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.
Appendix A Industrial Research and Corporate Restructuring: An Overview of Some Issues Kenneth Flammi On October 12, a conference convened by the Academy Industry Program of the National Research Council met to consider the significance for industrial research of recent trends in corporate restructuring activity. This brief overnew essay was intended to raise some questions for that discussion.2 Are there grounds for concern with respect to the recent behavior of American industrial research and development activity? What is the avail- able evidence on the relationship between recent corporate restructuring activity and industry R&D? What are the positive and negative perspectives on how corporate restructuring affects industrial R&D? Finally, can any tentative assessments of its impact be ventured? IS THERE REASON FOR CONCERN? Since about 1987, it has been clear to both business analysts and public officials that industrial research in the United States has slowed considerably after almost a decade of rapid growth. Figure 1 shows both absolute levels ~ Consultant, Academy Industry Program of the National Research Council. The author is a Senior Fellow at The Brookings Institution, 1775 Massachusetts Ave., N.W, Washington, DC 20036. The opinions expressed in this essay are the authorts alone, and should not be ascribed to officers or staff members of the National Research Council or The Brookings Institution. 2 Without implicating her in my errors, I thank Margaret Blair for her very helpful comments. 121
122 60000 ~ ° RED Funcls, 1982 $ CORPORATE RESTRUCTURING % Growth Rate, Inflation-Adjusted R&D Funds 15 ~ Co R&D regrowth Rate ~ 50000 / \ _< / 20000 1977 1978 1879 1880 1981 1982 1983 1984 1985 16386 1987 FIGURE 1 Company-funded R&D expenditures. 10 5 o and growth rates for company-funded research and development funds as measured by the National Science Foundation (NSF).3 For the period from 1977 to 1985, growth rates in inflation-adjusted company funds for R&D did not fall below 6 percent Since 1986, however, the growth rate has fallen below half of that minimum, and forecasts suggest further decline in 19~.4 Why that R&D spending growth rate has fallen is less clear. One way of gaining insight is to examine the it&D-to-sales ratio, a measure of the research intensity of a company (or industry). Figure 2 shows the recent historical behavior of research intensity, as measured by company-funded 3The NSF figures on company-funded R&D used here include all company funds expended on the NSFs definition of R&D (which differs from that used in FASB-based accounting reports, Business Week, corporate annual reports, and many internal company cost accounting systems3 on projects undertaken within the company's U.S. facilities. Excluded are company-funded R&D conducted by external organizations, company-funded R&D undertaken within foreign facilities, and federally funded R&D performed within domestic facilities. The NSF data used in this overview are the most recent data available, unpublished statis- tics furnished to the author by the NSF. The precise firms used to estimate these figures, and their industrial classification, are based on the construction of a new panel for the NSF survey in 1987. Earlier data released by the NSF are based on a sample put together for 1981. For that reason, these data may not exactly match earlier industrial R&D data produced by the NSF. 4 See National Science Foundation, Science Resource Highlights, "Modest Increase in Company R&D Funding Estimated for 1989, " NSF 89-310, June 30, 1989; Gene Koretz, "Business Bilks a Better R&D Game Than It Plays," Business Week, August 21, 1989, p. 20.
INDUSTRIAL RESEARCH R&D as % Sales i All Manufacturlng ~ Chemicals -x- Petroleum -v- Computers -A Electrical ~ Motor Vehicles ~ Aerospace -a- Instruments 10 5 123 v V 1977 1978 1979 1980 1981 1982 1983 1884 1985 1986 1987 FIGURE 2 Research intensity of it&D-performing manufacturing companies. R&D per dollar of sales, in key industries. Research intensifier generally grew during the late 1970s and early 1980s, but with a few exceptions the computer and chemical industries, in particular stabilized or even dropped slightly in the mid-1980s. The percentage rate of change in inflation-adjusted R&D spending can be broken down into two components~hange in research intensity, and change in inflation-adjusted sales.5 Figure 3 shows that the declines in 1986 and 1987 in inflation-adjusted R&D did not reflect a continuous decline in either research intensity or inflation-adjusted sales of American research-performing firms in manufacturing.6 In the face of a declining sales base in 1985 and 1986, the relative size of the research effort of American it&D-performing companies increased. When sales picked up in 1987, the intensity of these companies' research efforts decreased. Breaking out growth in inflation-adjusted company R&D funds by major it&D-performing industry, a similarly heterogeneous picture emerges 5If R is R&D, 5 is sales, P is the price index used for inflation adjustment, and be ~ ~ denotes taking the natural logarithm of some argument, we have: In (R/P) = ~ (R/~ + h (S/P) Take fimt differences on both sides of this equation,note that fimt differences of natural logs are approximately the percentage rate of change, and we have the procedure described here. The GNP implicit price deflator is used for inflation adjustment. 6Manufacturing firms were estimated to account for about 9l percent of U.S. company-funded R&D in 1987.
