. Burl Econ. Res., Working Paper No. 2895, March 1989.
12 CORPORATE RESTRUClURING so involved. At the individual industry level, however, the results were too imprecisely measured to draw solid conclusions.33 Abbie Smith, in a study of 58 management buyouts (MBOs) completed between 1977 and 1986, finds a "substantial increase in profitability follow- ing the MBo."34 She concludes, however, that these increased profits are apparently not due to "pervasive cutbacks in 'discretionary expenditures' such as maintenance and repairs, advertising, or R&D which might lead to a longer run decline in cash flows."35 In particular, Smith finds that the "median ratio of R&D expense to sales increases from .012 in the year preceding the MBO to .018 in the year following the MBO, with a median change of 0.00 for the seven firms with available data."36 Moreover, there is anecdotal evidence that R&D expenditures may actually increase in some situations following a takeover or restructuring. For example, after Hoechst's purchase of Celanese Corporation, R&D spending increased by 10 percent annually.37 Data prepared by Kohlberg, Kravis & Roberts, America's leading leveraged buyout firm, also confirms that R&D expenditures decline prior to leveraged buyouts and suggests that KKR, at least, budgets for aggregate increases in R&D expenditure.38 On the other side of the ledger, however, stands a recent NSF study that examined R&D expenditures at the 200 largest it&D-performing firms in the United States.39 These firms account for almost 90 percent of industrial R&D spending in the United States. Within this sample, the 33Hall, supra n. 28, at 93. 34A. Smith, Corporate Ownership Structure and Performance: The Case of Management Buyouts," Univ. Chi. June 1989, at 1. 35Id.at2. 36Id. at 24. 37Testimony of Dr. Julie Fox Gorte, supra n. 29, at 93. 38Kohlberg, Kravis, and Roberts, Presentation on Leveraged Buyouts at 8-1 (Jan. 1989~. 39National Science Foundation, Corporate Mergers Implicated in Slowed Industrial R&D Spend- ~ng, Washington, March 1989 ("NSF Study"~; Testimony of Erich Bloch, director, National Sci- ence Foundation, House Ways and Means Committee, March 14, 1989 ("Bloch Testimony"~. I exclude from consideration the findings of the Ravenscraft-Scherer studies, which, based on 1977 data, find that "lines of business originating from mergers had significantly lower compa- ny-financed R&D to sales ratios" than similar companies without a merger history. D. Raven- scraft and F. M. Scherer, The Long Run Performance of Mergers and Takeovers, at 44, in M. L. W~edenbaum and K ~ Chilton, Public Policy Toward Corporate Takeovers (~1988~. See also D. Ravenscraft and F: M. Scherer, Mergers, Sell-Offs & Economic Efficzeng~ (~1g87~. These endings do not shed much light on the current controversy because the data result primarily from a con- glomerate restructuring wave that is substantially different from current restructuring phenom- ena. Moreover, the low R&D intensity found at merged plants may simply reflect the finding that industry laggards in R&D are more likely to be involved in restructuring and may not support the hypothesis that restructuring causes a reduction in R&D efforts.
EVENING SESSION, OCTOBER 11, 1989 13 NSF identified 33 firms that were merged into 16 companies, as well as 8 freestanding firms that were involved in LBOs. Interestingly, in the NSF sample the average R&D outlay for the 16 merged firms was $575 million per year, whereas the average R&D outlay for the 8 LBOs was only $75 million per year.40 This statistic supports the view that LBO transactions tend to concentrate in mature, stable industries with "reliable and stable cash flows necessary to amortize the acquisition debt.... LBOs in research industries are rare."4i The market thus does not want to load debt onto it&D-intensive firms because, among other reasons, debt capital is relatively more expensive than equity for R&D applications.42 The NSF study found, however, that these 24 restructured companies reported a 5.3 percent reduction in R&D spending while all other com- panies in the NSF sample reported a 5.4 percent increase.43 A118 of the LBO firms reduced their R&D expenditures, and the aggregate decline in R&D expenditures at these firms was 12.8 percent.44 Indeed, even in the chemical and pharmaceutical industry, where merged companies reported a 5.4 percent increase in spending, the rest of the industry reported a 9.8 percent increase, suggesting that the merged firms were not keeping pace with industry R&D developments after restructuring.45 Why do the NSF results differ from the other findings that indicate no statistically significant change in R&D following restructuring? One In addition, shortly before the date of delivery of this address I received a copy of Hitt, Hoskisson, Ireland, and Harrison, Acquisitive Growth Strategy and Relative R&D Intensity: The Effects of Leverage, Diversification, and Sue (Texas A&M, Baylor, and Clemson universities, May 1989) ("Hits Study"~. The Hitt Study examined 191 mergers of publicly traded firms conducted between 1970 and 1986. Conglomerate acquisitions were the dominant form of transaction. Id. at Table 1. Conglomerate acquisitions are, however, quite different from current restructuring ef- forts, and many of the factors that caused conglomerate acquisitions to fail provide incentives for current restructuring activity. See also, Porter, From CompeictiveAdvantage to Corporate Strayer, 64 Harvard Bus. Rev. 43 (1987) (documenting the failure of conglomerate acquisitions by large corporations during the period 1950-19863. In particular, spin-o~s, bust-ups, and downsizings are all aimed at undoing many of the inefficiencies associated with the conglomerate form. The Hitt Study found that in conglomerate acquisitions, "acquisitive growth, leverage, diversification and size were negatively related to R&D intensity, adjusted for industry R&D intensity." Hitt Study, Abstract. 40Bloch Testimony at 34. (Derived from an NSF Study statistic reporting that the 16 merged firms spent $9.2 billion on R&D and the 8 LBOs spent $600 million.) 41 Merrill Lynch, Leveraged Buyouts in Perspective, at 7 (March 1989~. 42See, e.g., B. Hall, How Is R&D Financed?, Univ. Calif., Berkeley, 1989. 43NSF Report at 3. These changes are measured in constant dollars over the period 1986 to 1987. 44Id.atS. 45Id.
14 CORPORATE RESTRUCTURING answer is that comparing the NSF study to the other studies is a bit like comparing apples to oranges. Aside from the obvious fact that samples and time periods differ, it should be noted that the other studies measure changes in R&D intensity, Epically expressed In terms of an it&D/sales ratio, while the NSF study measures aggregate R&D expenditures. One of the consequences of restructuring is typically a downsizing of the firm's scale as it focuses on more profitable market niches. Thus, in order to compare the NSF results with prior research, it may be necessary to recalculate the NSF findings in terms of research intensity.46 A second potential explanation of the reduction in aggregate expendi- tures is, as the NSF study itself notes, that "firms may simply be eliminating duplication and inefficiency within their R&D programs."47 Here, there is at least some anecdotal evidence supporting the view that postrestructuring reductions in aggregate outlays do not necessarily imply a weakened R&D initiative. For example, in 1986 Exxon spun off its Reliance Electric division to a management-led LBO. Management recognized that Reliance had been spending $30 million a year on overlapping research efforts and proceeded to rationalize its expenditures so that it didn't "have three people working on the same thing."48 This rationalization chopped R&D expenditures by $25 million in 1987, or 17% of the firm's total budget.49 At the same time, however, Reliance increased spending on related productivity tools, such as computer software and custom chips, which may not show up in R&D statistics. As one of Reliance's vice presidents explains, "We are executing projects faster, more efficiently, and experiencing less waste because we have to. Our livelihood depends on it. We're now competitive after the LBO, no question about it."50 Restructuring can also cause changes in the focus of research even if it does not change aggregate expenditures. Japan has built its enviable com- merical position not by concentrating on basic research but by emphasizing superior commercialization. The world's videorecorder, semiconductor, and television markets are all built on basic U.S. research and Japanese commercialization. Given a choice between being a hero for doing prof- itless basic research or, at the margin, moving resources more vigorously into profitable commercialization, it may well make sense to reallocate resources toward the commercialization end of the R&D spectrum. 46Accord, B. Hall, Leveraged Buyouts, Corporate Debt, and R&D Investment: Is There Any Con- nect~on? Univ. of Calif., Berkeley (Work in Progress, Version of Sept. 1989, at 7~. 47NSF Study at 4. 48A. Ramirez, What LBOs Really Do to R&D Spending, Fortune, March 13, 1989, at 98. 49Id. 50Id. (quoting Peter lLivitse, vice president, Reliance Electric).