124 .15 .05 -.2 -.25 CORPORATE RESIRUCTlJRlNG >>>-.0>5 -.1 1 = -.15 1 1 -.15 i ]:L Cbe Rate of Change Due to Research Intensity Due to Inflation-Adjusted Sales 1978 1979 1980 1981 1982 1883 1884 1985 1988 1987 .15 .1 -.2 -.25 FIGURE 3 Decomposition of change in inflation-adjusted R&D; U.S. manuf~ctunng, company funds. 12000 9000 6000 3000 o Co R&D Funds, 1982 S . _ HAWAII Manufacturing Chemicals -x- Computers -v- Electrical _ -A Motor Vehicles Aerospace Instruments 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 55000 50000 45000 40000 35000 30000 25000 FIGURE 4 Growth in company R&D funds, by major sector. Left axis: all manufactunng; right axis: other sectom (see Figure 4~. While there was little growth in such major sectors as computers, motor vehicles, aerospace, and instruments, in 1987 chemicals and electrical equipment grew respectably. Is there some deeper logic to the overall sluggish growth that emerges from the aggregation of these seemingly diverse industry experiences? The importance of the issue seems obvious enough: R&D investment is central
INDUSTPL4L RESEARCH 125 to the long-term competitiveness of high-technology industries, and those high-technology industries are at the heart of any strategy to increase American living standards. The fact that the National Science Foundation had received numerous telephone inquiries concerning the impact of a wave of corporate restructurings, which became highly visible around 1983, apparently led it to examine the impact of restructuring on R&D. THE RISE OF THE LBO Corporate restructuring is a vague term which encompasses many types of activities. It is often used to describe merger and acquisition activity. In the United States, a surge in corporate takeovers in the 1980s was especially notable; this activity coincided with the rapid diffusion of a major financial innovation, the so-called leveraged buyout (LBO). Within the science and technology community, however, other events, like the divestiture of the regional operating companies by AT&T in 1984, are also included in discussions of the potential impacts of corporate reorganization on research investment in the United States. The creation of the AT&T Bell Laboratories and Bell Communications Research from the predivestiture Bell Labs; the transfer of RCA's Sarnoff Laboratories to SRI International in the wake of General Electric's acquisition of RCA (and sell-off of its consumer electronics business); and the reorganization of research at G.E.'s Schenectady central research facility are examples of possibly significant "corporate restructuring" which are not directly linked to the high-profile leveraged buyout. The rise of the leveraged buyout was a relatively recent development on the American financial scene. The use of so-called "junk bonds" (low- quality bonds) to finance takeover activity was considered a significant innovation in financial markets when it first became widespread in the 1980s. LBO activity appears to have been virtually nonexistent prior to the early 1980s. Data cited by Lichtenberg and Siegel show LBOs rising from under 4 percent of the value of all merger and acquisition deals in 1981, to about 27 percent in 1986.7 What defines a leveraged buyout? It is helpful to think of "leveraged corporate takeovers" as a subset of all acquisition activity, in which a group of investors utilizing significant borrowed funds buys out a controlling in- terest in a publicly held corporation. "Leveraged buyouts" are generally understood to be the subset of leveraged corporate takeovers in which a group of investors buys out all the outstanding equity, and takes a company private, while making use of substantial borrowed funds. What constitutes 7Frank R. Lichtenberg and Donald Siegel, "The Effects of Leveraged Buyouts on Productivity and Related Aspects of Firm Behavior," NBER Working Paper 30=, June 1989, liable 1.