EVENING SESSION, OCTOBER 11, 19~9 15 Several companies have recently reached just that decision. Xerox's Palo Alto Research Center spawned several successful innovations that have failed to earn Xerox a fraction of what they could have. Xerox's inability to capitalize on its development of Ethernmet, of the laser printer, and of the icon-based operating system popularized by Apple Computer is perhaps the most poignant example of a company that did its "R" brilliantly only to watch the profits slip away as a result of poor "D."5i ~ minimize the chance of this happening again, Xerox has taken strong steps to assure that the research for which it is paying develops products that return value to the corporation and to its shareholders. Success in the marketplace thus requires a balance between "R" on the one hand and "D" on the other. Without great research, there is nothing to commercialize. Without great commercialization, you never earn the fruits of your research. It is my subjective assessment, based on recent developments in Japan and elsewhere, that a shift in our emphasis toward commercialization might be the most profitable change American industry could make in allocating its R&D. If that is a direction in which restructuring is driving America's R&D efforts, it's hard to conclude that it's all for the bad.52 Evidence of reduced or dramatically changed R&D expenditures there- fore does not, in and of itself, suggest a weakening of a company's com- mitment to R&D. Nor does it necessarily suggest a reduction in the effec- t~veness of a company's R&D program. Instead, what we need to measure is how "smart" we are in spending our R&D dollars, because the elimi- nation of a "dumb" R&D dollar resulting from waste, duplication, or bad planning means something quite different than the elimination of a "smart" R&D dollar that reflects a potentially profitable gamble on the scientific unknown. THE STOCK MARKET AND R&D EXPENDITURES But how do we tell "smart" R&D from "dumb" R&D? The short answer Is that there is no easy answer. Research and development is a gamble on the unknown. It will always be impossible to know whether two 51Pitta, Bean Counters Invade Ivory Tower, Forbes, Sept. 18, 1989, at 198. 52Accord, Reich, 17'e Quiet Pads to Technological Preemmence, Scientific American, Oct. 1989, at 41. ("If the U.S. is to regain its technological prominence, it must improve the capacity of Americans to use technology. This quiet path back to competitiveness depends less on ambitious government R&cD projects . . . than on improving the way by which technological insights whenever they may be discovered around the globe are transformed by American workers into high quality products") This may be one of the few points regarding economic policy on which Reich and I agree.
16 CORPORATE RESTRUCTURING guys wearing white coats in a Topeka lab will, if left alone for a decade, come up with cold fusion or superconductivity. While there is no easy answer to this question, many critics of restruc- turing and of the stock market would be quick to conclude that the stock market is incapable of judging the value of R&D projects and invariably penalizes companies that are committed to substantial, long-term R&D expenditures.53 Therefore, however one judges R&D, one should surely ignore the stock market's valuation, at least according to these critics. But is this highly negative view of the stock market's response to R&D supported by the evidence? 1b pose the issue most starkly, let me begin by asking a question: Which company does the stock market value more highly, Merck, a research-intensive pharmaceutical firm with 1988 sales of $5.9 billion, or General Motors, the automotive giant with 1988 sales of $110 billion that are 19 times as large as Merck's sales? Believe it or not, as of December 31, 1988, the stock market valued Merck's stock at $26.433 billion, about $400 million more than General Motors stock which was trading at an aggregate value of $26.027 billion.54 But how can that be? After all, Merck is one of the most R&D- intensive companies in one of the most it&D-intensive industries in the world. In 1988 Merck spent $669 million on R&D: that's 11.3 percent of its sales, 34.9 percent of its profits, and $15,962 per employee.55 These expenditures are for R&D projects that are wildly expensive,56 more likely to fail than to succeed, and certain not to yield revenues in the United States for about 8 to 10 years from inception.57 Yet Merck's stock trades at a price-earnings ratio of 23, more than triple the multiple of 7 accorded GM's shares.58 If the critics are right, and if the stock market is simply too impatient or myopic to wait for the payoff from R&D, then Merck's shares should be trading at an aggregate value far below General Motor's. But Merck's shares aren't trading below General Motors, and that fact takes at least some of the wind out of the sails of market critics. While this simple comparison of Merck and General Motors is not enough to sustain any broad hypothesis about stock market behavior, it is enough to force critics of takeovers to take pause and to reconsider some 53See, e.g., the materials cited in notes 1 and 2, supra. 54The World's loo Largest Public Companies, Wall SL J., Sept. 22, 1989, at R14. 55Rk£D Scoreboard, Business Week, Innovation 1989, at 198. 56The average cost of developing a new dmg (new molecular entity) through approval By the Food and Drug Administration is $125 million measured in 1986 dollars. Pharmaceutical Man- ufacturers Association, Facts at a Glance, 13 (1989~. s7Id. at 15, 18. 58New York Stock Exchange Composite Transactions, Wall St. J., Oct. 9, 1989, at C4.