126 CORPORATE RESTRUCTURING "substantial" or "significant" use of borrowed funds evades precise defi- nition; the borrowing is generally considered to result in a "significant" increase in a company's debt-to-equity ratio. Hostile takeovers are con- ducted over the objections of current management, while friendly takeovers involve the acquiescence of managers. LBOs, like other takeovers, can be hostile or friendly. "Management buyouts" are LBOs in which current man- agement participates in the acquiring investor group; such deals presumably are friendly by definition. EXPIAINING THE SLOWDOWN IN R&D GROWTH More than telephone calls from the curious pointed the NSFs attention to the LBO phenomenon. The NSFs annual surveys of R&D directors at major corporations, and comments received in response to its spending survey, had suggested that research directors, as a group, often perceived the impact of corporate restructuring to be a significant factor in cutting back research at their own and other companies. In 1988, the NSF initiated an attempt to identify the impact of re- structuring on companies responding to its annual expenditure survey.8 The 200 largest it&D-performing companies (accounting for about 90 percent of company funds) were divided into two groups. One group consisted of firms uninvolved in any sort of merger activity. Another group consisted of 33 large firms which had merged into 16 companies, and 8 companies involved in "LBO and other restructurings" (which included leveraged buy- outs, as well as stock buybacks and other restructurings to defend against leveraged buyouts) over the 1984 1986 period. Companies in which R&D- performing divisions had been sold off were deliberately excluded by NSF from its sample. The results seemed quite striking.9 In the 16 "merged" companies, R&D expenditure declined by 4.7 percent over 19~1987, while the 8 "LBO/other restructuring" companies had R&D declining by 12.8 percent. The remaining 80 top-200 companies had seen R&D expenditures increase by 5.4 percent. While the NSF did not publish sales data which would allow one to look at research intensity in the different sets of firms, such data apparently will be made available shorter. In thinking about these results, two points are worth bearing in mind. First, after inflation, the "normal" firms in the NSF sample had R&D growth rates which remained substantially below the pre-1986 norm (up 2 Amy description here is based on various unpublished memos from the NSF. 9 See National Science Foundation, "An Assessment of the Impact of Recent Leveraged Buyouts and Other Restructurings on Industrial Research and Development Expenditures," February 1, 1989.
INDUS77~1AL RESEARCH Billions 1982 S 30 All Other Countries 127 US Only 55 ~ France WOW Germany -x-Japan -v-UK US l 25 20 15 o 1977 1978 1979 1980 1981 FIGURE 5 Industry funds for R&D. 50 / 1982 1983 1984 1985 1986 1987 45 40 35 30 25 percent after inflation). Thus, even if there is a link between restructur- ing and R&D, the slowdown in the nonrestructured firms remains to be explained, and the attendant concerns persist. Second, growth in foreign countries of inflation-adjusted company R&D funds was not totally dissimilar to that in the United States. Figure 5 charts NSF data on industry funds for R&D (a slightly different definition than that used earlier, since it includes funds for R&D performed outside of the company itself) in the United States and five other industrialized countries. It is interesting, and possibly important, to note that like the United States, Japan and West Germany seem to have experienced a slowdown in industry-funded R&D growth after 1985.~° This of course suggests that general business conditions in globalized high-technology industries, rather than country-specific phenomena (like the U.S. LBO wave) are the root cause of the slowdown. At least two other large studies have attempted to look at the impact of selected restructurings on corporate R&D investment in the 1980s, and 10 These data are computed from numbers found in National Science Foundation, Internation`;d Science and Technology Data Update: 1988, Special Report NSF 89-307, December 1988, pp. 4, 16. Foreign industry R&D funds have been converted to U.S. dollars at purchasing power parity exchange rates.