EVENING SESSION, OCTOBER 11, 1989 17 of their prejudices. Apparently, the relationship between R&D expenditure and stock price valuation is much more sophisticated than a simple "increase your R&D and the market will knock your stock price down" correlation. Indeed, I suspect that the easiest way for Merck to slash its stock price would be for it to cut back dramatically on its R&D. Further support for the view that the stock market does not invariably penalize increased R&D expenditures is found in a recent study by Randall Woolridge, who examined the stock price effects of announced changes in R&D budgets.59 Woolridge found that in the two days following an- nouncement of increased R&D budgets by 45 companies the value of those companies' shares increased by an average of 1.2 percent, net of overall market changes.60 For example, after du Pont announced on August 12, 1983, that it would spend an additional $100 million on R&D to improve automotive and industrial coatings, its stock price rose 2.54 percent. The market did not penalize du Pont with a decline. Similarly, a study by SEC economists found that a sample of 62 R&D announcements were associated with significantly positive stock price returns.6i A study of the stock price effect of 658 announcements of changes in planned corporate capital expenditures also found that announcements of increased capital expenditures are correlated with significantly positive stock price effects while reductions in capital expenditures are correlated with declines.62 No doubt, these average statistics mask significant mistakes on both sides of the R&D fence. For example, when Federal Express in 1984 announced its plan to spend $1.2 billion over ten years to develop its Zapmail service, the stock market was as enthusiastic as Federal Express' management and the company's shares rose 2.27 percent63 Time proved that Federal Express and the stock market were both wrong about the promise of Zapmailbut the market's initial response was hardly hostile to management's long-term and expensive technology gamble. Similarly, Genentech was able to raise $40 million in its initial public 59J.R. Woolridge, Carnpetinve Decline: Is a Myopic Stock Market to Blame?, 1 J. Applied Coup. Fin. 26 (1988). 60Id. at 31. 61G. Jarrell, K. Lehn, and W. Marr, Institun~1 Ownash~ Tender Offers, and Long Term Invest- nzents, Office of the Chief Economist, Securities and Exchange Commission (April 19, 1985~. 62J. McConnell and C. Muscarella, Corporam CapualExpendinueDec=ans and the Market Vane of the Firm, 14 J. Fin. Econ. 399 (1985~. The sample in this study had only eight announcements of changes in R&D expenditures and did not generate statistically significant results for this subsample. 63Woolridge, supra n. 59, at 33.
18 CORPORATE RESTRUCTURING offering at a time when it had no meaningful revenues much less profits.64 All Genentech had was the dream that one day it might be able to develop useful products that might gain FDA approval and that might earn a profit for investors. The dream was, however, a distant one viewed from Genentech's initial public offering and, as events have subsequently proved, the gamble has not worked out as well as many scientists and investors had hoped.65 Thus, there are several examples of situations In which the market has been willing to reward high-intensi~ R&D companies with rich stock-price multiples and start-up funding. However, these are not the situations in which friction is likely to arise. Friction arises when management wants to pursue an R&D project or capital expenditure plan that the stock market won't support. In that situation managers often scream that the stock market is crazy and that it is only because of the market that they can't engage in valuable new investment. But when management and the market disagree is it invariably true that the market Is crazy and management wise? I think not, and there are several examples of projects in which management was willing foolishly to spend hundreds of millions of dollars despite the market's clear warnings to the contrary. My personal favorite example of a management that wouldn't listen to or learn from the market is the management of Unocal and its devotion to a shale oil conversion project that could be feasible only at sky-high oil prices. What Unocal's management overlooked, however, was that before the price of oil could rise to $50 or $60 a barrel, there would be so many other alternate sources of energy and conservation called on- line that the demand for high-priced shale oil might be problematic even if the technology was feasible. Neither the market nor I believed that this project made any sense. Nonetheless, Unocal's management pumped enough money into this project both taxpayer dollars and shareholder dollars that its expenditures substantially depressed its stock price and became a major magnet for Boone Pickens's attempted takeover of the company.66 64See Investment Dealers Digest, Oct. 21,1980, at 10. 65See, e.g., C. Bartlett, Jr., Special Sedations, Forbes, June 26, 1989, at 266; R. Stern and P. Bornstein, winy New Issues Are Lousy Investments, Forbes, Oct. 2, 1985, at 152, 154; Waiting for a Payoffin Biotech Stocks, Fortune, Nov. 26, 1984, at 185, 186. 66See, e.g., Coming Up Dry: Unocal Straggles On With Attempt to Get Crude Odirom Shale, Wall St. J., May 14, 1989, at 1; Synfuels Corp. Issues New Grant, Defies DOE; $550 Million Approvedfor Od Shale Profits, Wash. Post, Oct. 17, 1985, at A21; Lawsuit Is Fded to Void Accords for Unocal Plant, Wall St. J., June 6, 1986, at A9. In addition, McConnell & Muscarella, supra, n. 62, ob- seme that in the late 1970s announcements of increased expenditures on oil and gas exploration were correlated with stock price declines. The market was apparently signalling that exploration expenditures were not profitable given the outlook for petroleum prices and the availablity of
EVENING SESSION, OCTOBER 11, 1989 19 No doubt, there have also been and will continue to be situations in which managements would have supported successful R&D projects but for the market's skepticism. Similarly, there are probably R&D projects the market would have supported, but for management's lack of vision or courage in proposing them. However, my goal is not to prove that markets are always right and that managements are always wrong. Instead, my goal is simply to get across the message that markets are not always wrong and managements are not always right. In particular, managements could often do themselves and their companies a great service if they just took some time to appreciate why the markets value some forms of long-term investment and penalize others. That simple step of market appreciation could probably work wonders for R&D budgeting, capital budgeting, and several other critical corporate decisions.67 CONCLUSION In sum, America needs to do more, much more, to strengthen and preserve its critical position in R&D. Stifling restructuring is not, however, the answer to America's R&D problem. Even if we brought restructuring to a screeching halt, our capital costs would remain far above our international competitors'. Moreover, most of the restructuring we would prevent would involve companies that do little or no R&D, and many of the companies that would be restructured are R&D laggards who spend less on innovation than their industry peers. What then are we to do? A two-step program appears necessary. First, we should take strong and immediate steps to reduce the cost of capital for R&D projects. Most fundamentally, America needs to increase its savings rate so that more domestic capital is available for R&D and for other investment projects. On a more targeted basis, R&D tax credits and reduced capital gains tax rates can also help lower the effective cost of capital for R&D projects. Second, we must focus more of our efforts on commercialization. Japan is eating our lunch not only because its capital costs are lower, but because it has mastered the art of commercialization. All too often the United States stands at the cutting edge of R&D only to watch Japanese and other foreign firms earn the lion's share of the profits. This Is not reserves that could be purchased at lower cost. Ibis divergence between market valuations and management expenditure plans can help explain much of the takeover activity in the oil and gas industry in the eartr 1980s. By the way, in this situation it appears that the market's valuation did a better pb than management's plans. 675eegenera~y, A. Rappaport, Creating Sh~hoh~ Cage (~19gT).
20 CORPORATE RESTRUCTURING smart R&D. This is charity R&D that does a disservice to the corporation, to its stockholders, and to its scientists who won't be able to do future R&D for that company unless it starts earning some profits from its past R&D efforts. Neither of these two steps is easy. Neither is dramatic. Neither will satisfy the critics of restructuring who want to stop change with any argument they can find. Either of these two steps will, however, help restore America's com- petitive R&D edge. If that's what we really care about, then that's clearly the direction in which we should go. DR. PRESS: This symposium is designed to present both sides of the issue. For those of you who think that the keynote speaker is ove~whelm- ingly biased toward one side of this issue, let me assure you that tomorrow you will hear the other side from our luncheon keynote speaker and during the course of the afternoon. It is rather late to take questions and since we have an early start tomorrow, may I ask you to reserve your questions for the discussion period tomorrow. And since this was such an erudite talk, full of statistics, I think you need a little bit of time, those of you who want to ask your questions, to prepare adequately.