128 CORPORATE RESTRUCTURING both studies contained at least some findings in apparent contradiction with the NSF results. A study by Bronwyn Hall, though primarily focused on models of corporate merger and acquisition behavior, tabulates data on the R&D intensity of merged firms one year before and after their merger, and compares it with that of firms not involved in merger activity within the same industry. Hall finds no obviously significant pattern of change in R&D intensity, either before and after merger, or between merged and nonmerging companies. The data cover the period 1976 to 1985 (which would cover mergers occurring in 1984), but Hall has stated that extending the sample to 1987 (covering mergers through 1986) leaves the conclusion unchanged.~3 The other major set of studies considered here were undertaken by Frank R. Lichtenberg and Donald M. Siegel. The first Lichtenberg and Siegel study examined the impact of ownership changes on employment in firms' production and non-production establishments,~4 over the 1977 to 1982 period. As part of that analysis, they show that ownership changes over that period appeared to have no significant impact on R&D employment in companies' central offices. The second Lichtenberg and Siegel study directly tackles the issue of the impact of LBOs on company R&D expenditure by examining changes over time in the average research intensity of 43 firms involved in LBOs, and comparing the LBO firms' research-to-sales ratio with the average research intensity of all it&D-performing firms.~5 The 11 We exclude from this discussion Kohlberg Kravis Roberts ~ Co., "Presentation on Leveraged Buy-Outs," January 1989, on the grounds that the data presented there which include projec- tions rather than actual historical numbe~are not comparable with the other studies discussed. See William F. Long and David J. Ravenscraft, "The Record of LBO Performance," July 1989. We also have omitted at least two other unpublished studies: Abbie Smith, "Corporate Own- ership Structure and Performance: The Case of Management Buyouts," University of Chicago, January 1989; Michael A. Hitt, Robert E. Hoskisson, R. Duane Ireland, and Jeremy S. Harrison, "Acquisitive Growth Strategy and Relative R&D Intensity: The Effects of Leverage, Diversifica- tion, and Size," May 1989. Finally, F.M. Scherer and David J. Ravenscraft, Mergers, Sell-o~s, and Economic Efficiency (Washington, D.C: Brookings Institution), 1987, though analyzing mergers of the late 1960s and early 1970s which differed in some important respects from the buyouts of the 1980s, contains a detailed analysis of the impact of those mergers on R&D. i2Bronwyn H. Hall, "The Effect of TakeoverActivityon Corporate Research and Development," in Alan Auerbach, Ed., Corporate Talceovers: Causes and Consequences (Chicago: University of Chicago Press), 1988. i3Bronwyn Hall, Testimony on "Corporate Restructuring and Its Effect on R&D," Hearings before the Science, Research, and Technology Subcommittee, House Committee on Science, Space, and Technology, July 13, 1989, p. 2. i4See Frank R. Lichtenberg and Donald Siegel, "The Effect of Takeovers on the Employment and Wages of Central-Office and Other Personnel," Bureau of the Census, Center for Economic Studies, Discussion Paper CES 89-3, June 1989. i5Frank R. Lichtenberg and Donald Siegel, "The Effects of Leveraged Buyouts on Productivity and Related Aspects of Film Behavior," NBER Working Paper 3022, June 1989.
INDUSTRIAL RESEARCH 129 study found that average research intensity increased in the LBO firms over time, and even showed some increase relative to research intensity in all firms. The second Lichtenberg-Siegel study is particularly interesting because it made use of data which included as a subset the same 200 largest R&D- performing United States firms used in the NSF study. While the NSF study used data on only 1 LBO in its "LBO/other restructuring" group (the remaining 7 firms fell in the "other" category, and largely were made up of firms undertaking stock buy-backs to defend against takeover bids), Lichtenberg found 43 firms undergoing LBOs within their sample. This difference is attributable to the much larger size of the Lichtenberg sample, the different definitions used to identify an LBO, the elimination from the NSF sample of companies with divisional sell-offs, and differing time periods over which the LBOs were counted (1978 to 1986, versus 1984 to 1986~. The simple fact is that LBOs as a group don't do much R&D. Note Bronwyn Hall's observation that the LBOs in her sample generally did much less R&D than the average manufacturing firm: "the recent increase in acquisition activity due to leveraged buyouts or other such private purchases is more or less orthogonal to the R&D activity in manufacturing. Even if all such purchases resulted in the complete cessation of R&D activity by the firm, this would amount to only around 500 million 1982 dollars annually compared to expenditures on R&D by the manufacturing sector of approximately 40 billion 1982 dollars annually." Other information also points to the lack of much overlap between LBOs and R&D activity.~7 How can these studies be squared with the seemingly contradictory NSF results? The first Lichtenberg-Siegel study is concerned with the earlier 1977-1982 period, and thus preceded the recent wave of leveraged takeovers. Furthermore, central office data on R&D employment include only central Et&D laboratories, which accounted for a little under half of all industrial R&D employment in 1982. One can think of admittedly speculative scenarios where the engineering and research staff at central laboratories might be more insulated from declines in R&D spending than their brethren engineers in the field. When R&D budgets slow, R&D personnel stationed out in production and development facilities might be reassigned to other engineering tasks, while R&D directors retain staff at the central R&D facility. i6Hall, in Auerbach, Ed., 1988, p. 82. i7Hall, House Subcommittee on Science, Research, and Technology, Jutr 13, 1989. The "LBO/ other restructuring" group in the NSF study also did only a small amount of R&D in the aggre- gate.
130 CORPORATE RESTRUCTURING Both the Hall and Lichtenberg-Siegel papers findings on research intensity (A&D-sales ratios) are not necessarily inconsistent with NSF's observation of an absolute decline in R&D in restructured firms. If sales fall as a consequence of an LBO, an increase in research intensity may accompany a decline in absolute levels of R&D. Since LBOs are thought to frequently be mothated by a desire to downsize a firm, to cut unprofitable lines of business and associated investments, this is not entirely unlikely. In short, the NSF appears to have selected a particular group of firms including both LBO/other restructurings, and more conventional mergers in which absolute R&D spending in the mid-1980s rose much less than average. Hall and Lichtenberg-Siegel, looking at somewhat different kinds of restructurings in slightly different time periods, find little difference in research intensity between restructured and other firms. The apparent contradiction is reconcilable if sales declined at least as rapidly as R&D in the NSFs restructured firms. Whether this is the explanation will be known when the NSF releases aggregate sales data for the groups of firms in its sample. Despite the ambiguity in the empirical evidence, opposing points of view on the consequences of LBO activity seem to concur in predicting diminished R&D activity in restructured firms. We next turn to a brief discussion of these opposing viewpoints. TWO POINTS OF VIEW ON LEVERAGED TAKEOVERS The rise of the LBO in the 1980s attracted the attention of researchers, and led to explanations for this activity in terms of a "market for corporate control." This market is the aggregate of transactions undertaken by shareholders to create incentive and discipline mechanisms that will induce corporate managers to act in their interest. The empirical consensus seems to be that takeover activity generally increases the market value of the stock of companies involved. One's view of whether this is socially desirable boils down to judgments about the source of that increased value. If the gain in market value results from the more efficient or productive utilization of a firm's assets or cash flows, then that ought to be counted in the benefit column and offset whatever costs involving R&D resources might be created. On the other hand, if the increase in the value of equity reflects a temporary misjudgment by the market, or comes at the expense of other groups with an interest in the firmexisting creditors who have seen their bonds or loans grow riskier without any compensating increase in return, or employees who have seen their wages and benefits drop when their contracts are renegotiated- then any negative impacts on R&D are not offset by gains for society as a whole.
INDUSTRIAL RESEARCH RESTRUCTURING AND R&D: A NEGATIVE VIEW 131 One pessimistic assessment of corporate restructuring's effects on re- search rests specifically on its assumed impact on the financial structure of the firm. The emphasis is on the "1:' in LBO. Restructuring is assumed to lead to an increase in a corporation's debt-to-equity ratio, and a continuing need to pay out relatively large amounts in interest to service the increased debt. Interest payments on debt are fixed charges, which must be paid regardless of business conditions. R&D expenditure, on the other hand, can be deferred in extremes. Hence when sales are down, or costs are up more than anticipated, R&D like other deferrable expenditures is cut back. Increasing the debt load shouldered by the firm, it is argued, makes firms worry more about having enough cash flow to meet interest charges in times of adversity, and therefore concentrate more on short-term projects generating quick paybacks than would have been the case with a less leveraged financial structure. ~ put this argument in perspective, it is helpful to sketch out how a firm's optimal financial structure is determined.~9 Because interest pay- ments are deductible from taxable income, and dividends are not, the tax system tips a company toward debt as a source of funds. However, in an uncertain world, increasing debt makes it more difficult to cope with unforeseen declines in earnings. In the worst case, the firm might be forced into bankruptcy. Because the process of bankruptcy is itself costly, the value of the firm's assets to creditors is reduced by the cost of their liquidation. Since the probability of bankruptcy depends on how leveraged the firm is, the real costs of bankruptcy make it desirable not to so increase leverage as to make bankruptcy highly likely. In equilibrium, "the tax advantages of an extra dollar of debt would be just offset by the additional bankruptcy and moral hazard costs implied by the replacement of a dollar of equity with debt. Since the bankruptcy costs will vary by firm for many reasons, particularly the variability of the firm's earnings, the equilibrium debt-equity ratio will vary by firm, and the firm with more variable earnings will choose a lower debt~quity ratio."20 (The i8This argument is laid out in Robert R. Miller, "The Impact of Merger and Acquisition Activity on Research and Development in U.S.-Based Companies," contractor report submitted to the Office of Technology Assessment, June 1989, referenced in Julie Fox Gorte, "Statement Before the Subcommittee on Science, Research, and Technology," House Committee on Science, Space, and Technology, July 13, 1989, p. 3. i9A clear and accessible exposition of the arguments sketched out here may be found in Roger H. Gordon and Burton G. Malkiel, "Corporation Finance," in Henry J. Aaron and Joseph A. Pechman, Eds., How Taxes Affect Economic Behavior (Washington, D.C: Lee Brookings Insti- tution), 1981. 20 Gordon and Malkiel, p. 149.
132 CORPORATE RESTRUCTURING moral hazard cost is the cost created by the possibility of shareholders or managers imposing uncompensated increases in risk on creditors holding debt instruments.) Where does the impact of increased financial leverage on R&D fit into this framework? R&D investments are inherently highly uncertain, and a high-technology firm with many such projects and relatively volatile earnings would normally be expected to choose a lower debt-to-equity ratio. If a high debt-to-equity ratio were suddenly imposed on such a firm, one adjustment in moving back toward equilibrium would be to shed some of the riskier investment projects like R&D, and particularly the most uncertain and long-term R&D projectsto reduce the variability of earnings. This is the essence of the argument made by critics of the impact of restructuring on R&D. But is this cutting back of R&D in the restructured firm necessarily bad? If one firm dumps an R&D project that is viewed by the market as worthwhile, some other firm is free to invest in it. One firm's cutback in R&D, brought about by having a higher debt-to-equity ratio imposed, ought be matched by another firm's willingness to jump into the profitable vacuum if the project was truly worthwhile. Of course, this may ignore other relevant costs. There may be firm- specific sunk costs associated with the start-up of an R&D project. These costs presumably must be duplicated if a terminated R&D project is started up elsewhere, and along with the opportunity costs incurred by delaying an innovation, represent a genuine social cost to the nation. And shuffling R&D resources among companies may itself involve real costs to society, costs divided among companies and researchers. Critics of corporate restructuring also argue that other costs, unrelated to R&D, are imposed on society by excessive restructuring activity. Sub- stantial resources may be absorbed by legal and financial maneuvers, with no real return to society.2i The increased leverage created by restructurings may increase the business failure rate during a downturn, creating macroe- conomic costs external to the firm. If the cash flows freed by cancelled investment projects are not reinvested productively (for example, if they are used to increase consumption), the future standard of living will fall. And some restructurings may increase the riskiness of existing debt, which lowers its value without compensating creditors for the increased risk. The moral hazard costs may create distortions in investment incentives, and a long-term efficiency cost to society. 2iSome argue that the lawyers and financiers involved would otherwise be involved in other, equally unproductive activity, so that little net cost to society is created.
INDUSTRIAL RESEARCH BUYOUTS AND MANAGEMENT REFORM: THE POSITIVE VIEW 133 Another view focuses on the "h" in hostile takeover. From this perspec- tive, the recent upsurge in hostile takeover activity particularly leveraged buyouts is beneficial to the United States economy. This view takes as its point of departure the so-called "free cash flow" theory of Michael Jensen.22 Briefly, Jensen notes that corporate managers are supposed to function as the agents of shareholders, implementing shareholder interests in mammiz- ing the value of the firm. But conflicts of interest between managers and stockholders may arise. "Agency costs" the costs of monitoring managers and devising incentives for them to act in the interest of stockholders, and the losses inflicted on stockholders by such conflicts of interest may be great. The possibility of a buyout or hostile takeover, from this point of view, may sense as a discipline and deterrent for managers who get too far out of line with the interests of stockholders. Such corporate restructuring, or its threat, serves the positive social function of reducing agency costs. Jensen argues that some firms may face the problem of free cash flow, "cash flow in excess of that required to fund all of a firm's projects that have positive net present values when discounted at the relevant cost of capital. Such free cash flow must be paid out to shareholders if the firm is to be efficient and to maximize value for shareholders."23 Managers, on the other hand, may prefer to use those resources in ways designed to increase their influence and compensation, by plowing funds into investment projects in order to increase the company's size or growth rate. It has been argued that such free cash flow and hostile takeover activity should increase when high real interest rates make many projects less attractive investment opportunities, or as industries mature and the most profitable projects are exhausted.24 Increased debt, in addition to making the restructuring possible, creates a fixed charge against future cash flows. If debt is exchanged for equity, argues Jensen, this has the effect of forcing managers to promise increased future cash flows to stockholders. R&D investments are presumably among some of the insufflc~ently profitable projects that managers may be pursuing at stockholder expense. Therefore, one consequence of corporate restructuring designed to over- throw incumbent management and divert free cash flow to stockholders would be to expect cutbacks in R&D investment projects. However, this 22See Michael C Jensen, "Takeovers: Their Causes and Consequences," Joumal of Ec~7nuc Perspectives, vol. 2, no. 1, Winter 1988. 23Jensen, 1988, p. 28. 24See Margaret M. Blair and Robert E. Litan, "Explaining Corporate Leverage and LBO Ac- tivity in the Eighties," mimeo (Washington, D.C.: The Brookings Institution), 1989.
134 CORPORATE RESTRUCTURING cutback in R&D would be viewed positive) by defenders of the takeover, since the resources would be shifted to more profitable activities. One important point made by Jensen is that the example created by hostile takeovers against unprofitable investment of free cash flow by management serves to discipline other managers observing these events. The hostile takeovers of some firms may induce other firms not subject to an actual takeover attempt to respect stockholder interests, and cut insufficiently profitable activities. SOME TENTATIVE ASSESSMENTS Because technology investments are central to our future standard of living, the impact of corporate restructurings on R&D is an important issue. More work on the interactions between distinct types of mergers, sell-offs, and research and development activity is clearly required before we can have any real confidence in our understanding of these linkages. 1b date, empirical work on the response of R&D investment to recent corporate restructurings has largely been confined to comparing investment ratios among broad groups of firms. At least one set of arguments suggests that the crucial relationship is between financial leverage (which may vary significantly among merged firms) and R&D investment. For free-cash-flow theorists, it is the hostility of the takeover which is the key to understanding its erects. Both per- specthes suggest that different types of mergers and acquisitions should be distinguished from one another for the purposes of analysis. Another set of arguments suggests that some significant effects of restructuring are felt outside the acquired firm. These issues complicate interpretation of existing work, and merit further empirical study. Nonetheless, a few preliminary judgments may be ventured. First, there is little evidence for the proposition that "restructuring" reduces research intensity in affected firms. Second, if LBOs generally lead to reductions in sales by acquired firms, it is entirely possible that declines in the absolute level of R&D spending might occur despite constant or increasing R&D intensity. Third, R&D activity on the part of firms taken over in recent LBOs is probably too limited to have had much more than a marginal direct impact on aggregate industrial R&D. The possible indirect impact on the behavior of other companies remains unexplored. The similarity of company R&D spending trends abroad to those in the United States suggests other influences may be more important. Even among companies undergoing no restructuring, R&D spending slowed significantly. Those concerned with the recent decline in company-funded
INDUSTRL4L RESEARCH 135 R&D growth would be well advised to investigate other explanations as well. Finally, the ultimate test of the impact of the recent LBO wave has yet to be faced. That will come when the United States economy enters a significant recession.
